Deposition in a glacial or glaciofluvial environment is supported by the presence of beds of matrix-supported gravel and of broken, previously rounded quartz grains. However, trial pits have shown that the gravels contain a range of relatively undisturbed sedimentary structures and glacial disturbance is likely to have been restricted in its effects. Transport by meltwater is more difficult to refute if the widespread breakage of quartz grains at depth within the Buchan Ridge Gravel Member is not a result of the drilling process.

Despite much investigation, the age of the member remains uncertain. The gravels are clearly post-Cretaceous in age, as they contain flint, and they predate the Late Pleistocene, as they are overlain by till (Whittington et al., 1998). The degree of postdepositional alteration is striking and certainly far exceeds that shown by the felsite-rich Leys Gravel of Middle Pleistocene age found at Kirkhill (Hall and Jarvis, 1993a). The constituents of the Buchan Ridge Gravel Member were derived in part from a highly weathered land surface and the kaolinitic style of weathering has been related to warm climates that prevailed before the Pliocene (Hall, 1985; Chapter 3). However, absolute dating of the weathering of bedrock and of the gravels is lacking and kaolinitic fluvial deposits are known from Europe to span the whole of the Palaeogene, Neogene and the earliest Pleistocene. Significant landscape modification has occurred since deposition, with probable uplift and topographic inversion (Figure 20), but the timing of these events is unknown.

If the Buchan Ridge Gravel Member is fluvial, rather than glaciofluvial in origin, then the upper parts of the former river catchment have been destroyed by erosion. Furthermore, the energy required for boulder transport is likely to have come from steep river gradients resulting from regional tectonic uplift and tilting towards the North Sea (Chapter 3). Uplift such as this brought about deposition of kaolinitic sands in the inner Moray Firth both in the Eocene and Pliocene (Andrews et al., 1990). If the gravels are marine in origin then deposition occurred during a period when sea level fluctuated over 75 m and all contemporaneous fine-grained sediments have been destroyed by erosion, which is perhaps unreasonable. Most would agree that the wide elevation range and localised distribution of deposits of the Buchan Ridge Gravel Member suggest that the individual bodies were deposited over a considerable time period. There is a distinct possibility that beach deposits formed during the post-Cretaceous marine regression were later fluvially reworked.

Channel-fill deposits at the mouth of the River Spey

A borehole survey of sand and gravel resources around Garmouth, at the mouth of the Spey (Aitken et al., 1979) revealed the existence of a deep rock-cut channel extending below present sea level. The channel is filled with medium- and fine-grained greenish grey quartzose sands. Thin seams of pale greenish grey, sandy silty clay are also present, containing illite with kaolinite; 20 m or more of glaciofluvial gravel cover the deposits. The channel is interpreted as a possible preglacial channel of the River Spey. An enigmatic kaolinitic ‘channel-fill’ deposit was also found at depth in a borehole sited on the ‘Mosstodloch’ glaciofluvial terrace on the Muir of Stynie [NJ 3284 6152] (Aitken et al., 1979).

Deep weathering and soil development

An important feature of the landscape of the lowlands of north-east Scotland is the extensive development of deeply weathered rock (Wright, 1997). Although deep weathering is known from other parts of Scotland, it tends to be relatively restricted in its depth and distribution. In Buchan, in particular, it is common to find extensive areas with few, if any fresh rock outcrops. Boreholes show the weathering commonly extends to depths of 10 to 20 m and may exceed 50 m. The characteristics of these saprolites (in situ weathered rock) and weathering profiles are now quite well known. Investigation of the deep weathering is important for understanding the long-term evolution of the landscape, the variable impact of glacial erosion and the origins of soils (Chapter 3). The deep weathering is also of importance in engineering geology, for the weathering reduces rock mass strength in a complex manner, varying in depth and degree of alteration over short distances (Chapter 2). Finally, the disaggregation that accompanies weathering allows certain types of weathered granular rock to become a source of aggregate (Appendix 2).

Distribution

Deep weathering is common throughout the lowlands of north-east Scotland, but varies in its frequency and depth (FitzPatrick, 1963; Hall, 1986). Information on weathering sites comes from exposures, both long-term and temporary, and from boreholes. Commonly, weathered rock has been wrongly identified as drift in borehole logs. A point distribution map of weathering sites (Figure 23) is useful in showing the widespread distribution of the 500 or so sites known in north-east Scotland. However, data on the degree of rock weathering are unevenly distributed, and it is likely that many areas that lack fresh outcrops and with a thin drift cover also conceal extensive zones of weathered rock. One such area lies around the headwaters of the River Ythan, where smooth, valley-floor slopes developed across pelite are mantled by periglacial slope deposits (Galloway, 1958), but may also hide pockets of deep weathering. In addition, a point distribution map may give a false impression of the importance of weathering in an area, as small pockets of weathered rock commonly occur in areas of dominantly fresh rock. Nonetheless, it is clear that rock type is a basic control over the distribution and degree of weathering.

Deep weathering is most commonly developed on biotite granites, basic igneous rocks and in feldspathic psammite bands within quartzites (Hall, 1986). It affects a wide range of metamorphic rocks, but is comparatively uncommon on pelites, although data on weathering in areas underlain by slate is sparse. Deep chemical weathering also has not been noted widely on the Devonian sandstones and conglomerates (Hall, 1986), although geophysical surveys suggest weak alteration on the Turriff outlier down to 10 m in places (Ashcroft and Wilson, 1976). Decomposed conglomerates occur in the Elgin district (Peacock et al., 1968) and between Stonehaven and Inverbervie (Auton et al., 1990).

Depths of weathering are recorded in deep quarries and in boreholes. Weathering to depths of more than 5 m is common in eastern Buchan: for example in the quartzites at the western edge of Mormond Hill [NK 950 568], in quartz-mica psammite at Northseat [NJ 930 408], near Auchnagatt, and in granites at Cairngall, Longside [NK 053 471] and at Hill of Longhaven [NK 084 423]. Weathering to a depth of over 60 m is recorded in a borehole south-west of the Hill of Dudwick [NJ 979 378] and in this part of the Buchan, depths of weathering exceed 20 m in numerous boreholes (Hall, 1985). A similar depth is recorded from a fracture zone in the Peterhead Granite (Edmond and Graham, 1977). Leaching of Permo–Triassic sandstone reaches depths of more than 50 m near Lossiemouth (Peacock et al., 1968). Seismic profiles indicate alteration to depths of 50 m on the sheared margins of the Arnage and Insch basic masses (Leslie, 1984). Weathering depths of 10 to 20 m or more are known from boreholes from the Huntly, Knock and Maud basic masses, the Skene and Peterhead granites and schists near Aberdeen, but these depths tend to be exceptional and normally correspond to fracture zones (Hall, 1983, 1986). Similar depths of weathering are expected to occur along the northern margin of the Insch depression and beneath the floor of the Drumblade depression (Ashworth, 1975).

Many sections and boreholes reveal rapid lateral variations in weathering depths. In sections at Northseat (Hall, 1986; (Map 6)), Cairngall (Hall, 1983), Kirkhill (Hall, 1983; (Map 7)) and Ythanbank (Hall and Sugden, 1987; (Map 6)), zones of weathering lie adjacent to fresh outcrops, yet extend more than 15 m below the surface. Very closely spaced boreholes at Cuttlehill [NJ 499 473], near Ruthven, and Lumphart [NJ 769 269], near Oldmeldrum, demonstrate variations in depth of more than 20 m over distances of less than 100 m. (Hall, 1983, fig. 5.3.ii). Boreholes in the Knock basin in the area east of Ruthven show local development of 20 to 30 m of weathered basic igneous rock, but again weathering depths are highly variable (Hall, 1983, fig. 12.6.iv). These variations are associated with fractured or hydrothermally altered rocks and with sequences of rock of widely different resistance. Rocks of more homogeneous composition tend to show more gradual changes in weathering depths. In general, however, lateral variations in weathering depths of over 5 m can be expected over short distances in most areas.

Given the irregularity of many rockhead profiles, stripping of saprolite should have resulted in slopes with upstanding rock knobs and ribs. In fact, slopes generally show a marked homogeneity in weathering patterns. Lowland areas north of the Don valley show gentle convexo-concave slopes with few rock outcrops. In section, smooth slopes are developed across thin drift covers resting on rocks at various stages of alteration. Bosses of fresh rock have protected adjacent pockets of weathering from erosion, yet these ‘risers’ commonly fail to have any topographic expression. This discordance seems to be largely a result of glacial and periglacial activity. Although glacial erosion has been generally moderate to low in its intensity, it has been sufficient to remove small, upstanding rock forms. Frost shattering has reduced rock knobs and solifluction has contributed to the smoothing of slopes (Hall, 1986).

Characteristics of the weathering profiles

Three basic types of weathering profile have been identified in north-east Scotland (Hall, 1983; (Figure 24)). The first, type A, is characterised by thorough disaggregation, with a gradual downward increase in coherence to hard rock. These profiles are typically associated with coarse-grained granites and other homogeneous or closely fractured rock types (Plate 6). A good example is seen at Hill of Dens quarry (Map 7) on the Hill of Longhaven (Hall, 1993c). The second type of profile (B) shows an upper zone of thorough disaggregation and a lower zone of corestone development, where kernels of fresh rock are isolated by penetrative weathering along joints (Plate 7). Profiles with corestones are typical of most widely jointed acid and basic igneous rocks. Good examples were formerly exposed in quarries to the south-east of New Pitsligo. Corestones occur widely on the Knock, Huntly and Insch basic masses and to the east of the Strichen granite. The third type of profile (C) is characteristic of weathered metamorphic rocks. At the base of the profile the metasedimentary rocks break up into angular blocks separated by thin seams of decomposed rock. The blocks become progressively reduced in size up-profile and are surrounded by weathered rock. The heterogeneous composition of many metamorphic rocks in the district means that zones of blocky disintegration persist even in the upper zones of weathering profiles.

Borehole records indicate that the transition from fresh to hard rock at the base of weathering profiles is variable in character. Around one third of boreholes examined show a well-developed basal surface of weathering. This abrupt change from weathered to fresh rock is most common where weathering is shallow. However, around half of the boreholes show that fresh rock is overlain by alternating bands of decomposed, weakened, shattered and fresh rock. Such transition zones may persist over depths of 15 m or more. In about one in five boreholes the transition from weathered to fresh, more competent rock is gradational over several metres. Overlying glacial and periglacial deposits generally truncate the weathering profiles, but exposures commonly show pockets of weathering protected by adjacent bands of hard rock.

In many deep exposures, such as Northseat [NJ 930 408] (Hall, 1986; (Figure 25); (Map 6)), there are few signs of soil horizon development within the weathering profiles. Colour, grain size and clay mineralogy all change gradually down the weathering profile. Downward translocation of clay is indicated by the presence of clay coatings along joint boundaries and around blocks. However, in more altered saprolites, near-surface zones of kaolinisation and reddening are replaced at depth by less weathered materials. The lateral variations in the degree of weathering of the Moreseat Sandstone on the Moss of Cruden have been interpreted as reflecting the differential truncation of a deep weathering profile (Hall and Jarvis, 1994). Reddening (rubefaction) is associated with saprolites developed on the Peterhead Granite where it probably results from the presence of iron minerals in the granite. Reddened quartzitic saprolites occur at several sites including Sunnyside [NJ 983 372], Drinnies Wood [NJ 973 497] and Howe of Dens [NJ 975 805].

At the base of weathering profiles, staining by manganese or manganese-iron oxides is widespread, especially on rocks rich in mafic minerals. The presence of manganese dioxide, in particular, has been interpreted as an indicator of present or former hydromorphic conditions at, or beneath, the water table (Koppi, 1977). This observation is significant as many weathering profiles that contain these minerals lie well above contemporary water tables, for example at Northseat. The relationship between weathering profiles and water tables is poorly understood, although normally the basal surface of weathering provides a permeability contrast, which corresponds with the level of the water table. On parts of the Buchan Ridge, where weathering depths exceed 15 m, water is not met for 10 m or more below the surface. Several deep quarries in partly weathered rock, such as Cairngall and Hill of Dens, have floors above the local water table. Here it appears that the free-drainage offered by the disaggregated rock allows a lowered water table. However, weathering is known to penetrate deep below the water table in some boreholes indicating that weathering processes must occur locally in the deeper zones of groundwater circulation.

Weathering patterns

At scales of 1 to 10 km², rockhead profiles show three main features:

• alternating low risers and depressions
• linear zones of deep alteration
• scarp-foot weathering zones

Each of these distinctive weathering patterns may have topographic expression.

In the lower Ythan valley, topography appears to reflect the progressive stripping of saprolites from a gently undulating rockhead surface (Hall, 1986, fig. 4). The intensity of glacial erosion increases towards the axis of the valley, where ice-moulded bedrock surfaces are locally developed (Hall and Sugden, 1987). The formation of the Barra basin, one of a series of small basins along the lower Don valley (Hall, 1983, fig. 12.5.ii), has involved the removal of weathered rock by a combination of fluvial and glacial processes. The lowest part of the basin floor corresponds with a zone of intense shearing and is underlain by deeply weathered rock. In contrast, the adjacent mass of Hill of Barra (Map 8) is composed of dislocated, but relatively unsheared, basic cumulates and is comparatively little weathered (Ashcroft and Munro, 1978).

Zones of deep linear alteration exist at a variety of scales. Trenches up to 300 m wide in the weathering front at Crichie [NJ 970 440] (Map 6) and Minnes [NJ 944 237] (Map 9) appear to relate to localised fracturing. Both have been partly excavated by meltwater. Boreholes for a bypass site investigation south-west and west of Keith, to the west of the district, showed that weathering beneath the floor of Strath Isla penetrated to depths exceeding 30 m in schists. Adjacent metalimestones showed cavities, presumably resulting from solution, down to depths of 20 m (Ove Arup and Partners, personal communication, 1993). These linear zones of alteration generally reflect fracturing or shearing (Leslie, 1984).

Scarp-foot weathering zones occur widely. The acceleration of weathering at the foot of water-gathering slopes has led to the development of deep weathering zones transverse to the scarp-foot, itself generally located on a geological discontinuity. These zones may be found at various stages of development dependent on the present drainage (Figure 26).

The key controls on the distribution of weathering at the small scale are rock type, structure and topography. At the regional scale (greater than 10 km²), the varied intensity of glacial erosion is an important factor. Identification of regional weathering zones in north-east Scotland (Hall, 1986; (Figure 27)) suggests that weathering is most sparsely distributed in areas of more vigorous ice flow. The most common occurrences of fresh rock typically occur in areas that were eroded by the coastal ice streams and by ice flowing down the Dee valley. This contrast is particularly apparent on a transect northwards from Aberdeen to the Buchan Ridge, where there is a close inverse correlation between the depth and frequency of weathering and the occurrence of icemoulded bed forms (Hall and Sugden, 1987; (Figure 21)).

These contrasts probably reflect differences in the basal thermal regimes of the ice streams. The coastal areas and the Dee valley were traversed by warm-based ice, capable of significant erosion. In contrast, inland areas were covered by ice that was usually cold-based and frozen to its bed, allowing only limited erosion.

Saprolite characteristics

Saprolites in the lowlands of north-east Scotland vary widely in grain size, geochemistry and clay mineralogy. Grain size within weathering profiles generally does not change systematically with depth and near-surface samples commonly have a similar texture to those at depth (Hall, 1983). Analysis of grain size for 78 exposures of saprolite shows that grain size characteristics vary according to rock type and degree of chemical alteration. It is also significant that grain size characteristics vary widely within a section. At the initial stages of weathering, granitic, basic igneous and metasedimentary weathering profiles form distinct granulometric populations (Figure 28) and the resultant saprolites possess different mechanical properties. On granites, initial disintegration forms granular grit, with median grain sizes above 1000 µ and very limited development of fines. This style of disaggregation is well displayed on the coarse Peterhead Granite at Hill of Dens quarry on the Hill of Longhaven ((Figure 28); (Map 7)). At that site, total fines (silt and clay) content is usually less than 5 per cent. The granulometry of such immature granite saprolites closely reflects the dimensions of the minerals in the fresh granite, with fine-to medium-grained granites breaking down into saprolites with median grain size of 400 to 500 µ and coarse-grained granites providing median grain sizes of 1200 to 1500 µ (Hall, 1983). On medium- to coarse-grained gabbro and norite, initial breakdown also produces granular sands (Figure 28). Fines contents are low (generally less than 10%), and median clay content (0.7%) is below that of the weathered granites (2.8%). Quartz psammites produce varied grain size characteristics that reflect the diverse lithology of the weathered, less resistant rock bands within the host rocks and the advanced degree of alteration at these sites. Other metasedimentary rocks produce saprolites with elevated fines (27.4%) and clay (4.5%) contents (Figure 28). In comparison with overlying soils developed on the saprolites, the soils are greatly enriched in fines and deficient in coarse sand.

The grain size characteristics of saprolites developed in the initial stages of breakdown on different rock types indicates a particular style of disaggregation (Figure 24). The initial breakdown of granitic and gabbroic rocks produces a polymineralic granular gruss (or sand). This process has been termed ‘granular disintegration’ (Basham, 1974). Metasedimentary rocks first break into angular blocks, a process termed ‘blocky disintegration’ (Basham, 1974). Alteration then penetrates along the fractures and produces thin seams of decomposition containing significant amounts of fines. These contrasts may reflect the ways in which residual stresses are released from the rock during weathering and unloading. In finer grained metasedimentary rocks, the opening of fractures involved in blocky disintegration largely relieves residual stress. In igneous rocks, stress retained after jointing is relieved by the propagation of micro-cracks, possibly after the modest expansion of certain minerals, notably biotite, in the initial stages of weathering.

With increasing alteration, the granulometric differences between rock types become less clear, although metasedimentary saprolites continue to have a relatively high fines content. The size of the clay fraction (< 2 µ) is a useful indicator of the degree of chemical alteration. Advanced alteration gives saprolites on all rock types with a clay content of more than 6 per cent, reaching 20 to 25 per cent on metasedimentary rocks.

Weathering types and age

Two distinct types of saprolite or weathered rock mantle are identified in north-east Scotland (Hall, 1985). A more evolved weathering type, clayey gruss, shows an elevated clay content (>6%), with clay mineralogy dominated by kaolinite and illite, and with small amounts of hematite. Within the district, this degree of weathering is known from only a small number of sites, mainly on the high ground of central Buchan (Hall, 1985; (Figure 22)). It is represented by kaolinisation of clasts within, and bedrock beneath, the deposits of the Buchan Gravels Formation. It also occurs within the Mormond Hill Quartzite. Kaolinised granite containing a hematite/layer-silicate clay mineral, macaulayite, is found at the foot of Bennachie [NJ 693 245] (Wilson et al., 1981, 1984; Koppi, 1977). Kaolinistion affects the pelites beneath the quartzite gravels at Windy Hills (Figure 22) and silicate clasts within the gravels themselves (Koppi and Fitz-Patrick, 1980; Hall et al., 1989). Koppi (1977) also describes a highly weathered feldspar-biotite psammite from Clashindarroch Forest [NJ 435 305], just west of the boundary of the district. Comparison with the mineralogy of North Sea sediments suggests that highly kaolinitic weathering mantles formed prior to the Pliocene in north-east Scotland under warm humid conditions (Hall, 1985).

The vast majority of the saprolites in the district fall within the gruss weathering type. The saprolites are dominantly sandy, with limited development of fines, and the clay mineralogy is closely controlled by rock type. Granitic saprolites contain kaolinite-mica clay mineral assemblages (Hall et al., 1989). Basic igneous saprolites contain a wide range of clay minerals. For example, weathering of the ‘Insch Gabbro’ has left feldspar and hornblende largely unaffected. Pyroxene alters initially to iron oxides and later to vermiculite, while biotite weathers to hydrobiotite and vermiculite and, locally, to kaolinite and gibbsite (Basham, 1974). Acid metamorphic rocks give kaolinite and mica clays, but an increasing content of primary ferromagnesian minerals leads to a reduction in kaolinite and an increase in smectite content (Hall et al., 1989). Altered metalimestones are dominated by smectite, partly inherited from the parent rock (Wilson et al., 1968). Grusses are thought to have formed in north-east Scotland under the humid temperate environments of the Pliocene and warmer periods of the Pleistocene (Hall, 1985).

Chapter 5 Quaternary Period

Global record of climate change

In the last two decades significant advances have been made in the understanding of Quaternary environmental change in Scotland. This is largely because the timing and pace of climatic change in Scotland can be viewed now as part of the broader pattern affecting the whole North Atlantic region (Boulton et al., 1991). Evidence of global and regional events is found in deep-sea sediment cores, in cores of ice through the Greenland ice sheet and from extended sequences of interbedded loess and organic deposits from continental Europe. The fluctuations in climate appear to be driven by minor orbitally controlled variations in solar radiation that are amplified through complex interactions between the atmosphere, oceans, ice sheets and global tectonics.

The onset of glaciation in the Northern Hemisphere probably began in the late Miocene with a significant build-up of ice over southern Greenland, although progressive intensification did not begin until 3.5 to 3 Ma when that ice sheet expanded into northern Greenland (Maslin et al., 1998). This suggests that the Scottish climate had begun to deteriorate long before the beginning of the Quaternary Period. The Quaternary is presently defined as beginning at about 1.77 Ma (Shackleton et al., 1990), but studies of ice-rafted debris in ocean floor sediments in the North Atlantic and elsewhere have shown that significant climatic deterioration occurred some 600 ka earlier when the mid-latitude continents of the Northern Hemisphere first became glaciated (Shackleton et al., 1984; Ruddiman and Raymo, 1988). Many now agree that the beginning of the period should be placed either at 2.4 Ma, as accepted here (Figure 6), or at 2.6 Ma (Funnell, 1995) in order to take into account the above evidence and recent advances in micropalaeontology, magnetostratigraphy and climatostratigraphy (Mauz, 1998). Although no direct evidence of glaciation at this time has been found on the Scottish mainland or neighbouring offshore shelves, the relatively minor decline in temperature needed for glaciers to develop, suggests that they did so, at least in the western Highlands.

Deep ocean record

The frequency, rapidity and intensity of climatic change are a key feature of the Quaternary. Studies of the isotope geochemistry and micropalaeontology of deep ocean-floor sediments have revealed that climatic conditions have fluctuated continuously throughout the period and that climate systems have switched rapidly between interglacial and glacial modes (Figure 29). At least 50 significant ‘cold–warm–cold’ oscillations have been recognised. Many theories have been put forward to explain the initiation of Northern Hemisphere glaciation (Maslin et al., 1998), most involving changes in atmospheric composition (such as caused by increased volcanism) or in total solar radiation. The initial cooling that began during the Neogene has been attributed to the slow changes in the global configuration of the continents as a consequence of sea-floor spreading. These include the emergence of the Panama Isthmus and the deepening of the Bering Straits, both of which had pronounced affects on the ocean circulation patterns (Raymo, 1994). The cooling has also been attributed to the uplift of high mountain ranges such as the Tibetan Himalayas and the Sierra Nevadan and Coloradan mountains of North America, causing perturbations of the upper atmosphere and subsequent climatic changes (Ruddiman and Kutzbach, 1991). Uplift of the Himalayas also may have resulted in a massive increase in chemical weathering during the late Cainozoic, leading to increased sedimentation of calcium carbonate and atmospheric depletion of carbon dioxide. Global cooling, the inverse of the ‘greenhouse effect’ would thus occur (Raymo and Ruddiman, 1992; Raymo, 1994). However, none of these mechanisms can wholly explain the rapid intensification of glaciation observed in the deep ocean record at about 2.7 to 2.5 Ma (Maslin et al., 1998).

The driving force of climatic change during the Quaternary appears to be the long-term cyclical variation in solar energy. Cyclical changes in solar insolation appear to be associated with the Earth’s orbital periodicities, rather than on long-term changes in energy output of the Sun, although the latter is difficult to disprove (Shackleton and Opdyke, 1973). There are three significant orbital (‘Milankovich’) cycles, caused by:

The periodicities of these cycles are roughly 23 ka, 41 ka and 100 ka, respectively. The climatic fluctuations that occurred during each major glacial–interglacial cycle have been attributed to the first two cycles (Imbrie and Imbrie, 1979), but now it is generally agreed that they alone could not have produced the magnitude or rapidity of the documented changes. Maslin et al. (1995) suggested that the increase in the amplitude of orbital obliquity cycles deduced from the deep ocean record from 3.2 Ma onwards may have increased the seasonality of the Northern Hemisphere, thus initiating the long term cooling trend. The subsequent sharp rise in the amplitude of precession between 2.8 and 2.55 Ma may have forced the rapid intensification of glaciations that began then.

The primary Milankovich fluctuations of solar insolation must have been amplified substantially by additional factors involving physical, biological and chemical interactions and ‘feedback loops’ between the atmosphere, oceans and ice sheets. For example, complex changes in the surface and deep water circulation patterns of the oceans, and the concentrations of carbon dioxide and other ‘greenhouse’ gases in the atmosphere all play a crucial role (Broecker and Denton, 1990). Of particular importance to the British Isles are changes in the position of the Gulf Stream. This northward-flowing current of warm surface water is compensated by the return southwards of cold, dense water at depth. Sudden changes in this circulation pattern, the so-called ‘North Atlantic conveyor’, may have had a major impact on climate (Skinner and Porter, 1995; Figure 30). For example, large volumes of fresh water released during rapid warming events may have temporarily ‘switched off’ the conveyor leading to cooling in north-west Europe (Lagerklint and Wright, 1999).

The ‘SPECMAP’ deep ocean oxygen isotope record (Figure 6) indicates that during the early Quaternary, up until about 760 ka, each dominant warm–cold cycle lasted about 40 ka (Ruddiman et al., 1989). Ice caps in Greenland, Alaska, Iceland and Scandinavia expanded to the coast and glaciers probably developed in the western Scottish Highlands at high elevations (Clapperton, 1997). Iceberg drop-stones found in sediment cores taken offshore from northwest Europe provide indirect evidence of these early glaciations (Holmes, 1997; Sejrup et al., 2000). Following a major change at about 760 ka, the build-up of much larger ice sheets in the Northern Hemisphere occurred, and to date there have been seven major glacial–interglacial cycles (Figure 29). Each ‘glacial’ episode lasted between 80 and 120 ka and was followed abruptly by an ‘interglacial’ lasting 10 to 15 ka. The rapid deglaciations are defined as ‘Terminations I–VII’ (Broecker, 1984; (Figure 29)). The glacial periods included long, cold intervals, termed stadials, and less cold, and even warm, intervals lasting for a few thousand years termed interstadials. Most terrestrial evidence preserved in Scotland relates to the last glacial–interglacial cycle (Holocene, Devensian and Ipswichian stages), but older deposits are preserved locally (Figure 7).

The SPECMAP timescale of Imbrie et al. (1984) has been the cornerstone of Quaternary palaeoclimate studies. It assumes that the climatic cycles observed in deep ocean sediment records were driven by Milankovitch orbital parameters. Although this assumption has been challenged (Schrag, 2000), the validity of the timescale has been confirmed independently by Raymo (1997). She hypothesises that the general intensification of the major glaciations during the Quaternary results from long-term reduction of atmospheric carbon dioxide levels by tectonic processes (as shown by the gently inclined straight line on (Figure 29)). This gradual weakening of the ‘greenhouse effect’ has caused cold periods to become colder, allowing ice sheets to expand in regions that were previously too warm. Raymo’s (1997) model shows that major glaciations (coloured areas on the insolation curve of (Figure 29)) only occurred when both ‘obliquity’ and ‘eccentricity’ Milankovitch cycles reinforced themselves, causing long intervals of low summer insolation in northern lattitudes. Furthermore, her model

Greenland ice core record

New, high resolution evidence derived from cores taken through the Greenland ice sheet near its summit (Figure 31), demonstrate that the Earth’s climate has been much more variable during the past 250 ka than previously thought (Dansgaard et al., 1993). Dramatic climatic changes have occurred, both on the millennial scale and over just a few years. Some 24 interstadial intervals have been identified in the last (Devensian) glacial stage alone, compared with the five or six that were recognised previously from the pollen record and formalised in the north-west European and British stratigraphy (Mitchell et al., 1973; (Figure 7)). Each of the 24 interstadials within the Devensian started with abrupt warming and typically cooled over a period of 1 to 3 ka, ending in a cold event (Broecker, 1994). These periods, known as Dansgaard–Oeschger (D-O) cycles, are sometimes grouped together within longer, Bond cycles, during which temperatures also declined gradually (Bond et al., 1993; (Figure 32)). It is important to note that even the milder interstadials in Greenland had a climate in which average temperatures were five to six degrees colder than on the summit of the Greenland ice sheet today (Dansgaard et al., 1993). Intervals when glaciers were probably present in Scotland during the past 25 ka are indicated in (Figure 31)).

Many Bond cycles appear to culminate in a Heinrich event (Figure 32). Originally it was concluded that temperatures rose substantially within a decade or so following these events (Bond and Lotti, 1995), but there is growing evidence that they are actually the result of extremely rapid warming (Lagerklint and Wright, 1999). Heinrich events, as originally defined, were massive discharges (‘armadas’) of icebergs from the Laurentide ice sheet, recognised in deep ocean cores in the North Atlantic by the presence of ice-rafted debris (IRD) at intervals in the sediment sequence (Heinrich, 1988; (Figure 31)). These horizons were thought to result from the sudden mass wastage, via iceberg calving, of the Laurentide ice sheet every 11 ka or so, possibly as a result of it having grown too large to sustain itself and resulting in catastrophic collapse. Such ‘binge-purge’ cycles are thought by some to have played a pivotal role in driving world climate variability (Broecker, 1994; Alley and MacAyeal, 1994). More recently, Heinrich events have been identified in North Pacific cores (Hicock et al., 1999) and ice sheets in Iceland, Britain and Scandinavia are now known to have contributed IRD during Heinrich events (Frontal et al., 1995; Sejrup et al., 2000). It therefore seems unreasonable that all of the ice sheets ‘binged and purged’ in harmony unless there was some independent forcing mechanism (Bond and Lotti, 1995; Kotilainen and Shackleton, 1995). Furthermore, it appears that there was a synchronous deposition of ice-rafted layers in the Nordic seas and North Atlantic (Dowdeswell et al., 1999) and glacial events in north-west Europe may actually have driven events across the Atlantic, not the other way around (Grousset et al., 2000). One suggestion is that Heinrich events are triggered every 6100 years as a result of direct changes in solar radiation (Lehman, 1996), their effects being greatest during the longer cold periods when glaciers, ice sheets and tidewater glacier margins were most extensive.

Heinrich events probably led to ‘armadas’ of icebergs in the seas around the British Isles. The release of large volumes of glacial meltwater and icebergs from North America possibly ‘switched off’ the ‘North Atlantic Conveyor’ for a while leading to cooling of the north-east Atlantic and north-west Europe (Lagerklint and Wright, 2000). Minor re-advances of ice streams flowing through the major firths of the west and east coasts of Scotland are possibly due to such events (Clapperton, 1997). The Heinrich events of the North Atlantic occurred at about 10 ka BP (H0), about 14.2 ka BP (H1), 21.4 ka BP (H2), 26.7 ka BP (H3), 34.8 ka BP (H4) and 47.2 ka BP (H5) (Chapman et al., 2000; (Figure 31)).

The future challenge is to determine to what extent the various fluctuations in climate have left a recognisable signature in the geological record of sediments, landforms, and fossil floras and faunas (Alverson and Oldfield, 2000 and papers cited therein). The offshore stratigraphical record obtained over the last two decades has already significantly augmented the relatively incomplete terrestrial record, particularly for the Early and Middle Quaternary. Onshore, recent integrated studies at the few sites at which organic materials are well preserved, involving sedimentology, the analysis of pollen, beetle remains and marine molluscs, and the use of a range of dating techniques, have also provided much new information on past Quaternary environments.

Offshore stratigraphical record

An understanding of Scotland’s evolving landscape requires knowledge of how the position of the coastline has varied throughout the Quaternary in response to changes in climate and relative sea level. The current position of the coastline is only a transient feature. For example, most of the southern North Sea was low-lying land like the Netherlands until the first major glaciation. Moraines are preserved at the shelf-edge north-west of mainland Scotland, more than 60 km west of the nearest coastline (Holmes et al., 1993; Stoker et al., 1993). Conversely, marine deposits are preserved more than 10 km inland from the modern coastline at the head of the Moray Firth and at least 6 km inland near Elgin. For the following account, the modern coastline is taken as the boundary separating onshore and offshore Quaternary sedimentary deposits.

Compared with the fragmented onshore Quaternary sedimentary record, that offshore is considerably more complete, especially towards the shelf-edge where large submarine fans formed during major glacations (Table 1).

The offshore sequence is generally thicker, more extensive and can be correlated regionally using seismostratigraphical methods (Stoker et al, 1985; Holmes, 1997). At least five major glacial episodes have been recognised in the North Sea basin, within a sequence that is dominanted by deltaic, low salinity cold-water marine and glaciomarine conditions (Sutherland, 1984a). The thickest and most complete sequence (more than 500 m) is preserved in the Central Graben of the North Sea, which has subsided tectonically throughout the Quaternary. At least ten till units are present within the Norwegian Channel with six or seven extending to the distal shelf break (Sejrup et al., 2000). An unconformity occurs at the base of the Quaternary and the magnitude of the hiatus increases away from the graben towards north-east Scotland (Andrews et al., 1990; Johnson et al., 1993; Gatliff et al., 1994). In the Moray Firth, most Quaternary sediments rest on Paleocene, Cretaceous, and Jurassic strata, with successively older formations cropping out as the modern shoreline is approached (Andrews et al., 1990, fig. 2). This style of basin–tectonic subsidence was initiated in the mid-Miocene, and was coeval with Miocene, Pliocene and Quaternary uplift of the mainland, which resulted in increased sediment supply to offshore areas (Jordt et al., 1998; Chapter 3). The truncation of Quaternary, Neogene and older sediments towards the margins of the North Sea basin is consistent with a hypothesis involving cycles of sub-glacial and marine erosion removing terrestrial sediments and contributing to long-term isostatic uplift of the mainland and near-shore areas (Japsen, 1998).

The following account is based largely on Holmes (1997) and the BGS regional reports covering the Moray Firth (Andrews et al., 1990) and the central North Sea (Gatliff et al., 1994). Details of map coverage are given in Information Sources. Interpretations are based mainly on seismic reflection data calibrated by borehole logs. Regionally mappable seismostratigraphical units have been established as formations and members where there is sufficient lithostratigraphical, biostratigraphical or chronostratigraphical control (Figure 33); (Figure 34), but difficulties in correlation are common. Moreover, the Moray Firth basin was investigated early in the offshore programme when seismic acquisition technology and many procedures and dating techniques were in their infancy. Hence the available data is not as high quality or detailed as that obtained more recently (Figure 35). A summary of important events is given in (Table 1).

The north-west European chronostatigraphical classification has been adopted offshore because it is more complete than the British one, especially in the Lower and Middle Pleistocene (Figure 33). The oxygen isotope stages (OIS) of Shackleton and Opdyke (1973) provide a link between the two schemes, but there is presently vigorous controversy concerning their correlation. The boundary between the Lower and Middle Pleistocene is taken here at the Brunhes–Matuyama boundary of the palaeomagnetic record of Funnell (1995; (Figure 33)).

Early Quaternary (2.44 to 0.78 Ma)

The ‘non-glacial’ Early Quaternary (2.44 to 1.2 Ma)

This part of the record is contained within the Aberdeen Ground Formation (Figure 33); (Figure 34), the oldest and thickest formation mapped offshore (Holmes, 1977; Stoker et al., 1985). The greater part of the formation off northeast Scotland comprises prodeltaic silts, clays and fine-grained sands deposited in a shallow sea (less than 50 m deep) that lay adjacent to extensive deltas occupying most of the southern North Sea basin (Stoker and Bent, 1985; The unconformable base of the Aberdeen Ground For-Jeffrey and Long, 1989; Cameron et al., 1992). The deltas mation is represented in the central North Sea by the were formed mainly by precursors of the Rhine and ‘crenulate reflector’ (Holmes, 1977, 1997). Locally it lies Thames and by a major river draining the area now largely 50 m above the Gauss–Matuyama palaeomagnetic occupied by the Baltic Sea. boundary ((Figure 33); (Table 1)). To the north and west of 58°N 01°E, the unconformity is interpreted from 3-D seismic evidence to take the form of fluvial channels. These features merge downslope into furrowed surfaces like those known to have been produced by grounded icebergs or sea-ice keels (Holmes, 1997). This evidence for ice scour within the Aberdeen Ground Formation is reinforced by the occurrence of periodic ice-rafted debris in sequences of equivalent age on the Atlantic margins off north-west Scotland (Stoker et al., 1994). Hence, the Scottish Highlands were probably glaciated to some extent during the colder stages of the Early Quaternary before 1.2 Ma.

Early Quaternary glaciations (1.2 to 0.78 Ma)

In the northern North Sea, the Shackleton Formation (Johnson et al., 1993) includes an erosion surface that separates sand-rich sediments below from mud-rich sediments above. The incoming of mud-rich sediments is reflected by a change of acoustic facies that is tentatively attributed by Johnson et al. (1993) to the first impact of regional glaciation on the depositional environment of the northern North Sea. The facies change is numerically undated, but on the basis of seismostratigraphical correlation, it predates the 1.07 to 0.99 Ma Jaramillo Subchron (Figure 33), which has been reported from the overlying Mariner Formation (Stoker et al., 1983; compare with Skinner et al., 1986).

There is more robust evidence in the Troll region of the Norwegian Channel for the ‘Fedje Glaciation’, which postdates the 1.19 Ma Cobb Mountain Event within the Matuyama Reversed Polarity Chron (Funnell, 1995; (Table 1)). On the basis of the micropalaeontology, Sr-isotopes, palaeomagnetism and amino-acid geochronology, this glaciation has been assigned an age close to 1.1 Myr (Sejrup et al., 1995). The base of the Mariner Formation is defined regionally on seismic reflection profiles by an irregular erosion surface, which is locally overlain by a diamicton (Johnson et al., 1993). The presence of the diamicton, the geometry of the reflector and other evidence have been put forward as evidence of shelf glaciation extending from Norway across the northern North Sea to areas west of Shetland (Holmes, 1997). However, there is no secure basis for such wide-reaching correlations and the early shelf glaciations may not even have affected offshore areas as far south as the Moray Firth.

The earliest evidence for glaciation of the sea bed offshore from north-east Scotland occurs in the Fladen area of the central North Sea where a 10 m-thick unit of diamicton lies just above the Jaramillo Subchron within the Aberdeen Ground Formation. As the diamicton is overlain by sediment containing microfossils with an interglacial aspect provisionally correlated with the Leerdam Netherlands Pollen Stage, Sejrup et al. (1987) place the glacial event within the Bavelian Complex, between 800 and 900 ka. It has not been established whether the ice that deposited the diamicton flowed from Scotland or Scandinavia, although Sutherland and Gordon (1993) have argued that the latter is more likely. Nevertheless, the Scottish mountains were most probably extensively glaciated during that event.

Middle Quaternary (0.78 Ma to 130 ka)

Cromerian glaciations (OIS 14, 16 and 18)

The oldest known glacial deposits laid down offshore by ice flowing from the Scottish Highlands have been found in boreholes in the Forth Approaches and the Moray Firth (Stoker and Bent, 1985; Bent, 1986). There, towards the top of the Aberdeen Ground Formation, glaciomarine sediments laid down by a grounded ice sheet occur in the west, with the facies becoming increasingly more distal to the east. The deposits lie immediately above the Brunhes–Matuyama palaeomagnetic boundary ((Figure 33); (Table 1)) suggesting that the glaciation occurred in OIS 18, within the Cromerian Complex.

The top of the Aberdeen Ground Formation is cut by a succession of isolated, anastomosing and locally stacked channels, some of which may have been formed as sub-glacial or ice-marginal ‘tunnel’ valleys. The channels, together with the oldest units of sediment contained within them (Ling Bank Formation), have been correlated on the basis of regional seismostratigraphy with the widespread Elsterian glaciation of the north-west European mainland (Stoker et al., 1985; Cameron et al., 1987). However, several lines of evidence indicate that the basal sediments of the Ling Bank Formation at its type locality were laid down during a late-Cromerian interglacial (Penney, 1990; Ansari, 1992 Knudsen and Sejrup, 1993; Sejrup and Knudsen, 1993) and that they are overlain by arctic glaciomarine sediments thought to be Elsterian in age, on the basis of amino-acid dating of shells (Sejrup and Knudsen, 1993). The channelled surfaces at the top of the Aberdeen Ground Formation may well have formed in more than one glacial cycle, including the severe Cromerian glaciation of OIS 16 (Holmes, 1997). It follows that the first major interruption of the growth of deltas across the southern North Sea may have occurred in the Cromerian and not the Elsterian as commonly believed (Table 1).

Elsterian glaciation (OIS 12) and Holsteinian interglacial deposits (OIS 11 and 9).

Sediments near the top of the Ling Bank Formation at its type locality have been correlated with ‘type’ Holsteinian sections in Denmark and Germany (Knudsen and Sejrup, 1993). However, it is still widely believed that marine Holsteinian deposits, in general, directly overlie the major erosion surface at the top of the Aberdeen Ground Formation (Figure 34). There was, therefore, a widespread glaciation during the Elsterian stage, which involved ice flowing from Scandinavia into the southern North Sea (Cameron et al., 1992) and extending to the continental shelf edge north-west of Scotland (Stoker et al., 1994). The timing of the glaciation is significant because the palaeogeography of the North Sea basin and hence the climate of the surrounding lands subsequently changed profoundly. Furthermore, Scotland and the adjoining continental shelves would have been glaciated during that event, which is generally correlated with the Anglian stage of the conventional British chronostratigraphy (Table 1).

Saalian glaciation(s) (OIS 6 to 8)

Evidence for Saalian glaciations off north-east Scotland occurs in the Fisher and Coal Pit formations. The predominantly arctic glaciomarine Fisher Formation rests on a major, gently undulating unconformity that cuts across (onlaps) the Ling Bank and Aberdeen Ground formations (Figure 33), (Figure 34). The unconformity results from a marine transgression and the overlying Fisher Formation is thought to be no older than OIS 7 (Jensen and Knudsen, 1988; Holmes, 1997). A till has been identified within the Fisher Formation (Sejrup et al., 1987) and contemporaneous subglacial erosion is thought to have occurred to the north of 58.6°N in the Moray Firth (Bent, 1986). The top of the Fisher Formation is defined by another regional unconformity, but unlike the one at its base, this one is typically crenulate (Figure 34) and is believed to be the result of the Saalian glaciation (Gatliff et al., 1994). The channels are mainly infilled with glaciomarine deposits belonging to the overlying Coal Pit Formation (Sejrup et al., 1987), although some fragmentary beds deposited in warmer waters have been identified within the formation in the northern and central North Sea (Gatliff et al., 1994).

The glaciomarine basal deposits of the Coal Pit Formation are succeeded by marine sediments containing foraminiferal assemblages typical of the Eemian–Ipswichian Interglacial (Cameron et al., 1987). These in turn underlie a horizon that has been correlated with the palaeomagnetic Blake Event at the OIS 5e/5d boundary (Stoker et al., 1985).

The evidence cited above suggests that there was a regional glaciation during OIS 6 of the Saalian and possibly an earlier, more limited Saalian glaciation in the north. The glaciation of OIS 6 is thought to have affected most of Scotland, because ice flowing from the mainland affected the northern North Sea at least as far south as 56°N (Sutherland and Gordon, 1993) and it reached the shelf break to the north-west of Scotland (Skinner et al., 1986; Stevenson, 1991; Holmes, 1997). Sejrup et al. (2000) conclude that ice streams occupied the Norwegian Channel during every glacial stage between OIS 12 and 6, each one representing a regional glaciation.

Late Quaternary (130 ka to present day)

Eemian Interglacial (OIS 5e)

A record of the Eemian (Ipswichian) is contained within the central part of the Coal Pit Formation, which fills channels that were probably eroded during, or immediately after, the Saalian glaciation in the Moray Firth and central North Sea (Andrews et al., 1990; Gatliff et al., 1994; (Figure 34)). Dinoflagellate cyst assemblages, like those found at present in the North Sea, have been identified in interbedded bioturbated sand and stiff, dark grey, shelly, pebbly clay with wood fragments in the type borehole (Stoker et al., 1985). The final occurrence of Elphidium ustulatum (Todd) has been noted; this is a foraminiferid that became extinct in the North Sea by the early Weichselian (Gregory and Bridge, 1979). In southern Norway, the Eemian is now thought to have been succeeded by a prolonged period of interstadial conditions with restricted mountain glaciation (Sejrup et al., 2000).

Early Weichselian glaciation (OIS 4)

No secure evidence of early Weichselian–Devensian glacial deposits has been reported off north-east Scotland, but an ice sheet is believed to have reached the shelf break to the north-west of the Scottish mainland where it formed submarine end-moraines (Stoker, 1988, 1990; Stewart and Stoker, 1990). Sejrup et al. (2000) conclude that an ice stream occupied the Norwegian Channel after 80 ka, suggesting that a regional glaciation occurred equivalent to the Karmoy Glaciation established in south-west Fennoscandia (Figure 43). Micromorphological studies of sediments from several boreholes in the central North Sea also suggest regional glaciation at that time involving coalescing Scottish and Scandinavian ice sheets (Carr, 1998).

Middle Weichselian (OIS 3)

The Coal Pit Formation in the central North Sea includes a sequence of shelly glaciomarine clays that have been placed tentatively in OIS 3 on palaeomagnetic evidence (Stoker et al., 1985) indicating that this area was free of glacier ice during this stage. In the northern North Sea and on the West Shetland Shelf a widespread surface of marine erosion has been correlated with the OIS 4/3 boundary. It is overlain by the mainly arctic marine Cape Shore Formation, which is securely placed in the Middle Weichselian on several lines of evidence (Johnson et al., 1993; Skinner et al., 1986; Sejrup et al., 1994; Holmes, 1997). Thus the seas to the north and west of Scotland were also free of glacier ice during this stage and perhaps much of the Scottish mainland.

Indirect evidence of former near-shore environments along the southern coast of the Moray Firth is provided onshore by glacial rafts that were transported by ice during the Late Devensian. At Clava, near Inverness, rafts of highboreal to low-arctic shallow marine mud originally deposited in Loch Ness, then a fjord, are probably early to middle Devensian in age (Merritt, 1992). The deposits correlate with those of the Bø Interstadial in Norway on amino-acid dating evidence (Figure 43). Rafts of broadly similar age have been located at the Boyne Limestone Quarry, King Edward and Gardenstown sites (Appendix 1).

Late Weichselian/Late Devensian glaciation (OIS 2)

The offshore record of Upper Weichselian deposits is of particular importance in establishing models of the main Late Devensian ice sheet in northern Britain, but it is complicated and, as explained below, controversial (Figure 3). BGS mapping suggests that an ice sheet extended to the continental shelf break, and beyond, to the north and west of Scotland (Holmes, 1991; Stoker and Holmes, 1991; Stoker et al., 1993; (Figure 41)). The resultant sediments are correlated on the basis of regional seismostratigraphy with Late Weichselian (OIS 2) deposits in the northern North Sea (Johnson et al., 1993). A radiocarbon date of about 22.5 ka BP from glaciomarine deposits within the limit of glaciation on the outer shelf to the west of St Kilda (Selby, 1989) appears to be consistent with the Late Devensian glacial maximum predating 18 ka BP (Figure 41).

The important units off north-east Scotland are, from west to east, the Wee Bankie, Marr Bank, Swatchway and Coal Pit formations (Figure 33), (Figure 34). The Wee Bankie Formation (Stoker et al., 1985) lies directly off the east coast of Scotland. It has a sheet-like geometry with an uneven, ridged top and comprises up to 40 m of stiff, matrix-dominated diamicton with some interbeds of sand, pebbly sand and silty clay. The formation is generally thought to have been laid down beneath the main Late Devensian ice sheet, and its mapped eastern boundary (Figure 44) has been used in many reconstructions of that ice sheet (e.g. Sutherland, 1984a; Boulton et al., 1985; Boulton et al., 1991), although some have envisaged more extensive glaciation at this time (e.g. Ehlers and Wingfield, 1991).

The Wee Bankie Formation is replaced eastwards by the Marr Bank Formation, commonly at a low, eastward-facing scarp interpreted as a former ice-contact slope, but the two formations probably interdigitate locally (Stoker et al., 1985). The Marr Bank Formation consists mostly of sands and muddy sands of Scottish provenance with a sparse microfauna indicative of shallow, high boreal to arctic waters. It forms a sheet-like deposit up to 25 m thick resting on an extensive surface of marine planation dipping north-eastwards (Holmes, 1977). As it is traced eastwards, the basal reflector of the Marr Bank Formation becomes acoustically indistinguishable from the upper part of the adjacent Coal Pit Formation, and the two formations probably pass laterally into one another locally (Gatliff et al., 1994).

In the central North Sea, the lower part of the Swatchway Formation, at Borehole 77/2 (Figure 44), is formed mainly of glaciomarine sediments from which AMS radiocarbon dates of 22.7, 20.9 and 19.7 ka BP have been obtained on in situ mollusc and benthic foraminiferids (Sejrup et al., 1994). This evidence suggests that the area was free of grounded ice during that period. However, a diamicton underlying the glaciomarine deposits in that borehole is interpreted as a till (Sejrup et al., 1994). It contains reworked arctic benthic foraminiferids that have provided a maximum AMS radiocarbon age of 42.3 ka BP. The diamicton rests on cold marine deposits assigned to the Ålesund Interstadial of north-west Fennoscandia and it is concluded by Sejrup et al. (1994) that the till was laid down between 28 ka BP and 22 ka BP during an initial, maximal stage of the Late Devensian glaciation. This conclusion hinges on the identification of the diamicton as a till and that the foraminiferids are unlikely to be in situ. Although Sejrup et al. (1994) describe deformation structures in the diamicton, they do not discriminate between subglacial deformation and disturbance by sea-ice. However, recent micromorphological evidence has revealed brittle shear at the base of the diamicton suggesting that it is indeed a subglacial, deforming bed till (Carr, 1998). Furthermore, Sejrup et al. (2000) and Carr (1999) both conclude that the central North Sea was also glaciated in the previous Skjonghelleren Glaciation of Fennoscandia, between about 50 and 40 ka BP (Figure 43).

There are important implications to these findings, discussed below, which indicate that the Scottish and Scandinavian ice sheets reached their maximum extent in the Late Devensian prior to about 22 ka BP and that they very probably coalesced (Figure 41). At least the uppermost part of the Wee Bankie Formation postdates this early phase. Following retreat to an unknown position at about 20 ka BP, during the Hamnsund Interstadial of Norway (Valen et al., 1996), the Scottish ice sheet probably then re-advanced to the eastern boundaries of the ‘Bosies Bank Moraine’ (Bent, 1986; (Figure 44)). This event probably equates with the Tampen Glaciation of Norway, when ice re-advanced onto the shelf and an ice stream reoccupied the Norwegian Channel (Sejrup et al., 2000; (Figure 43)).

Deglaciation, the Late-glacial period and Holocene (OIS 2 and 1)

Much of the northern North Sea north of 58°N had been deglaciated by 16 to 14 ka BP, and was either subaerially exposed or inundated by a very shallow sea (Peacock, 1995). This is compatible with a maximum age of about 14.1 ka BP for the onset of glaciomarine sedimentation following retreat of ice from the Witch Ground area ((Figure 44); Sejrup et al., 1994) and at about 15 ka BP in the Norwegian Channel (Sejrup et al., 1995). The onset of deglaciation on the Hebridean Shelf has been dated to about 15.2 ka BP, predating the onset of warming in the North Atlantic (Peacock et al., 1992; Austin and Kroon, 1996). Shells within glaciomarine sediments occurring onshore near Peterhead have yielded ages of about 14.3 and 14.9 ka BP (Appendix 1 St Fergus).

Radiocarbon dates from Portlandia arctica Gray, a high arctic marine bivalve, indicate that polar water, and probably seasonal sea-ice, remained in the northern North Sea until at least 13.2 to 13.1 ka BP when warmer waters arrived. Glaciomarine and estuarine silts and clays of the Errol Formation accumulated along the coasts (Peacock, 1999), while the muddy St Abbs Formation was laid down in this polar sea off the eastern coast of Scotland (Stoker et al., 1985; (Figure 34)). Warm North Atlantic waters did not reach the north-east Atlantic and western Scotland until about 13 ka BP, when conditions changed from high arctic to boreal possibly in less than 50 years (Kroon et al., 1987; Peacock and Harkness, 1990).

A record of the Windermere Interstadial is probably contained within the Swatchway Formation, which occurs to the north-east of Buchan (Figure 34). It comprises shelly muds and sands with a mixed northern temperate to arctic microfauna (Stoker et al., 1985; Harland, 1988). It occurs more certainly in the Largo Bay Member of the Forth Formation, which is more widespread off north-east Scotland (Stoker et al., 1985; (Figure 34)). It includes up to 30 m of silty muds that become coarser grained and pebbly upwards with concomitant decreasing faunal diversity. The trends probably reflect lowering sea level and cooling seas towards the onset of the Loch Lomond Stadial. The overlying St Andrews Member was laid down as coastal sand bars in a very shallow sea during the subsequent Loch Lomond Stadial. High arctic marine fauna returned during the stadial, during which nearshore marine summer temperatures were approaching 10° below present levels (Graham et al., 1990; Peacock, 1996).

Most sea-bed sediments of Holocene age have been mapped lithologically rather than lithostratigraphically. The return of warm North Atlantic Drift waters to the Scottish seas occurred within a few decades just prior to 10 100 ka BP (Peacock and Harkness, 1990). At first sea temperatures were 2 to 3° lower than those of the present day, but a warming occurred at about 9600 BP.

Onshore stratigraphical record

Despite over a century of research, the record of events prior to deglaciation of the main Late Devensian ice sheet on land is less well understood than offshore. This is mainly because of the fragmentary nature of terrestrial deposits, the scarcity of organic material that is suitable for dating by radiocarbon methods and the relative unreliability of age determining methods beyond the range of radiocarbon dating, such as thermoluminescence (TL) techniques and amino-acid geochronology. Although more information is normally available at onshore sites compared to boreholes offshore, the absence of seismostratigraphy and palaeomagnetic chronology makes any correlation between those sites very difficult indeed. Inevitably there is controversy and the chapter concludes with a review of the published ice-sheet models for north-east Scotland. A summary of the important events is given in (Table 1).

The heirachy and description of named lithostratigraphical units are given in Chapter 8, and correlations between these units and the Oxygen Isotope Stages are presented in (Table 7). Many correlations between units are based on a ‘count from the top’ basis, which means that any reinterpretation of the status of the uppermost unit has consequences for others lower in the sequence.

Early Quaternary (2.44 to 0.78 Ma)

No Early Quaternary deposits have been identified in north-east Scotland, but mollusc shell fragments in the Kippet Hills Sand and Gravel Formation of the Logie-Buchan Drift Group (Chapter 8) have most probably been derived from the Aberdeen Ground Formation lying offshore. The Early Pleistocene age assigned to the shells on faunal grounds by Jamieson (1882a) has been confirmed by amino-acid analysis (Appendix 1 Kippet Hills).

Middle Quaternary (0.78 Ma to 130 ka)

Anglian glaciation (OIS 12) (=Elsterian)

The oldest known deposits in the district have been found at Kirkhill (Connell et al., 1982; Appendix 1). The basal fluvial or glaciofluvial sands and gravels at this site contain erratics derived from the west, as does the basal till (Leys Till Formation) occurring nearby. The two units were originally assigned to the Anglian glaciation on the basis of minimum ages for the overlying glacial and interglacial deposits using a simple chronostratigraphical model (Connell and Hall, 1984a; Hall and Connell, 1991). However, they are now assigned tentatively to OIS 8 on the basis of new luminescence dates on the Corse Gelifluctate Bed higher in the sequence (Appendix 1; (Table 7)).

Glacial erratics of Norwegian origin are relatively common within glacigenic deposits in north-east Scotland (Read et al., 1923; Bremner, 1939; Hall and Connell, 1991). Although none of these clasts occurs in a till assigned to OIS 12, it is possible that they were transported to north-east Scotland during that stage (Ehlers, 1988).

Hoxnian interglacial (OIS 11) (=Holsteinian)

The oceanic climate record of the North Atlantic suggests that there have been only two periods in the last 600 ka that were as warm, or warmer, than the Holocene, namely the Hoxnian (OIS 11) and Ipswichian (OIS 5e) interglacials (Ruddiman and McIntyre, 1976). Other interglacials were temperate (e.g. OIS 9 and 7), but probably not as warm. The sands overlying the Kirkhill Palaeosol Bed (Lower Buried Soil) at Kirkhill (Appendix 1 Site 7) contain detrital organic matter derived from the erosion of soil and vegetation formed in an interglacial climate, presumed from its stratigraphical position to be Hoxnian (Connell, 1984a; Lowe, 1984). The presence of pollen grains of pine, alder and lime appear to corroborate this presumption. However, while an OIS 11 age cannot be confirmed, the absence of significant unconformities within the Kirkhill sequence indicate that the soil and overlying organic sediment (Corse Gelifluctate Bed) may date from OIS 7, as indicated by luminescence dates reported by Duller et al. (1995). The Kirkhill Palaeosol Bed itself was originally thought to be of interglacial origin (Connell et al., 1982), but has also been compared to a cold-water gley soil formed in an interstadial climate (Connell and Romans, 1984). However, the strong development of the truncated Ea horizon, together with the presence of the mineral proto-imogite/allophane in the Bs horizon (Appendix 1 Site 7 Kirkhill and Leys), suggests the interglacial interpretation is correct. The nearest other site with a possible Hoxnian record is Dalcharn, near Inverness, where the presence of significant quantities of holly pollen implies that the palaeosol there did form in a particularly warm interglacial (Walker et al., 1992). However, there too, subsequently obtained luminescence dates suggest a much younger age (Duller et al. (1995).

‘Wolstonian’ glaciation(s) (OIS 6–10) (=Warthe/Saal/Drenthe/Fuhne)

Two distinct phases of ‘Saalian’ glaciation are thought to have affected continental north-west Europe during this period (Ehlers et al., 1991) and three regional glaciations affected the North Sea basin (Sejrup et al., 2000), but no equivalent terrestrial glacial deposits have been shown unequivocally to occur in Scotland. Several tills in north-east Scotland may date to this period, but arguments for their ‘Wolstonian’ age rely mainly on indirect evidence. This includes stratigraphical position, whether the tills are judged to have been weathered in an interglacial climate (as against incorporating materials weathered previously) and whether that weathering is judged to have occurred in the Ipswichian (OIS 5e). The Rottenhill Till (Kirkhill Lower Till) has been assigned to the ‘Wolstonian’ (Hall, 1984; Hall and Connell, 1991; Appendix 1) and luminescence dates from Leys Quarry nearby apparently confirm an OIS 6 age for the unit. Several other tills have been correlated with the Rottenhill Till ((Table 7)), including the Red Burn Till at Teindland (Hall et al., 1995a), the Camp Fauld Till (Corse of Balloch Till) on the ‘Buchan Ridge’ (Whittington et al., 1993), the supposedly weathered ‘Kings Cross’ till at Aberdeen (Synge, 1963), the Craig of Boyne Till at Boyne Bay (Connell and Hall, 1984b; Peacock and Merritt, 2000), and the Bellscamphie Till near Ellon (Hall and Jarvis, 1995). Apart from the last two named tills, the others were laid down by ice that flowed out of the Moray Firth and south-eastwards across Buchan towards Aberdeen (see ice sheet models below). A ‘Wolstonian’ age is also likely for the Benholm Clay Formation occurring to the south of Inverbervie (Auton et al., 2000; Appendix 1 Site 26 Burn of Benholm).

Late Quaternary (130 ka to the present day)

Ipswichian Interglacial (OIS 5e) (=Eemian)

Evidence from England and continental north-west Europe indicates that at least the early part of the last interglacial was a little warmer than the present day (Aalbersberg and Litt, 1998). Teindland and Kirkhill represent two of the four sites in Scotland at which organic deposits or palaeosols have been assigned to the Ipswichian (Sutherland and Gordon, 1993). Lowe (1984) disagreed with Fitz-Patrick (1965) and Edwards et al. (1976) that the Teindland Palaeosol Bed (Teindland Buried Soil) could not be attributed to the interglacial with any confidence, but more recent work there, and at Red Burn nearby, largely confirms the earlier conclusions (Hall et al., 1995a; Appendix 1 Site 1 Teindland). For example, the soil contains pollen indicative of deposition towards the end of an interglacial with the progressive replacement of hazel-alder woodland, first by pine woodland, then by heathland and, ultimately, sparse grassland. Luminescence dating suggests an Ipswichian age.

The truncated Backfolds Palaeosol Bed (Kirkhill Upper Buried Soil) is now more confidently assigned to the Ipswichian, following the acquisition of new luminescence dates on sediments underlying it at Leys (Duller et al., 1995). Other supposed interglacial deposits at Tipperty and Balmedie (Bremner, 1938, 1943), north of Aberdeen, and at King’s Cross (Synge, 1963) in Aberdeen, have not been confirmed (Peacock, 1980; Edwards and Connell, 1981). Another site at Errollston, Cruden Bay (Appendix 1 Site 17 Errollston Clay Pit), includes pre-Quaternary palynomorphs and clearly contains much reworked material (Peacock, 1984b; Connell et al., 1985). Reworked shell fragments of Ipswichian age are possibly present at Kippet Hills (Appendix 1 Site 16).

Early Devensian (OIS 5a-d) (=Early Weichselian)

The north-west European pollen and beetle record indicates that the last interglacial was followed by two short, cold stadials (the Herning 5d and Rederstall 5b) separated by cool and wet interstadials (the Brörup 5c and Odderade 5a) (Aalbersberg and Litt, 1998; (Figure 6), (Figure 7). Organic deposits of one or more of these interstadial periods appear to be more common in Scotland than those of the Ipswichian, possibly because the conditions were more conducive to the accumulation of peat. The most complete record in the region of an early Devensian interstadial, probably 5c, comes from the Allt Odhar site, near Inverness (Walker et al., 1992). Here, the pollen and beetle evidence indicates deteriorating climatic conditions towards the end of a cool interstadial as reflected in the replacement of birch woodland and willow scrub with grassland and heath, and then by open communities of grass and sedges. The record of the latter stages of this interstadial, followed by a colder phase of tundra environment, may be represented at both the Camp Fauld and Crossbrae sites in Buchan (Whittington et al., 1998; (Table 7); Appendix 1 Sites 5 and 14). The Burn of Benholm Peat Bed at the Benholm site, south of Inverbervie (Appendix 1 Site 26), is also now correlated tentatively on pollen evidence with OIS 5a or 5c (Auton et al., 2000).

Evidence of major glaciation was formerly reported during OIS 5d and 5b in the north-west Fennoscandian record (Baumann et al., 1995; Mangerud et al., 1996), but it now seems that only limited glaciation occurred (Sejrup et al., 2000; (Figure 43)).

Early Devensian glaciation (OIS 4)

It has been suggested for some time (Sutherland, 1981) that much of Scotland was glaciated during this very cold stage, but no unequivocal evidence has been presented (Bowen et al., 1986; Worsley, 1991). Nevertheless, judging from the record of events in the North Sea basin and north-west Fennoscandia (Figure 43), a regional glaciation occurred during OIS 4 (Sejrup et al., 2000), and much of Scotland is likely to have been glaciated. Several sites in the district probably include a record of this period in the form of periglacial phenomena, for example Corsend Gelifluctate Bed (Gelifluctate IV) at Kirkhill. The Woodside Diamicton (Teindland Till) at Teindland was probably laid down during an Early Devensian glaciation (Hall et al., 1995a), but similar claims for the Hythie Till (Kirkhill Upper Till) (Hall and Jarvis, 1993a), and the Pitlurg Till near Ellon (Hall and Jarvis, 1995) are less secure; the units are assigned here tentatively to OIS 2 (Table 7).

Sutherland (1984c) concluded that the clayey deposits containing beds of cold-water, interstadial-type marine molluscs beneath till at King Edward were in situ (Appendix 1 Site 4). The mid-Devensian amino-acid and radiocarbon ages obtained from the shells (Miller et al., 1987) therefore indicated mid-Devensian marine inundation. For this to have occurred during a period of low global sea level (Pirazzoli, 1993), there must have been substantial glacio-isostatic depression following an Early Devensian glaciation. However, the shelly deposits at King Edward almost certainly have been glacially reworked and form part of the Whitehills Glacigenic Formation (Peacock and Merritt, 1997). Sutherland (1981) used a similar argument in connection with the Clava Shelly Formation at Inverness (Fletcher et al., 1996). However, despite these shelly deposits also being glacial rafts, their derivation from Loch Ness, then a fjord, does seem to imply Early Devensian glaciation of at least the western Highlands (Merritt, 1992).

Middle Devensian (OIS 3)

Periglacial and glacial environments prevailed across much of continental north-west Europe during OIS 3 and glaciers probably existed in the western Highlands for most of the time. There were relatively warm periods between 50 to 41 ka and 37 to 36 ka on the Continent (Huijzer and Vandenberghe, 1998), the former being correlated with the Upton Warren Interglacial of the British chronostratigraphy (Figure 6); (Figure 7). There is also evidence of two interstadials at roughly equivalent times in the Scandinavian record (Figure 43). The younger of the two, the ‘Sourlie Interstadial’ is apparently represented by organic deposits beneath till in the lowlands around Glasgow, where reindeer, woolly rhinoceros and mammoth roamed in a tundra-like environment (Jardine et al., 1988; Sutherland and Gordon, 1993). Reindeer bones found in caves near Inchnadamph, Ross-shire, also date from the Middle Devensian (Lawson, 1984). However, there are no known representative organic deposits in north-east Scotland, although the district was probably free of ice. The Crossbrae Peat (Appendix 1 Site 5) was originally thought to date from between 22 and 26.5 ka BP, but it is now correlated with OIS 5a or 5c (Whittington et al., 1998).

Sand within the glaciofluvial Byth Gravel at the Howe of Bythe Quarry (Appendix 1 Site 6) has yielded luminescence ages of about 45 and about 37 ka, implying the presence of glacier ice in the vicinity (Hall et al., 1995b). This Middle Devensian glaciation would correlate with the Skjonghelleren glaciation of Norway (Figure 43), when ice probably crossed the North Sea basin (Carr, 1998; Sejrup et al., 2000).

Amino-acid ratios and radiocarbon dates on shells within rafts and tills in the Whitehills Glacigenic Formation (Chapter 8) suggest that the deposits are derived from cold-water marine muds of Middle Devensian age. More specifically, the ratios correlate with the Bö Interstadial of Norway (Figure 43), for which ages from 40 to 80 ka have been proposed by Miller et al. (1983) and the higher estimates are favoured by Sejrup et al. (2000). The rafts of the Clava Shelly Clay near Inverness are also thought to have been originally deposited during that interstadial (Merritt, 1992b). Oddly, there appear to be nocorrelatives of the younger Ålesund Interstadial of the Norwegian sequence in north-east Scotland (Peacock and Merritt, 1997).

Dimlington Stadial of the Late Devensian (28 ka to 13 ka BP) (OIS 2)

Although most of the drift deposits in the district were laid down during this period, the sequence of events that occurred is not fully understood. Several conflicting models have been published (Figure 3), and the controversy is discussed more fully below, where a more detailed history is given. It is the view of the authors that the district was overwhelmed entirely by ice in the Late Devensian, approximately coeval with the mainland ice sheet reaching the continental shelf break to the north-west of Scotland (Figure 41). There is growing evidence from Scandinavia, the North Sea and elsewhere that the ice reached its maximum extent early in the Late Devensian, between about 28 ka and 22 ka BP, not at about 18 ka BP, as was widely believed until recently (e.g. Bowen et al., 1986). The Scottish and Scandinavian ice sheets coalesced during this early phase (Carr, 1999; Sejrup et al., 2000). It was probably also at this time that ice of the Moray Firth ice stream was forced to flow south-eastwards across Buchan to lay down at least part of the Whitehills Glacigenic Formation and its correlatives (Table 7). There probably followed a period of glacial retreat that lasted from about 21 to 18 ka BP, roughly equivalent to the Hamnsund Interstadial of Norway (Valen et al., 1995; (Figure 43)). A major re-advance then probably occurred between 18 and 15 ka BP equivalent to the Tampen Re-advance of Norway (Sejrup et al., 1994; (Figure 43)). This would have mostly involved the coastal ice streams. The Moray Firth ice stream had retreated from the north-east coast of Buchan by about 15 ka BP, and the whole district would have been ice-free by the beginning of the Windermere Interstadial at 13 ka BP.

Throughout this memoir, the Scottish ice sheet that formed between 28 and 13 ka is referred to as the ‘Main Late Devensian’ in order to to distinquish it from that of the Loch Lomond Stadial, also part of the Late Devensian. The term ‘Dimlington Stadial ice sheet’ is not used because there are now strong arguments for redefining the time span of the Dimlington Stadial to 18–13 ka BP (Sejrup et al., 1994).

Windermere (‘Lateglacial’) Interstadial (13 ka to 11 ka BP) (OIS 2)

A sudden amelioration of the climate occurred at about 13 ka BP when the oceanic polar front in the North Atlantic migrated northwards (Ruddiman and McIntyre, 1973, 1981; Ruddiman et al., 1976) and temperate waters returned off the western coasts of the British Isles (Peacock and Harkness, 1990). Atmospheric temperatures rose within decades to close to present values (Bishop and Coope, 1977; Atkinson et al., 1987). It is almost certain that no ice survived the interstadial in north-east Scotland, although glaciers might not have disappeared completely in the central and western Highlands (Sissons, 1974a; Sutherland, 1984a; Sutherland and Gordon, 1993; Clapperton, 1997). Masses of ice buried within sediments from all phases of the Main Late Devensian glaciation (if not before) commonly melted out relatively slowly resulting in kettleholes (Peacock in Harkness and Wilson, 1979).

Organic sequences preserved in some kettleholes in the region provide a record of environmental change during the Late-glacial period (Donner, 1957; Vasari and Vasari, 1968; Vasari, 1977; Gordon, 1993; Appendix 1 Mill of Dyce, Rothens, Loch of Park). Elsewhere, beds of peat preserved beneath solifluction deposits and landslips provide important evidence (Godwin and Willis, 1959; Clapperton and Sugden, 1977; Hall, 1984; Connell and Hall, 1987; Aitken, 1991; Appendix 1 Site 25 Glenbervie). Radiocarbon dates are given in (Table 8). The freshly deglaciated ground was first colonised by a pioneer vegetation of open habitat species followed by the immigration of crowberry heath, juniper and dwarf varieties of birch and willow. Eventually open birch woodland developed with juniper and isolated stands of Scots pine locally. A stepwise climatic deterioration occurred throughout the Windermere Interstadial (Lowe et al., 1999; Mayle et al., 1999) and it is likely that glaciers had already started to build up in the Western Highlands before more sustained cooling began at about 11.2 ka BP.

Relative sea level in north-east Scotland probably had fallen to below its present level by the beginning of the interstadial and it continued to fall, especially in the west of the district, where glacio-isostatic rebound was greater.

Loch Lomond Stadial (11 ka to 10 ka BP) (OIS 2)

The North Atlantic Polar Front migrated rapidly southwards to the latitude of northern Portugal at about 11 ka BP (Bard et al., 1987), and the climate of the British Isles reverted temporarily to arctic conditions.

Although no glaciers are thought to have developed in the district they did so in the corries of the Cairngorms and the Mounth, and substantial ice caps formed over Rannoch Moor and in the North-west Highlands (Sutherland and Gordon, 1993).

A tundra environment existed in north-east Scotland during the stadial. Soils that had developed during the Windermere Interstadial were largely destroyed by periglacial processes such as solifluction and frost churning (geliturbation) (FitzPatrick, 1956, 1958, 1969, 1972, 1976, 1987; Galloway, 1961a–c, Connell and Hall, 1987). Vegetation was quickly replaced by open-habitat plant communities tolerant of the Arctic conditions and unstable soil (Gunson, 1975). Some fossil frost polygon networks observed in the district (Chapter 7) might have been formed during the stadial (Clapperton and Sugden, 1977), but most probably formed earlier during the retreat of the Main Late Devensian ice sheet (Gemmell and Ralston, 1984, 1985; Armstrong and Paterson, 1985; Connell and Hall, 1987; Appendix 1 Ugie valley). Fluvial and debris-flow activity would have been enhanced during the stadial, especially during springtime snow-melts (Ballantyne and Harris, 1994; Maizels and Aitken, 1991). Slopes would have been particularly prone to land-slipping (Appendix 1 Glenbervie) and head deposits formed widely across the district. Much of the coarse-grained alluvium in minor valleys across the district would also have been laid down then.

Holocene (10 ka BP to the present day) (OIS 1)

At about 10 ka BP the global climate changed rapidly, possibly within an average human lifetime, to its present interglacial mode, marking the beginning of the Holocene epoch. The oceanic polar front in the North Atlantic rapidly shifted northwards and the warm Gulf Stream current became established, providing an ameliorating influence on the climate of Scotland. At first the glaciers in the western Highlands might have re-advanced locally as a result of increased snowfall (Benn et al., 1992), but rapid disintegration soon followed. The widespread occurrence of bare, unstable soils at the beginning of the Holocene apparently led to a period of intense fluvial erosion and deposition with enhanced debris flow activity on mountain sides and extensive formation of landslips as the ground thawed (McEwen, 1997). Soils gradually became more stable following the establishment of vegetation: firstly shrubs and scrub communities were established during the early part of the Windermere Interstadial, and was replaced later by woodland (Durno, 1956, 1957; Vasari and Vasari, 1968; Edwards, 1978). A distinct phase of juniper dominance was replaced by birch woodland, followed by the arrival of hazel, elm and shortly after, by oak (Appendix 1 Loch of Park, Nether Daugh). Pine forest became established in the East Grampians by the mid-Holocene whereas birch-hazel forest was predominant in the east and across the coastal lowlands (Vasari and Vasari, 1968; Gunson, 1975; Birks, 1977).

Climatic deterioration may have begun soon after the beginning of the Holocene and it has continued in a stepwise fashion. Distinct layers of pine stumps preserved quite widely in upland blanket peat deposits in the East Grampians indicate that pine forest locally succumbed to the spread of Sphagnum moss between 6.7 ka and 5 ka BP (Pears, 1968, 1970). Colder and wetter climatic conditions caused this change, rather than the influence of people, but human impact is apparent in the pollen and sediment records before 5 ka BP (Edwards, 1978, 1979b; Edwards and Rowntree, 1980). It can also be seen in the late Holocene pollen and Coleopteran record obtained from a silted-up ox-bow lake on the floodplain of the River Don at Nether Daugh, near Kintore (Appendix 1).

Although the imprint of glaciation remains dominant, postglacial processes have superimposed subtle, but distinctive, modifications on the landscape. Steep hillsides have been modified by landslips, soil-creep and debris flows, whereas valley floors have been sculptured by rivers forming spreads of alluvium and river-terrace deposits. In general, braided rivers with gravelly beds would have been replaced by the single-thread, and locally meandering, rivers of the present day by the mid-Holocene (McEwen, 1997). The Spey is an exception in that it still maintains a braided, gravelly floodplain in its lower reaches (Lewin and Weir, 1977).

Sea level was below that of the present day during the early Holocene, but the Main Postglacial Transgression resulted in the formation of raised beaches and marine inundation of the lower reaches of valleys. The latter led to the deposition of estuarine silts, fine-grained sands and clays, locally on terrestrial peats, tree stumps and fluvial muds (Chapter 7). The transgression reached its maximum after 6.1 ka BP in the Ythan valley (Smith et al., 1999; (Figure 48)) and between 6.3 ka and 5.7 ka BP in the Philorth valley, between Fraserburgh and Peterhead (Smith et al., 1982; Appendix 1 Site 9). The subsequent regression, which has been attributed to a general lowering of global sea level following expansion of the Antarctic ice cap (Goodwin, 1998), was accompanied by renewed terrestrial sedimentation in the valley mouths. Horizons of fine-grained sand identified in both estuarine and near-shore terrestrial deposits of the lower Ythan and Philorth valleys were laid down at about 7200 BP by a tsunami caused by a major submarine slide on the Norwegian continental slope (Dawson et al., 1988; Long et al., 1989; Appendix 1 Philorth).

Previous models of glaciation

The systematic study of glacial deposits and landforms in north-east Scotland began in the mid-19th century with the pioneering work of Thomas Jamieson. During the following 150 years of painstaking research there have been numerous attempts to unravel the complicated record of glacial events. Although the glacial reconstructions that have been published are generally based on sound evidence, they were inevitably influenced by the contemporary knowledge of glacial processes and events. The conflicts that have arisen result largely from the limited stratigraphical record of events in the region and the general absence of materials that can be reliably dated, a situation that still exists today.

The work of Jamieson

In a series of papers stretching over a period of fifty years, Jamieson developed modern concepts of Quaternary glaciation from previous theories of the great biblical submergence and erosion by floating ice (Jamieson, 1858, 1859, 1860a, b, 1862, 1865, 1866, 1874, 1882a, b, 1906, 1910). He published a wealth of stratigraphical detail establishing sequences of glacial events and reconstructed former flow lines from the distribution of erratics, striae, and orientations of roches moutonnées. He was one of the first to realise that the glaciation of Scotland had involved the formation of a major ice cap like that of Greenland rather than Alpine-style valley glaciers.

In Jamieson’s final model (Figure 36) the oldest glacial event involved ice from the Moray Firth laying down ‘shelly indigo boulder clay’ in the Ellon district. The deflection was interpreted to have been caused by Scandinavian ice impinging on the eastern coast of Buchan. The indigo diamicton was largely stripped away by ice flowing from inland during a subsequent event, which laid down the locally derived, non-shelly ‘Lower Grey Boulder Clay’ of Aberdeenshire. That till was subsequently overlain by the ‘Red Clay Series’ along the coast to the north of Aberdeen. These vivid reddish brown deposits with Old Red Sandstone erratics had been deposited by ice flowing northwards from Strathmore, but deflected onshore by Scandinavian ice in the North Sea. Jamieson noted that the Red Clay Series intermingled with, and thus was contemporaneous with, the ‘Dark Blue Boulder Clay’ in the Peterhead area. The latter till contained many large erratics derived from Mesozoic rocks in the Moray Firth. Local ice had retreated far to the west during this incursion, although the Lower Grey Boulder Clay was regarded as broadly similar in age to the red and dark blue coastal tills. There was local marine incursion following the decay of the coastal ice streams leading to the deposition of clays. The final event involved an expansion of ice from inland. During this ‘Aberdeen Re-advance’, a glacier descended the valley of the River Dee to reach the coast, where it scoured away the Red Clay Series deposits locally.

The work of Bremner and Read

The mantle of Jamieson was taken over by Alexander Bremner in the early 20th century. He also published a wealth of detail on the Quaternary of north-east Scotland, paying particular attention to the pattern of supposed ice-marginal glacial meltwater channels formed during glacial retreat (Bremner, 1912, 1915, 1916, 1917, 1919, 1920a, b, 1921, 1922, 1928, 1931, 1934a, b, 1936, 1938, 1939, 1942, 1943). Bremner developed the triple glaciation model, although he disagreed with Jamieson about the sequence of events and many details. Unlike Jamieson, Bremner claimed to have found evidence confirming that the main glacial episodes were separated by warm, nonglacial periods.

In Bremner’s final model (Figure 37), the first glaciation involved ice sourced in Sutherland, Ross and Cromarty and the Great Glen flowing into the region from the west and north-west. It picked up Mesozoic rocks and shells from the Moray Firth and laid down dark blue shelly till along the northern coast. The Moray Firth ice stream coalesced with Scandinavian ice off Buchan and the combined ice masses flowed southwards leaving a train of boulders of Peterhead Granite and sparse erratics of rhomb-porphyry and larvikite. Peat preserved in the vicinity of the Burn of Benholm site (Appendix 1) supposedly formed in the subsequent interglacial period.

Ice traversed Buchan from the south-west and south in Bremner’s second glaciation. He used evidence cited by Read (1923) to show that it flowed northwards into the Moray Firth along the northern coast. Ice from Strathmore deposited red tills along the eastern coast and laminated clays were deposited in freshwater lakes ponded against the landward margin of that ice stream during its retreat.

Tills resulting from Bremner’s third glaciation were supposedly difficult to identify owing to the glacial reworking of older deposits. Patterns of ice flow were determined mainly from the orientation of supposed ice-marginal drainage channels and other morphological evidence. During this glaciation, the ‘Aberdeen Re-advance’ was supposedly contemporaneous with another advance of Moray Firth ice that affected the northern coast as far east as Fraserburgh. Widespread glaciofluvial deposits and morainic deposits in the lower Dee valley were attributed to the Aberdeen Re-advance and another spread at Dinnet, in the Tarland Basin, was attributed to a subsequent re-advance.

Bremner disagreed with H H Read regarding the evidence for the number of glacial episodes that had affected the northern coast. However, he acknowledged the careful observations that Read (1923) had made during the detailed geological survey of Banffshire, in particular Read’s wide knowledge of the source rocks from which the glacial erratics were derived. Read deduced that the earliest ice-movement to affect the Banffshire coast came from the west and north out of the Moray Firth, conditioned by the presence of Scandinavian ice at the mouth of the firth. Glaciolacustrine deposits of his ‘Coastal Series’ formed in lakes following the partial withdrawal of the Moray Firth ice stream. Inland ice then ‘crept’ northwards towards the coast, redistributing erratics, but depositing little recognisable till.

Work in the post-war period

Detailed, accurate stratigraphical work in the Aberdeen area was continued by Simpson (1948, 1955). He demonstrated that the moraines and meltwater channels attributed by Bremner (1938, 1943) to the Aberdeen and Dinnet re-advances of his third glaciation were simply produced during periods of stagnation (still-stands) that interrupted the decay of contemporaneous coastal and inland ice masses. Simpson (1955) supported Jamieson (1865, 1906) in that he believed the dark shelly tills of the northern coast were contemporaneous with the red tills of eastern Aberdeenshire, an interpretation disputed earlier by Bremner (1943).

Charlesworth (1956) and Synge (1956) both concluded that a substantial portion of the Buchan Plateau had not been glaciated during the last glaciation. This resulted in the concept of ‘Morainless Buchan’. They supported Jamieson (1865, 1906) believing that ice from the Moray Firth and Strathmore had coalesced in the vicinity of Peterhead while there was no ice immediately inland. Charlesworth concluded that many valleys became blocked by ice at the coast causing a series of proglacial lakes to form. These lakes were supposedly interconnected by glacial spillways. The concept of ice-free enclaves during the Devensian and earlier glaciations was supported by the recognition of widespread saprolites, weathered tills and periglacial phenomena (FitzPatrick, 1958, 1972; Galloway, 1961a, b, c).

Synge (1956, 1963) recognised two major glaciations in his model (Figure 38), with two re-advance stages occurring during the decay of the final ice sheet. Ice first moved into Buchan from the east or north-east, leaving Scandinavian erratics along the coast. There followed a ‘Greater Highland Glaciation’ (Wolstonian or Anglian) in which the Scandinavian ice caused Scottish ice to bifurcate over Buchan, thus preserving the Buchan Gravels in an enclave of minimal glacial erosion. Extensive weathering of bedrock supposedly occurred in the region during the succeeding interglacial period. Ice reoccupied the coasts during the second (Devensian) glaciation, but a substantial area of central Buchan purportedly remained ice free. The decay of the Strathmore ice lobe was accompanied by the deposition of estuarine red clays. Two short cold intervals lead to re-advances of Grampian ice at Aberdeen and Dinnet, the latter being correlated incorrectly with the Loch Lomond Re-advance of western Scotland.

Sissons (1965, 1967) placed Synge’s Greater Highland Glaciation in the Devensian and correlated the Aberdeen and Dinnet events with his Aberdeen–Lammermuir and Perth re-advances, respectively (Figure 39). He later abandoned the concept of the first re-advance and agreed that there was no irrefutable evidence for the latter at Perth (Paterson, 1974; Sissons, 1974b, c). Peacock et al. (1968) proposed that another ‘oscillation’ occurred in the Elgin area and that earlier in the deglaciation, ice-dammed lakes developed agaist the Moray Firth ice stream along the Banffshire coast.

Late 1960s to the early 1980s

The evidence for multiple glaciations and active, phased retreat of the last ice sheet became seriously questioned during this period of research. This stemmed from the general recognition that complex depositional sequences could result from rapidly changing processes and environments within single glacial events. Most drift deposits in the region were assigned to the last glaciation (Murdoch, 1975; McLean, 1977; Clapperton and Sugden, 1975, 1977), although doubts remained about its timing and the main directions of ice flow (Sissons, 1981).

The apparent spatial continuity of the pattern of glacial landforms and deposits, particularly of glaciofluvial origin, led Clapperton and Sugden (1975, 1977) to conclude that the entire area of north-east Scotland had been covered by the Late Devensian ice sheet (Figure 40). They concluded that the ice sheet was largely wet-based, and it was composed of three distinct regional ice streams with a triple junction zone located over central Buchan. The ice sheet decayed comparatively quickly leading to widespread stagnation in many lowland sites and no major re-advances were recognised. Central, northern and eastern Buchan became free of ice ‘early’ in the deglaciation, allowing time for proglacial lakes to form against the more persistent coastal ice streams and for previously deposited sediments and saprolites to become periglacially modified.

It was widely held that the whole of Scotland together with the northern and central North Sea basin was glaciated during the Late Devensian (e.g. Sissons, 1976; Boulton et al., 1977; Price, 1983).

Early 1980 to 2002

The discovery that stiff, stony lodgement tills were absent beneath most of the northern North Sea basin had a profound influence on modelling work during the 1980s for it was generally assumed that this indicated that Scandinavian ice had not crossed the North Sea during the Late Devensian glaciation (e.g. Cameron et al., 1987). This led to Sutherland’s (1984a) reconstruction of an independent British ice sheet of restricted size, in which ice terminated at the eastern margin of the Wee Bankie Formation off the eastern Scottish coast ((Figure 3)d). Sutherland (1984a) also cited evidence from the Crossbrae and King Edward sites (Appendix 1) to argue that the northern coast of Banffshire and Buchan to the east of Portsoy had not been glaciated during the Late Devensian. The nonglaciated enclave he proposed differed from that envisaged by Synge (Figure 38) and enclaves of various extent have appeared on subsequent reconstructions of the north-eastern sector of the last British ice sheet (e.g. those shown in (Figure 3) and Bowen et al., 1986; Sejrup et al., 1987; Nesje and Sejrup, 1988; Lambeck, 1993, 1995), some long after Sutherland’s evidence had been seriously questioned (Peacock, 1985; Hall and Bent, 1990; Hall, 1997). Ehlers and Wingfield (1991) also recognised an unglaciated enclave in Buchan, but suggested Scottish ice expanded farther east than the Wee Bankie Moraine, coalescing in places with Scandinavian ice ((Figure 3)f).

Over the past twenty years, field-related research into the glaciation of north-east Scotland has been undertaken predominately by the present authors and contributors. Modified versions of the single stage Clapperton and Sugden model for the last glaciation have been adopted in recent publications and maps of the Geological Survey on north-east Scotland over this period (e.g. Merritt, 1981; BGS, 1992). However, although stratigraphical evidence now appears to confirm that the ‘inland’, Moray Firth and Strathmore ice streams were confluent, both Hall (1983, 1984) and Hall and Connell (1991) reverted to a two-stage model similar to that originally proposed by Jamieson (1906).

This was because there was conflicting geochronometric dating evidence for the age of the ‘inland’ tills in Buchan.

The whole of Scotland was ice-covered during the first phase of Hall’s 1983 model, but ice from Strathmore reached only as far as Stonehaven. Ice flowing down the Dee valley and Insch depression (Wilson and Hinxman, 1890) crossed the coast between Aberdeen and the Ythan, depositing basic igneous erratics at Nigg Bay (Appendix 1). In Buchan, the pebbles of the distinctive Buchan Gravels from Whitestones Hill and of erratics from the quartzbiotite norite of the Maud intrusion indicates flow towards the north-east and east (Wilson, 1886). ‘Inland’ ice may have been confluent with Moray Firth ice along the valley of the North Ugie Water. Ice flowed north-eastwards across Banffshire to join the Moray Firth ice stream at the coast, west of the mouth of the Deveron (Read, 1923).

The East Grampian ice sheet retreated to an unknown position in the west during the second phase. Strathmore ice flowed northwards until it met the more vigorous Moray Firth ice stream, which diverted ‘inland’ ice eastwards into the lower Spey valley (Read, 1923) and pushed across northern parts of Buchan.

New information from offshore allowed the model to be refined (Hall and Bent, 1990; Hall, 1997). The maximum limit of the main Late Devensian ice sheet was placed at the Bosies’ Bank Moraine (Figure 44), where a tide-water margin was identified some 25 km north-east of Buchan. East Grampian ice had already retreated before the maximum offshore extent was attained, allowing incursion of Moray Firth ice across northern Buchan. This timing of maximum expansion was not known, but deglaciation was well advanced by 15.3 ka BP, when ice had retreated from the Buchan coast (Appendix 1 St Fergus). Glaciomarine sedimentation was also occurring to the west of Bosies’ Bank at the retreating ice margin at this time (Bent, 1986).

In the last decade or so it has been demonstrated that low gradient ice sheets crossing wet, deformable beds, especially in marine areas, are unlikely to form thick stony lodgement tills. Instead, fine-grained diamictons (deformation/deforming bed tills) are produced that can be mistaken easily for undeformed glaciomarine deposits. Additionally, some diamictons are too thin to be detected on seismic records. This knowledge, together with new AMS 14C dates on individual molluscs and hand-picked foraminiferids, led Sejrup et al. (1994) to return to the view that the British and Scandinavian ice sheets had coalesced in the northern North Sea basin during the early part of the Late Devensian, between 28 and about 22 ka BP. This agreed with the revised interpretation of the Dimlington type locality in Yorkshire (Eyles et al., 1994). The central part of the northern North Sea was deglaciated by 20 ka BP, followed by a re-advance in the eastern North Sea between 18.5 and 15.1 ka BP. A similar re-advance probably affected north-east Scotland during this period, when ice possibly reached the limits that Sejrup et al. (1987) used in their earlier reconstruction of the main Late Devensian maximum (Figure 41). Peacock (1997) argued that the glacitectonic disturbance of the St Fergus Silts (Appendix 1) could have resulted from a subsequent re-advance of the Moray Firth ice stream between 15 and 14 ka BP.

Recent micromorphological examination of sediment thin sections from boreholes has enabled Carr (1998) to demonstrate that ice crossed the North Sea basin on three occasions during the Devensian–Weichselian. Sejrup et al. (2000) conclude that it did so during the Karmoy, Skjonghelleren and main Late Weichselian glaciations of Fennoscandia (Figure 43).

The glacial record in Banffshire has been re-appraised by Peacock and Merritt (2000). Ice first flowed south-eastwards from the Moray Firth during the first phase of the Late Devensian glaciation, as originally deduced by Read (1923). The Moray Firth ice stream then withdrew from the coast, allowing ice from the East Grampians to creep northwards until being steered eastwards by the more powerful coastal ice-stream. The inland ice subsequently withdrew, allowing glacial lakes to form against the Moray Firth ice stream, which continued to remain active and made forays southwards from time to time (Figure 42).

Present model of glaciation

The glacial reconstructions described above are clearly conflicting, and some of their aspects are certainly flawed, but they all contain elements that remain valid. It is salutary that the stratigraphical approach of Jamieson has perhaps best survived the test of time. The sequences he described have generally been confirmed in recent site excavations and most of his deductions about the origins of deposits presented in his final paper are sound. Although there is still doubt about the time span involved, the basic tripartite sequence he established in the Ellon area still holds, and his evidence supporting the Aberdeen Re-advance has not been rejected beyond doubt.

Bremner’s final model was not so firmly based on stratigraphy, but his evidence for incursion of Scandinavian ice is compelling. Although some of the meltwater channels that he thought were formed at the retreating margin of the last ice sheet contain pre-Late Devensian deposits, and hence are older features, most indicate the pattern of ice retreat that he deduced. Conversely, the evidence does not support Clapperton and Sugden’s (1977) model of a regionally integrated, contemporaneous network of channels forming beneath a warm-based ice sheet. Nevertheless, the continuity of channel development is strong evidence in support of full ice cover during the last glaciation of the region. The pattern of ice movement in Clapperton and Sugden’s model is in broad agreement with available stratigraphical evidence, although it has to be modified in eastern Buchan to take into account the evidence of west to east glacial streamlining (Map 6), rather than north-west to south-east. Furthermore, their flow lines for the Strathmore ice stream do not account for the significant onshore carriage of Permo–Triassic lithologies into eastern Buchan.

Extent of glaciation

Until recently, the tills occurring in central Buchan could only be loosely dated as Devensian. This allowed repeated claims, discussed above, that parts of the area remained ice-free during the Main Late Devensian glaciation ((Figure 3)c, d, e). The evidence for so-called ‘Moraineless Buchan’ was equivocal, but the discovery in 1980 of the Crossbrae Peat with its interstadial pollen record soon become central to the debate (Appendix 1 Site 5). The peat appeared to rest on till deposited by ice flowing from the west and it was overlain by gelifluction deposits, but not by till. Furthermore, radiocarbon dates implied that the peat formed early in the Late Devensian. The site seemed to provide crucial evidence that ‘inland series’ tills in the area predated the Late Devensian and that central Buchan had indeed not been glaciated during that period (Sutherland, 1984a; Connell and Hall, 1987). However, on re-examination in 1992, the peat was found to rest on bedrock and to be overlain by coarse gravels of possible glaciofluvial origin (Whittington et al., 1998). Furthermore, new dating and palaeoenvironmental evidence from the peat indicate that it was formed most probably during the Early Devensian in OIS 5c or 5a (Table 7). The coarse gravel unit is probably, but not unequivocally Late Devensian in age.

Evidence from Crossbrae can no longer be used confidently to support the argument that parts, or all, of Buchan escaped glaciation in the Late Devensian. Recent mapping confirms that the entire area is crossed by a system of glacial meltwater channels that formed diachronously during the general westward retreat of ice (Chapter 7). Several lines of reasoning suggest that this was the Main Late Devensian ice sheet.

1. Time-dependent amino-acid epimerisation data on reworked shells in the Whitehills Glacigenic Formation at Boyne Limestone Quarry, Gardenstown and King Edward Quarry (Appendix 1 Sites 2, 3, 4) suggest that this unit and overlying tills can be attributed either to a Middle Devensian (OIS 3) glacial event equivalent to the Skjonghelleren glaciation of Norway (Figure 43), or to the Late Devensian (Peacock and Merritt, 1997).
2. Full ice cover in Buchan during the Late Devensian is implied by the projection of ice surface profiles from terminal moraines of probably Late Devensian age in the Outer Moray Firth (Hall and Bent, 1990) and by evidence that the Scottish and Scandinavian ice sheets coalesced in the northern North Sea between 28 and 22 ka BP (Sejrup et al., 1994, 2000).
3. The existence of raised glaciomarine silts and a raised beach at St Fergus (Appendix 1 Site 11), radiocarbon dated to 14 to 15 ka BP (Hall and Jarvis, 1989; Peacock, 1997), implies the previous removal of a thick ice cover, because considerable isostatic depression would be required for such sediments to be laid down during a period of low global sea level.
4. Glaciofluvial gravels beneath till of westerly origin at the Howe of Byth (Appendix 1 Site 6) have yielded luminescence dates of 45 ± 4 ka and 37 ± 4 ka BP (Hall et al., 1995b), although doubts have been expressed about the validity of these dates (Peacock and Merritt, 1997).

It is concluded that all of Buchan was glaciated during the Late Devensian. Furthermore, apart from nunataks in the north-west Highlands and Inner Hebrides (Ballantyne et al., 1998), it is unlikely that ice-free areas existed anywhere in Scotland during the Main Late Devensian glaciation, including Caithness and the islands of Orkney and Lewis (Hall and Whittington, 1989; Hall, 1995). A return to earlier models (e.g. Sissons, 1976; Boulton et al., 1977, 1985) of extensive ice sheet glaciation in the Late Devensian is suggested, with Scottish ice terminating at, or close to, the continental shelf to the north and west of Scotland and covering most of the northern North Sea basin (Hall and Bent, 1990; Sejrup et al., 1998, 2000). This view is supported by the recent reconstruction of the last ice sheet in north-west Scotland (Ballantyne et al., 1998), based on the detailed mapping of glacial trimlines, and assuming low gradient profiles for ice streams crossing deformable sediments.

Multistage models of the Main Late Devensian glaciation

The detailed record of climate change derived from the Greenland ice sheet (Figure 31) together with the discovery of various horizons of ice-rafted debris (Heinrich events) in North Atlantic cores, has led to increased scrutiny of the evidence for Late Devensian glaciation in Britain. It had been widely accepted that the last major glaciation of the Northern Hemisphere peaked at about 18 ka BP (Climap, 1976), and evidence from the type locality of the Dimlington Stadial in Eastern Yorkshire appeared to support a glacial maximum between 18 and 16 ka BP (Rose, 1985). However, subsequent Late Devensian amino-acid ratios on shells from the supposed pre-Ipswichian Basement Till at Dimlington indicate that the maximum extent of the main Late Devensian glaciation in eastern England occurred before 18.4 ka BP (Eyles et al., 1994). Several surges of coastal ice followed. As explained above, Sejrup et al. (1994, 2000) have presented evidence that the maximum Late Devensian glaciation also occurred before 18 ka BP in the northern North Sea. Further support for an ‘older’ Late Devensian maximum is provided by a 14C date of 22.5 ka BP for glaciomarine deposits associated with the maximum position of the independent Outer Hebrides ice cap near the edge of the continental shelf (Selby, 1989). Similarly, Peacock (1997) argued for a re-advance of Scottish ice into the western North Sea between 15 and 14 ka BP.

Supporting evidence from Ireland and the northern Irish Sea basin also suggests that maximum glaciation occurred early in the Late Devensian (by about 22 ka BP), followed by stillstands and at least one re-advance that has been age-constrained by AMS radiocarbon dating on foraminiferids (McCabe, 1996; McCabe and Clark, 1998). Following a period of ice retreat between 16.7 and 14.7 ka BP (Cooley Point Interstadial), ice re-advanced, reaching its maximum position at about 14 ka BP (Killard Point Stadial), which is roughly coincident with Heinrich Event 1 (McCabe et al., 1998).

Arguably, the north-west Fennoscandian sequence of Sejrup et al. (2000; (Figure 43) provides the best yardstick against which events in north-east Scotland can be judged. However, there are outstanding difficulties requiring further work and clarification. For example, Olsen (1997) concludes that Scandinavian ice advanced on to the continental shelf four times between 40 and 14.5 ka BP, with ice advances peaking at about 40, 30–29, 24–21 and 17–14.5 ka BP. Furthermore, radiocarbon ages on bone fragments at the type locality of the Hamnsund Interstadial suggest that it occurred at about 24.5 ka BP and not at about 20 ka (Mangerud et al., 1996). There are also difficulties with the radiocarbon dates for the Swatchway and Witch Ground formations in the Fladen area and the correlation of these units with the Wee Bankie and Tampen formations (Carr, 1998). In essence, ice might have been present in the Witch Ground area until almost 13 ka BP. It is speculated here that this ice could be the remains of a previously more extensive ice dome (Figure 41). The dome had possibly expanded in size during the ‘Dimlington Advance’ event (18 to 13 ka BP) causing the deflection of ice from Strathmore into the lower Ythan valley to lay down the deposits of the Logie-Buchan Drift Group. The speculative ice dome shown on (Figure 41) is based on evidence of a more or less radial pattern of channels (incisions) of Late Weichselian age in the Witch Ground area (Ehlers and Wingfield, 1991).

Another consequence of accepting a more widespread glaciation between 28 and 22 ka BP is that there must have been considerably more glacio-isostatic depression of the land in north-east Scotland, and hence higher relative sea levels, than determined by Lambeck (1993b) from numerical, geophysically based modelling.

Likely sequence of events during the Late Devensian in north-east Scotland

Early stage of the Main Late Devensian glaciation (28 to 22 ka BP)

The district was probably overwhelmed entirely by ice in the early part of this period (Hall and Bent, 1990; Peacock and Merritt, 1997; Whittington et al., 1998; (Figure 41)). Initially, ice would have flowed from centres in the western and north-western Highlands, followed by the establishment of an independent ice cap over the Cairngorms and eastern Grampian Highlands. This probably began prior to the Late Devensian. A major ice stream occupied the Moray Firth where it would have flowed relatively swiftly across the underlying soft, deformable, muddy sea bed and Mesozoic sedimentary rocks. Another Highland-sourced terrestrial ice stream flowed north-eastwards along Strathmore towards the North Sea and a further one flowed down the Spey Valley towards the Moray Firth (Sutherland and Gordon, 1993). The interaction between the relatively fast-moving and probably warm-based ice streams and the relatively sluggish, and probably largely cold-based, East Grampians ice sheet, has profoundly influenced the distribution of drift deposits in the district, leading locally to complicated sequences (Peacock and Merritt, 2000; (Figure 4)). Each ice stream produced distinctive suites of deposits, forming the basis of the five lithostratigraphical groups described in Chapter 8.

Many of the terrestrial deposits of this early phase of glaciation would have been eroded or reworked during subsequent phases. Evidence is preserved that indicates that the Moray Firth ice stream invaded the coastal lowlands of Moray, Banffshire and Buchan to lay down the dark grey shelly tills and rafts of Mesozoic strata, the main components of the Whitehills Glacigenic Formation. The ice possibly crossed the Buchan Plateau towards Aberdeen during this event (Hall and Jarvis, 1995), but if not, it did so in a pre-Late Devensian glaciation, laying down ‘indigo’ tills around Ellon (Table 7). On all occasions, the deflection of Moray Firth ice across Buchan was probably caused by the presence of Scandinavian ice in the central North Sea.

It is clear that following the deposition of the White-hills Glacigenic Formation at Boyne Bay, Gardenstown and up the Deveron valley, the power of the Moray Firth ice stream waned sufficiently to allow ice sourced inland to flow eastwards towards the coast and to lay down the Old Hythe, Crovie, Hythie, Byth, Sandford Bay and laterally equivalent tills (Table 7). Although there is no apparent unconformity in the sequence, there was possibly a significant hiatus accompanying the change in direction of flow, perhaps an interstadial event coincident with this glacial reorganisation.

A mid-Late Devensian interstadial? (22 to 18 ka BP)

Judging from the north-west Fennoscandian record (Sejrup et al., 2000; (Figure 43)), it is likely that there was substantial ice-sheet recession in Scotland at about 20 ka BP during a cold, dry interstadial period. The initial retreat of the East Grampian ice sheet from central Buchan had probably occurred by then, with the development of ice-marginal lakes between it and the coastal ice streams (Figure 42). Deposits of the Ugie Clay Formation were laid down in these lakes. At about this time, the lower reaches of the valley of the River Deveron north of Turriff were blocked by the Moray Firth ice stream causing meltwaters to flow south and east, first via the Towie spillway, near Fyvie, then into the valley of the River Ythan (Figure 42).

Later stages of the Main Late Devensian glaciation (18 ka to 13 ka BP)

There is growing evidence elsewhere that a significant build up of ice occurred during this period, but there is no unambiguous record of any substantial re-advance of the East Grampian ice sheet. There is evidence that the Moray Firth ice stream expanded across Lake Ugie, because a thick sequence of laminated clay (Ugie Silt Formation) is capped by blue-grey deformation tills (Essie Till Formation) inland of Peterhead (McMillan and Aitken, 1981; Chapter 8). The Logie-Buchan ice also advanced over glaciolacustrine deposits (Appendix 1 Site 12 Sandford Bay), and the moraines at Cross Stone, near Ellon, indicate a late re-advance onshore from the North Sea (Map 6). The position of the Cross Stone moraine and the composition of the Logie-Buchan Drift Group deposits associated with it, suggest that Scandinavian ice was present in the North Sea basin, causing ice derived from Strathmore to flow northwards, then westwards and northwestwards penetrating onshore. However, the central North Sea, at least, is thought to have been ice free at this time (Sejrup et al., 1994). If correct, the Moray Firth and Strathmore ice streams possibly advanced to the position of the Bosie’s Bank and Wee Bankie moraines (Figure 44), equivalent to the Tampen Re-advance of north-west Fennoscandia (Sejrup et al., 1994).

The Moray Firth ice stream possibly stood at the entrance of the Moray Firth at about 15 ka BP, where its activity disturbed the deposits of the St Fergus Silt Formation (Appendix 1 Site 11). There was substantial ponding of meltwaters between the slowly retreating East Grampians ice sheet and the Moray Firth ice stream. The glaciolacustrine Kirk Burn Silt Formation was deposited in these lakes along the Banffshire coast and its hinterland (Figure 42). Local glacial readjustments led to a minor re-advance of the East Grampians ice sheet towards the Banffshire coast following the retreat of the Moray Firth ice stream (Peacock and Merritt, 2000). A subsequent oscillation of the Moray Firth ice stream affected the area around Elgin, possibly at about 14 ka BP, and a fourth one formed a prominent push moraine at Ardersier, near Inverness, possibly at about 13 ka BP (Merritt et al., 1995). This event has since been correlated tentatively with the Killard Point Stadial at about 14 ka BP (McCabe et al., 1998). Deglaciation of the whole district was complete 13 ka (Clapperton and Sugden, 1977).

In the Aberdeen area, substantial ponding occurred between the slowly retreating East Grampian ice sheet and ice offshore (Thomas and Connell, 1985). The lakes gradually expanded southwards as the two bodies of ice ‘un-zipped’ in that direction (Simpson, 1954; Aitken, 1991; Appendix 1 Strabathie, Mill of Dyce).

On mainland Scotland, the eustatic lowering of sea level caused by the abstraction of water to form the continental ice sheets was more than offset by the isostatic depression of the land. Relative sea level was probably high during the periods of deglaciation, making the coastal ice streams particularly vulnerable to rapid retreat by the process of iceberg calving (Stewart, 1991). The resulting glaciomarine deposits and shorelines formed during deglaciation have been subsequently raised above sea level (Chapter 7). The St Fergus Silt Formation, dated to about 14 to 15 ka BP, was laid down when sea level stood at least as high as 12 m OD (Figure 48). The Spynie Clay Formation, associated with raised beaches lying at about 30 m above OD near Elgin, has yielded an arctic fauna and is comparable with the Errol Clay Formation of the Firth of Tay (Chapter 8). It was laid down in arctic conditions preceding the Windermere Interstadial (Peacock, 1999).

Chapter 6 Quaternary deposits

This chapter provides a systematic description of the drift deposits depicted on the eleven sheets covering the district (Figure 1). The deposits are classified here following the current BGS morpho-lithogenetic scheme (Figure 5). Names of units vary from sheet to sheet, but some of the most common synonyms are given in (Table 3) for comparison. Some deposits on the more recently published maps have also been classified lithostratigraphically and descriptions of these units are given in Chapter 8. Further descriptive accounts may be found in the older memoirs covering specific sheets, and in various sand and gravel assessment reports (Information sources). Geomorphological features are described in Chapter 7. The engineering properties of glacigenic deposits are described in Chapter 2.

A set of generalised maps showing the distribution of drift groups and a selection of deposits, landforms and localities mentioned in the text is provided at the back of this publication (Maps 1 to 11).

Glacial deposits

Till

Tills are the most widely distributed of the glacial deposits. They crop out over much of the district and also occur beneath younger superficial deposits. Tills are composed mainly of diamictons, materials that are characterised by a lack of sorting (in the geological sense). They are mostly matrix supported, dense, cohesive, commonly unstratified and comprise a mixture of rock fragments, gravel, sand, silt and clay (Table 6). Not all tills are diamictons and not all diamictons are tills. For example, many diamictons were formed as cohesive debris flows and mudslides, not necessarily in a glacial environment: others were formed by periglacial processes involving repeated freezing and thawing. The mapped tills probably include some of these nonglacial diamictons because it is not practical to delineate them separately.

Tills consist of ice-transported material laid down sub-glacially at the base of active glaciers and ice sheets, proglacially and supraglacially at the margins of retreating ice sheets, and paraglacially soon after glaciation when sediment deposited earlier is remobilised. Deposits formed in the proglacial, supraglacial and paraglacial environments commonly occur at the surface and comprise a metre or so of heterogeneous, very poorly sorted, crudely stratified, gravelly diamicton that is commonly intercalated with gravel, silty sand, silt and clay. These sediments accumulated at ice-sheet margins, mainly as debris flows that were modified and redeposited by ephemeral meltwater streams and sheet wash. They are generally permeable and include large boulders, up to several metres in diameter. In contrast, tills formed in the subglacial environment are relatively homogeneous (massive) and impermeable. In general, it is not practical to map out the various types of till separately. Locally, where proglacial, supraglacial and paraglacial deposits are several metres or more thick and form constructional mounds, they have been mapped out as ‘hummocky glacial deposits’ (see below).

The subglacially formed diamictons include lodgement tills and deformation (deforming-bed) tills, although the two types may be difficult to distinguish, resulting as they do from a continuum of processes. There is commonly a sharp boundary between deformation till and underlying units of pervasively sheared and deformed sediment or weathered bedrock. The amount of deformation in such units, which are now commonly referred to as glacitectonites (Table 6), generally decreases downwards. Some sandy deformation tills are weakly stratified on a scale of a few millimetres to a few centimetres and the stratification is formed of laterally impersistent laminae and wisps of pale, very fine-grained sand and silt. The presence of these laminae may indicate that some resedimentation in running water has occurred at the ice–till interface during melt-out, in which case the deposit may be classified as a melt-out till (Boulton, 1970; Shaw, 1979). However, the stratification can also result partially, if not wholly, from subglacial shearing and associated comminution of granular material (Boulton, 1979; Boulton et al., 1974). True melt-out and ablation tills are rare because they have poor preservation potential (Paul and Eyles, 1990).

Tills vary considerably in thickness and lithology across the district where they tend to underlie poorly drained ground of low relief and smooth slopes (Plate 8a), (Plate 8b), (Plate 8c), (Plate 8d). The tills of the district are described in terms of the five lithostratigraphical groups to which they are assigned (Figure 4). The groups relate to the major bodies of ice responsible for depositing the sediments (Chapter 8). However, the groups have been recognised only on the more recently published 1:50 000 maps. Diamictons belonging to different groups interdigitate locally, especially towards the coasts. The following account describes tills that are most likely to be encountered at the surface and in shallow excavations. They are predominantly Late Devensian in age, although much of the material forming them has probably been reworked from older glacial deposits and weathered bedrock. Information on the sparse, but stratigraphically important occurrences of older till is contained within Chapter 8 and Appendix 1.

Central Grampian Drift Group

This group of deposits, which mainly lies to the west of the River Spey (Figure 4), includes some of the thickest tills in the district. Sequences of glacigenic deposits with individual till units up to 15 m or so in thickness are common, and similar to successions described in detail nearer to Inverness by Fletcher et al. (1996). Tills occurring at the surface to the north of a line roughly between Rothes and Portsoy are predominantly reddish brown to brown sandy lodgement tills derived from sandstones and conglomerates of the Old Red Sandstone Supergroup cropping out to the west (Peacock et al., 1968; Aitken et al., 1979). However, a significant proportion of the clasts in these diamictons are ‘Moine’ (Grampian Group) psammitic granulites and quartzites together with minor amounts of local Permo–Triassic rocks, granite from Nairnshire, quartz porphyry and Jurassic limestone derived from the bed of the Moray Firth.

East Grampian Drift Group

Tills of this group are generally less than 5 m thick. They become patchy and are rarely thicker than 2 m across central Buchan. They are generally very sandy because they incorporate much decomposed, grussified rock (Chapter 4). Many are deformation tills derived largely from local decomposed rock, but with well-dispersed pebbles derived from farther afield. Rubbly, clast-supported diamictons up to about 1.5 m thick are common where the ice has overridden relatively fresh, quartzitic rocks. The colour, matrix and clast composition of the tills of this group, more than any other, reflect the nature of the underlying bedrock, or rocks cropping out within a kilometre or so ‘up-glacier’ (generally west or south-west). They are generally yellowish brown in colour, but reddish brown, orange-brown and olive-grey tills occur locally depending on the colour of the local bedrock. The tills are pale grey to white, kaolinitic and contain many well-rounded pebbles of quartzite, vein-quartz and flint where ice has crossed outcrops of the Buchan Gravels Formation (Chapter 4). The tills are very gritty inland from Aberdeen, where they include much decomposed granitic and psammitic bedrock (Figure 11). Where medium- to coarse-grained igneous rocks have been overridden, there are numerous boulders strewn across the surface that were probably formed as ‘core-stones’ within bedrock weathering profiles before being glacially transported.

Banffshire Coast Drift Group

The tills of this group are typically clayey, bluish grey to brown in colour and contain erratics, fossils, microfossils and shell fragments derived from the Moray Firth, as well as more local rocks. They crop out extensively at the surface only between Fraserburgh and Peterhead, but they occur patchily at depth beneath tills of the Central and East Grampian Drift groups between Elgin and Aberdeen (Figure 4). Thicknesses of 10 m or more are common. In the Elgin area, the tills are typically dark olive-grey to greyish brown and in addition to psammitic granulites and quartzites, include clasts of calcareous sandstone, glauconitic sandstone, black shale and shelly limestone from offshore (Peacock et al., 1968; Aitken et al., 1979). Fossils of Rhaetic to Early Cretaceous age are common. These dark tills of the Elgin area are almost certainly equivalent to diamictons within the Whitehills Glacigenic Formation, which has been recognised from Cullen eastwards (Chapter 8). Clasts of Jurassic mudstones and shales are common in these latter deposits, whereas the matrix has been formed partly from comminuted Mesozoic shales and partly from reworked Quaternary marine deposits from which the comminuted shell fragments have been derived. Fragments of Lower Cretaceous rocks and fossils are common in tills that outcrop farther east, towards Fraserburgh, where they include sparse clasts of chalk.

The diamictons of the Whitehills Glacigenic Formation are mostly deformation tills and glacitectonites. They commonly include lenses (boudins) and partially disaggregated masses of a range of Permo–Triassic, Jurassic and Cretaceous lithologies and Quaternary sediments (Plate 17a), (Plate 17b). They include dark grey to black mudstones, shales, clays and lignite, and yellowish brown fine-grained sands. Lenses of white, friable, kaolinitic sandstone are found sporadically in the west of the district, whereas vivid reddish brown and orange marls occur to the east of Troup Head. The lenses of all the aforementioned rock-types range in size from a few centimetres up to large glacial rafts several hundreds of metres across (Chapter 7). The latter have been confused with bedrock in some site investigations. Contact relationships between the rafts and enclosing diamictons are locally highly complex and in places confusing owing to considerable glacitectonic disturbance (Appendix 1 Boyne Limestone Quarry and Gardenstown).

Logie-Buchan Drift Group

This group underlies the coastal lowlands to the north of Aberdeen, east of Ellon and south of Peterhead (Figure 50). It comprises a complex sequence of relatively poorly consolidated, thinly interbedded, vivid reddish brown, clayey, calcareous diamictons, waxy clays, muds, sands and gravels. A few metres of stiff, stony lodgement till commonly occur at the base of the sequence, locally resting on yellowish brown lodgement tills of the East Grampians Drift Group. Otherwise the sequence is composed largely of laterally impersistent, and mainly cohesive, debris flow deposits (flow tills) that collectively may reach over 25 m in thickness, especially over hollows in the bedrock surface (Merritt, 1981; Munro, 1986) and close to contemporary ice sheet margins. The beds and lenses of sand within the sequence are typically fine to medium grained, silty and micaceous. Shell fragments are common. In addition to locally occurring rock-types such as amphibolite, psammite, quartzite and metagreywacke, the diamictons include rocks derived from the sea bed to the east, including dolomite, limestone, calcareous siltstone, both white and red friable sandstones and red-stained, rounded quartzite pebbles from Old Red Sandstone conglomerates to the south. Palynological analysis of the diamicton matrix has revealed complex assemblages, including Permo–Triassic pollen and Early Pleistocene dinoflagellate cysts (information from R Harland, 1980).

Mearns Drift Group

Tills of this group are invariably reddish brown in colour and contain a significant proportion of clasts derived from andesitic volcanic rocks, red sandstones, siltstones and mudstones of the Old Red Sandstone Supergroup cropping out in Strathmore. Unlike the Logie-Buchan Drift Group described above, the matrices are only weakly calcareous. Characteristically, they also contain rounded quartzite boulders, derived from Old Red Sandstone conglomerates. These clasts of sedimentary and volcanic rocks predominate in tills developed within the outcrop of the Old Red Sandstone, but they become less numerous where the tills extend onto the outcrop of adjacent Dalradian and Highland Border Complex rocks. These latter diamictons, which commonly abut and, in places, interdigitate with glacigenic sediments of the East Grampian Drift Group, contain clasts derived from local Caledonian igneous and Dalradian metamorphic rocks in addition to Old Red Sandstone material.

The tills are generally less than 5 m thick and are typically clayey, silty, pebbly diamictons that are matrix supported. Extremely overconsolidated clayey and silty lodgement tills are developed abundantly at the base of thick glacigenic sequences. More friable, sandy diamictons, generally less than 2 m thick, occur as flow tills capping deltaic spreads of glaciofluvial sand and gravel. Both of these types of till resemble diamictons within the Logie-Buchan Drift Group, but the presence of offshore-derived shell fragments and exotic clasts of dolomite, limestone and calcareous siltstone, which are only present in these deposits, serve to distinguish them from diamictons within the Mearns Drift Group.

Glacial deposits

Hummocky glacial deposits

Hummocky glacial deposits form a distinctive sedimentlandform association that has been identified on some of the more recently published maps. The deposits are highly variable lithologically and include complex interdigitations of matrix- and clast-supported diamicton, stratified and unstratified silty boulder gravel, and lenses of sand, silt and clay (Plate 9). Most are formed primarily of very poorly sorted materials that were deposited during deglaciation. They spilled from the ice-fronts as mud-slides and debris flows, either subaerially, or into standing water. Some of the deposits are constructional moraines formed at active ice margins, but most examples in the district apparently formed when large masses of ice stagnated. This situation commonly occurred towards the margin of the East Grampian ice sheet as it retreated westwards across north–south-trending ridges. For example, sandy morainic deposits derived largely from disaggregated granite occur quite extensively on Sheet 76E Inverurie in the lee of the Hill of Fare, between Banchory and West Cullerley [NJ 766 030] (Map 8). These mounds are commonly strewn with large boulders and blocks, but locally contain better-sorted materials at depth.

Similar deposits occur within a col at Tillyfourie and between the Hill of Fare and Cairn William, where they form irregular mounds and ridges with intervening peat-filled hollows (Map 8). These morainic deposits contain a high proportion of silty, matrix-supported bouldery diamicton as well as stratified sandy boulder gravel. Those at Tillyfourie are well seen from the A994 Alford–Aberdeen trunk road.

Hummocky glacial deposits are also common within the East Grampian Drift Group on Sheet 66E Banchory, where ice stagnated in the lee of high ground (Map 10). They occur extensively on the southern side of the Water of Feugh catchment and on the interfluve between the Dee, Carron and Cowie waters. Notable examples are present in the valley of Burn of Greendams, south-east of Strachan and in the valleys of the burns of Knock and Curran, to the south of Banchory. Most of the deposits form boulder-strewn lateral moraines, up to 10 m high in places, resting on granite bedrock. They are composed of boulder gravel with a coarse sandy matrix of disaggregated granite, derived from the underlying bedrock. Good examples of lateral moraine ridges, occur between Powlair [NO 621 912] and Green-dams [NO 649 900]. Similar landforms are present on the south-eastern side of the col between Craigbeg [NO 770 913] and Cairn-mon-earn [NO 783 920]. They were visible for many years from the A596 Banchory–Stonehaven trunk road (Slug Road), where it crosses the Dee–Strathmore watershed, but most of these ridges are now obscured by coniferous forest.

Few cross-valley retreat moraines have been recognised, but a good example is the Lady’s Moss Moraine [NO 782 901] in the valley of the Black Burn, south of Cairn-mon-earn (Auton et al., 1988). Other examples include north-east-trending morainic ridges in the col between Shillofad [NO 724 888] and North Dennetys [NO 710 877], and a north-west-trending ridge up to 18 m high, at Rouchan [NO 640 897] (Auton et al., 1990). The cross-valley moraine at Lady’s Moss and the morainic ridges between Shillofad and Kerloch lie close to the limit of the Aberdeen–Lammermuir Re-advance of Sissons (1967; (Figure 39)). The former moraines occur at elevations of 170 to 200 m OD, the latter at 275 to 320 m OD. They appear to mark still-stands in the late-stage retreat of small glaciers in the cols rather than the limit of a regionally significant ice advance. Although the concept of Sissons’ Aberdeen–Lammermuir Re-advance is now largely discredited, it was the recognition of morainic features such as these that first led to its proposition.

Constructional moraines were formed locally at the margin of the East Grampian ice sheet, especially where it ‘actively’ retreated from high ground and bedrock ridges trending at right angles to ice flow. In central Buchan, this situation led to the cutting of groups of north–southtrending meltwater channels (Chapter 7), but the ice was locally sufficiently active to form low morainic ridges parallel to the channels. Good examples of such ridges occur on Sheet 76E Inverurie in the vicinity of Burnhelvie [NJ 724 198] and Moss-side [NJ 710 183] (Map 8).

Push moraines appear to be rare in the district, but to the south of Ellon, two concentric ridges in the vicinity of Cross Stone Wood [NJ 955 278] are examples ((Figure 50); (Map 6)). The ridges stand up to about 8 m high and are apparently formed mainly of sand and gravel with a capping of red diamicton. The ridges are asymmetrical in cross-section, with steeper slopes facing south-westwards. They were probably formed at the margin of the coastal ice stream that was responsible for laying down the deposits of the Logie-Buchan Drift Group. Other moraines of this type lie at the southern end of the Den of Boddam [NK 102 408], where the coastal ice pushed northwards towards the Hill of Longhaven (Map 7). The form of all these features strongly suggests that the coastal ice was pushing towards ice-free ground.

Glaciofluvial and glaciolacustrine deposits

Glaciofluvial deposits

Glaciofluvial deposits are sediments laid down primarily by waters issuing from ice sheets and glaciers. The source of the water also includes rainfall and run-off from ice-free slopes as well as melting ice. Two main categories of deposit have been distinguished on the basis of their geomorphology and shown on all the 1:50 000 maps of the district, but the name applied to these categories varies from map to map (Table 3). On the more recently published maps, moundy deposits are classified as ice-contact deposits whereas terraced spreads are termed sheet deposits. The two categories are described in general terms below, before details are given. As glaciofluvial deposits are commonly laid down some distance away from former ice margins, the link between ice-source and deposit is less strong than in tills, and it is therefore more practical to describe their areal distribution rather than by lithostratigraphical group.

Glaciofluvial ice-contact deposits

Glaciofluvial ice-contact deposits consist mainly of sand and gravel, but include subsidiary beds of diamicton, silt and clay ( Plate 10a), (Plate 10b), (Plate 10c), (Plate 10d). The sediments were laid down in supraglacial, englacial, subglacial and ice-marginal drainage systems. Hummocky topography is characteristic, but flat-topped plateaux are also included. Steep-sided ridges of gravel (eskers) are shown individually on the 1:50 000 geological maps where space permits, whereas rounded hillocks of sand and gravel (kames) and flat-topped mounds (kame-plateaux) are not delineated from other, more irregularly shaped mounds and undulating spreads. Ice-contact deposits are typically associated with kettleholes, which are the result of buried masses of ice melting out slowly after deposition of the sediment surrounding and overlying them. Steep slopes that were formerly in contact with the ice are commonly still recognisable. The kames and kame-plateaux are composed typically of laminated muds that coarsen upward into sands and then gravelly deposits that formed as fan-deltas in ephemeral ice-marginal lakes. Lenses of diamicton and large boulders are common, having been deposited as debris flows, slurry and falls directly from the ice surface onto the accumulating glaciofluvial sediments.

The moundiness of ice-contact deposits is mainly the result of postdepositional collapse of the sediment bodies once the ice supporting them had melted. The sediments forming such mounds are commonly offset by normal faults. Reverse faults commonly occur where sediments have sagged into basinal structures (McDonald and Shilts, 1975). As glacier-ice rarely becomes totally stagnant as it melts out, many ice-contact deposits have been ‘nudged’, and in places more severely deformed, by minor glacial advances. Deposits formed at such active ice fronts are commonly disrupted by thrust faults and have been glacially overridden locally, leading to compaction and the deposition of till. Sequences such as these can be very complicated and therefore difficult to appraise and exploit commercially as aggregates.

The district does not include many large eskers and those that do occur are generally isolated, low-lying ribbon eskers that were laid down by subglacial meltwaters at retreating ice margins (Chapter 7). Eskers are formed mainly of gravel, but it is commonly very coarse and bouldery.

Glaciofluvial sheet deposits

Glaciofluvial sheet deposits were laid down mainly by braided streams in a proglacial environment leading to the accumulation of fan-shaped, elongate bodies and spreads of ‘outwash’ sand and gravel (sandur). Distinct terraces formed when parts of the outwash plain, or fan, were abandoned as streams graded to lower levels. Although sandar deposits generally become less coarse and more sandy downstream from the ice margin, many sheet deposits in the district coarsen upwards, which indicates that they accumulated during still-stands or minor glacial re-advances. Many sheet deposits pass upstream into glacial drainage channels.

Kettleholes do occur within sheet deposits, but they generally are not as large or as common as in the ice-contact deposits. The sheet deposits are also relatively more homogeneous and dense, less disrupted by faults and glacitectonic dislocations, contain fewer lenses of till, fewer large boulders and are consequently relatively easy to appraise and exploit commercially as aggregates.

Descriptions of glaciofluvial deposits by geological sheet

The resource potential of the glaciofluvial deposits is described more fully in Appendix 2.

Sheet 95 Elgin

A large part of this sheet is underlain by glaciofluvial deposits that form significant resources of sand and gravel. The deposits are varied, both in composition and distribution, and the observations given below are based on the more detailed descriptions of Peacock et al. (1968) and Aitken et al. (1979).

Most of the deposits west of the Spey valley are moundy and were laid down during deglaciation at the margin of ice occupying the Moray Firth as it retreated west-north-westwards. Meltwaters draining that ice, together with streams draining the partly deglaciated areas to the south, were constrained to flow around the retreating margin, creating deep drainage channels such as the Black Burn valley, to the west of the River Lossie, and the Blackhills channel, south of Lhanbryde (Chapter 7; (Map 1)). The Spey valley would also have functioned as an overflow channel. Much of the sand and gravel lying between Elgin and the Spey was laid down as fans at the mouths of these, and other minor channels, or as deltas in temporary lakes held back by the ice cap ( Plate 10a). Good examples of deltaic sand and gravel laid down by southeastward flowing meltwaters occur at Lochinvar Pit, near Elgin and at Rothes Glen pit [NJ 254 527], south of Elgin (Map 1).

Many of the deposits around Elgin were laid down on remnants of stagnant glacier ice that later melted to form kettle-holes and brought about the generally moundy topography. The deposits of glaciofluvial sand and gravel tend to coarsen upwards, but to fine north-eastwards away from each source. Kameplateaux amongst the mounds are commonly formed of fine-grained glaciolacustrine deposits capped by sand and gravel. Sand predominates to the north of the A96 trunk road, around Elgin and Lhanbryde, and it forms particularly large mounds towards the coast, for example at Binn Hill [NJ 305 656].

The glaciofluvial deposits to the north and north-east of Elgin become less moundy towards the coast where those occurring below about 35 m above OD appear to have been wave-washed locally. These undulating spreads are formed mainly of sand, silt and clay with subordinate gravel, and they merge with Late-glacial raised beach deposits, at about 30 m above OD, around Kinloss and Loch Spynie, indicating contemporaneous glaciofluvial and marine sedimentation at the former ice front (Chapter 7). To the east of the Spey, two gently undulating to moundy spreads of sand and pebbly sand lie between Fochabers and Buckie.

A distinctive suite of deposits lie between Kinloss and Lossiemouth where several east-north-east-orientated ridges, with flat tops descending towards the east-north-east, are separated by depressions. The sand and gravel forming the ridges tends to coarsen upwards into boulder gravel and the sequences are locally capped by till. The depressions are underlain by silts and clays, and parts of the ridges are formed of these lithologies too, commonly displaying complex soft-sediment gravitational deformation structures. The sands and gravels forming the ridges were probably laid down within ice-walled chasms at the ice margin, which possibly connected with the sea. The silts and clays were deposited following retreat of the ice margin westwards. A similar style of sedimentation, possibly ice-proximal glaciomarine, has been proposed for the Ardersier Silts and Alturlie Gravels formations of the Banffshire Coast Drift Group occurring farther to the west (Merritt et al., 1995; Fletcher et al., 1996).

A prominent feature of the lower Spey valley is a series of gravelly terraces that formed during the whole period between the retreat of the last glaciers and the present day. The river floodplain and low-lying alluvial terraces are deeply incised into a set of older late-glacial ones. The most extensive ‘Mosstodloch’ terrace, and higher terraces to the south of Mosstodloch, are regarded as glaciofluvial because they contain kettleholes or merge into moundy glaciofluvial deposits (Map 1). Lower terraces are probably mainly of Loch Lomond Stadial age, or younger. The Mosstodloch terrace merges downstream into moundy ice-contact deposits around Garmouth, indicating that ice still lay to the north-west when the terrace formed. Ice probably completely blocked off the lower reaches of the Spey valley downstream of Fochabers at an earlier stage, causing a large lake to be ponded up — the ‘Fochabers Glacial Lake’ (Peacock et al., 1968). Thinly laminated glaciolacustrine clays underlie parts of the Trochelhill and Balnacoul glaciofluvial terraces in the vicinity of Fochabers.

Sheet 96W Portsoy and sheet 96E Banff

The glaciofluvial sands and gravels occurring on these sheets have been divided into two assemblages. The predominantly moundy deposits inland of Cullen, Portsoy, Whitehills and Banff have been placed in the Blackhills Sand and Gravel Formation of the Banffshire Coast Drift Group (Chapter 8). Deposits in the valley of the River Deveron, upstream of Bridge of Alvah, are not divided, and are placed in the East Grampian Drift Group, together with terraced deposits in the valleys of the burns of Boyndie, Durn, and Deskford (Map 3).

The Blackhills Sand and Gravel Formation includes shell fragments and sparse clasts of sedimentary rocks and fossils derived from the floor of the Moray Firth in addition to rocks occurring locally, or farther to the west and south-west. Many of the well-rounded clasts in these sediments have been derived from Old Red Sandstone conglomerates. Some deposits accumulated as outwash fans at the margin of ice retreating westwards. Other deposits accumulated as deltas in lakes held up by glacier ice that remained offshore (Figure 42), abutting the coast. A reference section occurs in Brandon Howe pit [NJ 667 637] within the Hills of Boyndie, south-west of Banff, where over 12 m of sand with minor lenses of gravel and widespread seams of silt was formed in a distal outwash fan that pro-graded south-eastwards. Sands and silts displaying soft-sediment deformation due to loading and dewatering are exposed in a small pit at Tipperty [NJ 671 608] ((Map 3); (Plate 11)).

Degraded glaciofluvial terraces lie within several of the valleys, particularly those of the Deveron and the Burn of Boyndie. Unlike the Blackhills Sand and Gravel Formation deposits, which were laid down mainly by meltwaters flowing south or south-eastwards from ice situated to the north, these deposits were laid down by meltwaters flowing towards the Moray Firth. The deposits therefore contain clasts derived from the south, including basic igneous rocks that are commonly weathered and friable. Such deposits are exposed at Danshillock pit ((Map 3); (Plate 10c).

Sheet 97 Fraserburgh

About one third of this sheet is underlain by glaciofluvial and glaciodeltaic deposits, which occur mainly in a broad swathe between New Aberdour and Loch of Strathbeg (Map 4). Apart from fragmentary terraces in the valley of the North Ugie Water, most of the deposits have been assigned to the Blackhills Sand and Gravel Formation (Chapter 8). These deposits contain clasts and shell fragments (including Nuculana pernula) derived from the floor of the Moray Firth, in addition to locally occurring clasts. Fragments of porous limestone and spiculitic sandstone of Cretaceous age are common. Both moundy and terraced spreads are equally represented on the sheet.

The moundy deposits were laid down as fans and deltas within lakes dammed up against ice of the Moray Firth ice stream to the north ((Figure 42)b). Typical deltaic coarsening-upward sequences of sand and gravel occur at two pits in the vicinity of Blackhills [NJ 924 612] (Plate 10b), the type locality of the formation, and at Pitnacalder [NJ 872 628], near New Aberdour. Deposition was mainly from flow directed towards the south-east. Several of the moundy sediments were glacitectonically disturbed during, and shortly after deposition, by the ice to the north, which remained active and locally re-advanced to lay down a local capping of till (Merritt et al., 2000, fig. 26). One of the larger moundy deposits form the Sinclair Hills, south of Fraserburgh (Map 4).

There are two major spreads of terraced glaciofluvial sand and gravel on the sheet. Both were formed by braided meltwater streams flowing east or south, and then towards the sea, which, at the time, stood no higher than 5 m above OD (Peacock, 1997). Meltwaters also flowed south-eastwards into the valley of the North Ugie Water to lay down terraces there, and others formed the upper part of an outwash fan at Todholes [NJ 843 570], where an important series of events has been established at Howe of Byth pit (Appendix 1).

Sheet 86E Turriff

The glaciofluvial deposits on this sheet are mainly restricted to the degraded terraces within the valley of the River Deveron north of Turriff, the ‘misfit’ valleys of the Idoch Water and Burn of King Edward, and the upper reaches of the Ythan upstream of Fyvie (Map 5). It is likely that the present northward drainage of the Deveron was re-established only at a late stage in the deglaciation of the district. Before then, while ice still occupied the Moray Firth and the coast, meltwaters discharged southwards into the Ythan system via a major spillway at Towie [NJ 746 440]. Substantial terraced deposits of sand and gravel were laid down around Turriff, and underlie the site of the town’s new sewage works. An excavation into a high terrace there revealed imbricate gravels deposited by southerly flowing water. Meltwaters also flowed southwards through the Afforsk spillway, south-east of Gardenstown (Map 4), where they fed into the catchment of the Burn of King Edward. Extensive terraces occur at the lower end of that valley, sloping towards the valley of the River Deveron. An isolated, but prominant terrace lies to the north of New Byth, at Tippercowan [NJ 815 551], whereas several north-north-east or south-southwest orientated kames and eskers lie between the Deveron, north of Turriff, and Aberchirder.

Sheet 87W Ellon

Glaciofluvial deposits are not common on this sheet, especially the moundy variety. This is a paradox, because there is extensive evidence of the activity of glacial meltwaters in the form of drainage channels (Map 6). Furthermore, many of the deposits that do exist were laid down by meltwaters entering the district from the north, via the valleys of the River Ythan and North Ugie Water, at a relatively late stage in the deglaciation of the region. Part of the explanation for the apparent scarcity of glaciofluvial deposits may be that the meltwaters laid down sands and gravels in the deeper valleys where they are now concealed beneath alluvial, solifluction and gelifluction deposits.

Glaciofluvial terraces lie within the valley of the Ythan, downstream of Methlick, and in the valley of the North Ugie Water, downstream of Strichen. Some of the deposits of the Ythan are locally very coarse indeed, especially at Bellmuir [NJ 875 365] and Ardlethen [NJ 920 320], and might have been deposited during catastrophic flood events (Maizels and Aitken, 1991). The terraces in both valleys were partly built up by meltwaters entering the main valleys from tributaries, to the north and west, during westward retreat of the East Grampian ice sheet. The gravels in both valleys tend to become finer downstream, where the lower lying terraces rest on extensive fine-grained glaciolacustrine deposits and descend to the level of the main postglacial beach. Downstream of Ellon, a former valley of the River Ythan lies to the south of the present one and is largely filled with glaciofluvial deposits (Chapter 7). Glaciofluvial terraces also occur in the catchment of the South Ugie Water, in the vicinity of Stuartfield and Old Deer.

The deposits described above have all been assigned to the East Grampian Drift Group, together with a solitary east-north-easttrending esker near Auchorthie [NJ 923 523]. Other moundy glaciofluvial deposits are situated close to the boundary of the Logie-Buchan Drift Group deposits in the south-east part of the sheet. This boundary is most clearly defined to the south of Ellon in the vicinity of Cross Stone, where a pit [NJ 9523 2822] reveals about 5 m of poorly bedded gravel becoming more sandy upwards. The clasts include gneisses, quartzites, psammites, mica-schist and granites, which are all typical of deposits derived from the East Grampian ice sheet, and the bedding appears to indicate an easterly palaeocurrent. There are glacitectonic structures present that have been caused by ice pushing from the east, and the deposit is locally capped by red till. It appears that the deposit formed at the margin of the ‘Logie-Buchan’ ice, which pushed inland to form the two arcuate, asymmetric moraine ridges that lie to either side of the deposit (Map 6).

Several moundy deposits occur at the boundary of the two groups in the vicinity of the Hill of Auchleuchries [NK 006 365]. However, the sandy deposits forming the hill (Auchleuchries Sand and Gravel Formation; (Table 7)) are capped by yellowish brown till of the East Grampian Drift Group, not red till, and at Tillybrex, 1.5 km to the south-southwest, at least 11 m of weathered gravels (Tillybrex Sand and Gravel Formation; (Table 7)) with no morphological expression, are capped by red till. The Tillybrex gravel was deposited by westward flowing meltwater. There are clearly complicated sequences preserved along the boundary of the two groups. For example, the gravels at Tillybrex almost certainly predate the Late Devensian, and other older glaciofluvial deposits crop out from beneath till in the vicinity of Leys pit [NK 004 524] and at Oldmill [NK 023 440] (Appendix 1 Kirkhill and Oldmill).

Sheet 87E Peterhead

The glaciofluvial deposits occurring on this sheet have been assigned to three groups (Map 7). Most of the deposits lie close to the boundaries of the former East Grampian ice sheet with the coastal ice streams that laid down the Banffshire Coast and Logie-Buchan drift groups (shown on the published map as inland, blue-grey and red series respectively). Extensive spreads of terraced sand and gravel lie at the confluence of the North Ugie and South Ugie waters, where they overlie up to 25 m of laminated silts and clays. The sediments were deposited by meltwaters flowing from the west into lakes held up against ice of the deflected Moray Firth ice stream to the east (Figure 42). This took place after the East Grampian ice sheet had begun to retreat westwards.

In the south part of the sheet, glaciofluvial deposits are commonly interbedded with red diamictons and silts of the Logie-Buchan Drift Group, but individual beds are generally too thin to map out, or they are concealed. Several, more extensive deposits occur around Hatton. A pit in the village [NK 054 371] revealed over 10 m of deltaic calcareous sand with seams of red diamicton beneath a drape of red clay. The sequence included a bed of brown diamicton possibly derived from East Grampian ice. A ridge of pebbly sand lying to the west of the village, at the boundary of the Logie-Buchan and East Grampian drift groups, is also capped by red till although the pebbles were probably derived from East Grampian ice. This body of sand coarsens upwards and is possibly similar to the deposit forming the Hill of Auchleuchries, on Sheet 87W Ellon, and both deposits rest on dark grey-coloured tills.

One of the best examples of a ‘ribbon’ esker in the entire district lies to the east of Meikle Loch (Plate 15). The Kippet Hills esker (Appendix 1) stands up to 15 m high and is partially buried beneath red diamicton. It appears to connect with three mounds of sand and gravel lying to the north. The easternmost mound at Knapsleask is flat-topped and formed as an outwash fan. The fan and esker were laid down by meltwaters flowing northwards through a lobe of ice that pushed onshore between Aberdeen and Peterhead (Figure 50). The deposits contain a significant proportion of dolomite, limestone and calcareous siltstone from the sea bed, together with shell fragments derived from the offshore Aberdeen Ground Formation.

Several isolated mounds of glaciofluvial and glaciodeltaic sand and gravel lie within the area formerly covered and weakly glaciated by the East Grampian ice sheet. Some of the deposits have poor surface expression and probably predate the last glaciation, like the one at Oldmill (Appendix 1). Others contain materials derived locally; for example, the deposit at Redleas [NK 093 430] is composed mostly of fragments of weathered granite, and gravelly deposits around the Buchan Ridge contain many rounded clasts derived from the Buchan Gravels Formation.

Sheet 76E Inverurie

Glaciofluvial deposits are widespread on this sheet (Map 8). They were laid down mostly at the margin of the East Grampian ice sheet as it retreated westwards across the area. The pattern of retreat was greatly affected by the topography, especially by the high ground formed by Bennachie, Cairn William, Green Hill and the Hill of Fare in the west of the sheet. As the ice margin reached this crescentic ridge that reaches 300 to 500 m above OD, meltwaters carved out several major channels across cols and deposited sands and gravels on, and amongst, stagnant ice left stranded in the lee of the hills. Some of this stagnant ice blocked the engorged stretch of the Don valley upstream of Port Elphinstone, causing meltwaters to be diverted eastwards via another set of channels across the lower ridge lying between Kemnay and Inverurie. Water from these channels debouched into the valley of the River Don in the vicinity of Kintore, giving rise to fans of sand and gravel. A similar ice blockage also occurred farther upstream, east of Castle Forbes [NO 622 191]. It caused meltwaters to be diverted northwards, where they cut a precipitous gorge into fresh granite bedrock called ‘My Lord’s Throat’ [NJ 635 197].

A good example of one of the routes taken by meltwaters across the sheet can be demonstrated around Kemnay. Meltwaters passed through the col crossed by the A944 road at Tillyfourie, where they laid down moundy glaciofluvial gravels and cut channels through morainic deposits dumped there. The water passed through stagnant ice to the east of the col to lay down the gravels that now form eskers in the valley of the Ton Burn and beyond towards Kemnay. The main esker terminates to the north of Kemnay amongst moundy deposits of sand and gravel. From here, meltwaters passed through the deep channel at Tom’s Forest and laid down the large fan of sand and gravel at Tavelty, just to the north of Kintore. At a slightly earlier stage in the deglaciation, meltwaters flowed directly eastwards via channels at Leschangie to lay down a sheet of sand and gravel to the south of Gauch Hill [NJ 787 151].

The gorge of the River Don upstream of Monymusk also served as a major conduit for meltwaters across the high ground. Initially, meltwaters flowed eastwards across the col (about 265 m OD) between Bennachie and Millstone Hill cutting channels through the morainic mounds in the valley bottom. At Pitfichie, they took a more direct route eastwards than the present river and laid down moundy deposits of sand and gravel around Blairdaff [NJ 703 180] (Appendix 1 Rothens). Other routes taken by meltwaters across the high ground led to the cutting of deep channels across several cols between the Hill of Fare and Green Hill [NO 636 098].

A similar pattern of westward ice retreat occurred in the south-eastern part of the sheet, where ice stagnated in the lee of the Hill of Fare. One route taken by subglacial meltwaters led to deposition of moundy deposits, including an esker, between East Finnercy [NJ 767 042] and Roadside of Garlogie. Another route is traced by the esker forming Horsewell Hillocks [NJ 779 019], west of Hardgate. The valley of the Corskie/Kinnernie Burn carried meltwaters that laid down moundy deposits of gravel, including eskers, upstream of Dunecht [NJ 754 091], and more sandy, terraced deposits downstream towards the Loch of Skene. Sandy glaciofluvial deposits also lie to the east of the loch, beneath alluvium, and they form an extensive spread in the valley of the Leuchar Burn downstream of Garlogie [NJ 782 057].

Glaciofluvial sands and gravels form mounds 5 to 15 m high and low-lying terraces around Torphins and Gallow Cairn [NO 639 010]. These deposits are associated with a major former route of glacial drainage along the valley of the Burn of Beltie. This drainage route continues south-eastwards, along the valley of the Burn of Canny, to meet the valley of the River Dee at Invercanny Waterworks, 2 km upstream of Banchory on Sheet 66E. Smaller mounds of gravel and sand south of Milltown of Campfield [NJ 648 004] almost block modern drainage along another former route of meltwater flow that also extends on to the adjoining Banchory sheet (see below).

The north of the sheet contains ground that is similar in many respects to that covered by the Ellon sheet, on which moundy glaciofluvial deposits are relatively sparse and terraced deposits are restricted to the main valleys. The valley of the River Urie was ponded up for a while during deglaciation by ice that blocked the valley of the River Don between Inverurie and Kintore. A large lake formed upstream of Inverurie in which deltas of sandy gravel were laid down. Many of these deltaic deposits now form terraces on either side of the Urie.

Sheet 77 Aberdeen

Two contrasting suites of glaciofluvial deposits occur on this sheet (Map 9). Most have been assigned to the East Grampian Drift Group, but moundy red, silty sands and gravels of the Logie-Buchan Drift Group lie along the coast, extending inland for up to about 3 km to the north of Aberdeen. The latter deposits were laid down at the boundary between the East Grampian ice sheet and ice that pushed onshore from the North Sea. During deglaciation, the parting (‘unzipping’) of the two bodies of ice appears to have started in the north, leading to ponding of meltwaters between them in the ice-free enclave thus formed (Figure 42).

The deposits of the Logie-Buchan Drift Group in the north of the sheet include a ridge that extends from Drums [NJ 990 227] towards Newburgh, where it has been truncated by the River Ythan (Figure 50). A short fragment of the ridge lies on the northern bank of the Ythan at Waterside, and the meltwaters that formed it probably continued northwards to form the Kippet Hills esker at Slains, on Sheet 87E Peterhead. To the north of Drums the ridge is locally boulder strewn and probably formed at the landward margin of ‘Logie-Buchan’ ice. To the south, the ridge merges into a belt of moundy deposits formed of interbedded reddish brown sands, gravels, silts and diamictons. The belt arcs around low-lying ground at Balmedie [NJ 964 177], which appears to have been occupied by ice while the sequence was laid down.

Most of the glaciofluvial deposits on the sheet were laid down by precursors of the rivers Don and Dee. Those of the latter river occur mainly as terraces within the present valley, although deposits lying to the south of Torry indicate that the river formerly took a more direct route to reach the sea at Nigg Bay (Chapter 7; (Figure 46)). Meltwaters also flowed into the Dee valley from the north, but they did not lay down deposits quite as extensively as shown on the 1:50 000 map published in 1980. Likewise the belt of deposits lying between Cadgerford [NJ 835 056] and Blackburn, and between Craibstone [NJ 867 111] and Bankhead are also not as extensive as shown on that map (Appendix 2).

The River Don formerly took a direct route to the coast instead of turning southwards at Dyce. This was probably partly due to ice blocking the valley downstream of Dyce and partly because the river was flowing towards the sea at a higher level than today. Large resources of sand and gravel in the form of mounds and ridges formerly occurred between Corby Loch [NJ 925 145], a large kettlehole, and the coast at Blackdog [NJ 963 141], but many of the deposits have been worked out (Appendix 2). Gravels laid down from eastward flowing braided rivers predominate around Bishops’ Loch and Corby Loch (Aitken, 1995, 1998), but many of the deposits to the east formed as deltas and consequently contain sequences that coarsen upwards from silty sands into coarse gravels. Particularly good sections in deltaic deposits were formerly exposed at Strabathie [NJ 955 136] (Appendix 1), in a pit that is now used as a landfill site.

A particularly large deposit of sand and gravel lies in the middle of the valley of the River Don in the vicinity of Liddell’s Monument [NJ 869 153], where it has been exploited at Mill of Dyce pit (Appendix 1). This deposit formed as a fan delta at the margin of the East Grampian ice sheet when it was situated immediately to the west. The ice then retreated upstream and a lake was held back behind the deposits, which blocked the valley. Meltwaters entered the main valley from side valleys to form kame terraces, especially around Hatton of Fintray. The ice front probably stood in the vicinity of Straloch [NJ 860 211] at about the same time as the Mill of Dyce deposit formed. However, although terraced deposits of gravel were laid down between there and Newmachar, they are not as extensive as shown on the 1:50 000 map published in 1980 (Appendix 2).

Sheet 66E Banchory

The glaciofluvial deposits on the northern two thirds of the sheet have been assigned to the Lochton Sand and Gravel Formation of the East Grampian Drift Group, whereas those in the south belong to the Drumlithie Sand and Gravel Formation of the Mearns Drift Group (Chapter 8). The deposits of the former group were laid down at the margin of the East Grampian ice sheet as it retreated slowly westward (Map 10). Clasts of coarse-and fine-grained granite predominate, but gritty and micaceous psammite and schistose semipelite become more numerous in gravels in the valley of the River Dee, downstream of Banchory.

Much of the meltwater flowed towards the valley of the River Dee where substantial amounts of sand and gravel accumulated as kettled terraces. The thickest known deposits in the valley form a moundy spread in the vicinity of Park Pit, Drumallan [NO 797 978], where a minor still-stand occurred during ice retreat (Brown, 1993). Significant meltwater flow (directed towards the east and north-east) also took place to the north of the Dee valley, along channels linking linear ice-scoured basins. Moundy and terraced spreads of sand and gravel flank the basin north of Upper Lochton and others occur along the Bo Burn, south of the Raemoir Hotel; similar sands and gravels are also concealed beneath alluvial deposits.

Large amounts of meltwater entered the Dee valley from the south via the valleys of tributaries such as the Feugh and the Burn of Sheeoch. Once the ice front had retreated, meltwater drainage along the Dee valley was generally unrestricted. However, the valleys of the southern tributaries lay more or less parallel to the retreating ice margin and as a result drainage was impeded by masses of decaying ice. Meltwaters were ponded up and sequences of laminated silts, coarsening upwards into sand and cobble gravel, were laid down as fan-deltas. These deltaic sediments now commonly occur as kettled terraces and as flat-topped and irregular mounds. A most impressive kettled terrace (the Pitdelphin Wood Terrace), which stands up to 30 m above the level of the floodplain of the Water of Dye (Auton et al., 1990), occurs between Pitdelphin Farm [NO 654 912] and Bogarn [NO 657 903]. The upward coarsening glaciofluvial sand and gravel forming the terrace averages about 13 m in thickness and overlies glaciolacustrine silt and clay resting on till.

Large masses of stagnant ice were stranded in the lee of hills as the margin of East Grampian ice retreated westwards, giving rise to hummocky glaciofluvial and morainic deposits and many eskers. Eskers are most common along the southern side of the valley of the River Feugh, in the vicinity of Strachan [NO 674 923], and in the valley of the Burn of Sheeoch farther east; they are formed mainly of coarse, clast-supported gravel. A group of eskers that have an unusual south-east trend occurs on the northern side of the valley of the River Feugh (Map 10). They contain clasts of fine-grained basic igneous rocks, which are generally absent in the adjacent glaciofluvial gravels such as exposed in Cammie Wood pit (Plate 10d); (Map 10)).

The glaciofluvial deposits in the south of the area were formed mainly when the East Grampian ice sheet began to retreat northwestwards and the Strathmore ice stream retreated south-westwards. Meltwaters formed lakes between the two ice masses, which acted as sediment traps in which coarsening-upward deltaic sequences were laid down. At first the ponding took place high in the valleys to the south-east of the main watershed, for example, in the upper reaches of the Carron Water around Snob Cott [NO 801 885], where a considerable thickness of coarse-grained sand and gravel was laid down in contact with decaying ice. These deposits, assigned to the Lochton Sand and Gravel Formation, are characterised by clasts of granitic and metamorphic rocks. Ponding subsequently occurred at lower levels in minor valleys, such as those around Pitdrichie [NO 795 825], and the valley of the Burn of Caldcotts, upstream of Fettercairn, before the valleys of the Bervie Water and Luther Water were inundated. Deposition of sand and gravel derived from southeastward flowing meltwaters from the retreating East Grampian ice sheet was concentrated around Auchenblae, where a large fan was laid down flanking the Luther Water.

Gravels of the Drumlithie Sand and Gravel Formation include moundy ice-contact sands and gravels interbedded with reddish brown flow-tills, which were laid down close to the glaciofluvial fan at Auchenblae. However, these moundy deposits were banked against the Strathmore ice as it retreated south-westwards. All of the sand and gravel deposited by meltwaters from the Strathmore ice stream in the Pitrichie, Glenbervie and Auchenblae areas, contains clasts derived from the adjacent Devonian bedrock and psammitic and granitic clasts, most of which were derived from the upland to the north-west. Most of the durable clasts within the gravels around Auchenblae consist of quartzite and psammite pebbles, the bulk of which were derived from nearby outcrops of Old Red Sandstone conglomerate. A significant proportion of the psammite and many of the weathered granitic clasts can be matched locally with adjacent outcrops of Dalradian and Caledonian granitic bedrock.

Much of the coarse gravel which forms mounds on the flanks of the western headwater tributary of the Burn of Caldcotts (the Burn of Balnakettle) is assigned to the Lochton Sand and Gravel Formation as it contains no Devonian clasts. A typical gravel deposit, 3.7 m thick, containing small intraformational ice-wedge casts, was exposed in a small working [NO 620 752] north-east of Mains of Balnakettle. Farther up the valley, a coarsening-upward sequence of reddish brown sands and gravels of the Drumlithie Sand and Gravel Formation, with abundant clasts of Devonian rocks, is exposed in the back-scar of a landslip [NO 618 757] on the southwestern side of the burn (Appendix 1 Balnakettle). The sand and gravel has been tectonised and overridden by East Grampian ice which laid down a thin, moderate brown, silty till, containing a preponderance of local Dalradian schistose psammite and semipelite clasts. These topographical and stratigraphical relationships suggest that a minor re-advance of East Grampian ice took place. This occurred after the initial retreat of Strathmore ice and deposition of its outwash on the flanks of the upland.

The deposits of the Drumlithie Sand and Gravel Formation within Strathmore itself are typically vivid reddish brown in colour and relatively silty. They contain abundant clasts of friable sandstone and mudstone and decomposed andestic lava. Many deposits are concealed beneath red, clayey flow tills, as for example, to the east of Auchenblae and to the south-east of Drumlithie. The final south-westward retreat of the Strathmore ice stream was marked by the deposition of eskers in the vicinity of Bomershanoe Wood [NO 734 754], Meikle Fiddes [NO 799 808] and Little Wairds [NO 798 785].

Sheet 67 Stonehaven

As on the adjoining Sheet 66E Banchory to the west (see above), the glaciofluvial deposits on the northern half of the Stonehaven sheet have been assigned to the Lochton Sand and Gravel Formation, whereas those in the south belong to the Drumlithie Sand and Gravel Formation. Again, ponding occurred between the East Grampian ice sheet and the Strathmore ice stream, with the latter retreating south-westwards as deglaciation proceeded (Map 11). The ponding occurred mainly in the valleys of the Cowie and Carron waters, but several minor valleys descending toward the North Sea were also affected. This led to the deposition of coarsening-upward, sandy deltaic sequences in the vicinity of Cantlayhills [NO 881 905]. Meltwater drainage carved a deep drainage channel between Ury Home Farm [NO 859 882] and Logie [NO 886 888] and laid down a fan of silty cobble gravel at its eastern end. The channel was formed near the front of the Strathmore ice stream, during a still-stand in its retreat, and shortly after its parting from the East Grampian ice sheet.

Thick, coarsening-upward, deltaic-style deposits form mounds on the sides of both major valleys. However, whereas in the Cowie catchment the deposits contain mainly durable clasts derived from metamorphic rocks and granites, those of the Carron catchment are dominantly more sandy and include friable sandstone and mudstone clasts and are commonly interbedded with reddish brown silt and clay. The terraced spreads of coarse gravel in the floor of the valley of the Cowie Water, which have been extensively worked inland from Stonehaven, were laid down by meltwater from the East Grampian ice sheet draining south-east towards the coast. These granite and psammite-dominated gravels were deposited on top of reddish brown glaciolacustrine silts and clays (Ury Silts Formation) of the Mearns Drift Group. The gravels contain isolated ‘rip-up’ clasts of reddish brown clay.

Meltwaters flowing north-eastwards from the retreating Strathmore ice stream deposited eskers that connect with moundy deltaic deposits in the vicinity of Brucklaywaird [NO 826 840], Muirtown of Barras [NO 837 813] and Fawsyde [NO 846 772]. Kame terraces and moundy deltaic deposits were also formed at several localities where the north-western boundary of the retreating Strathmore ice stream withdrew south-eastwards from high ground. For example, deposits occur in the vicinity of Foggie Brae [NO 827 845], Dunnottar Square [NO 864 848], to the west of Roadside of Catterline [NO 860 792] and at the mouth of the Catterline Burn [NO 866 777]. Moundy kame deposits occur at Uras Knaps [NO 876 810] and Greenden [NO 811 776], and terraced deposits lie on both banks of the Bervie Water at Inverbervie.

Glaciolacustrine deposits

Fine-grained sand, silt and clay laid down in standing water form part of many deposits categorised as ‘glaciofluvial’, especially those forming moundy topography. They are especially common at the base of coarsening-upward deltaic sequences, and interbedded with sands and gravels of distal outwash fans. However, only those occurring at the surface and that are extensive enough to justify mapping out separately have been identified as glaciolacustrine deposits. They are typically thinly laminated, micaceous, and commonly contain dropstones. These sediments were deposited mainly in ice-marginal, proglacial lakes that formed in the upper reaches of valleys when their lower reaches were blocked by ice. Such lakes formed extensively during deglaciation as the East Grampian ice sheet withdrew from the coasts while the coastal ice streams remained largely intact (Figure 42). Several of the more extensive glaciolacustrine deposits in the district have been worked in the past for clay for making bricks, tiles and drainpipes (Chapter 2).

Description of glaciolacustrine deposits by geological sheet

Sheet 95 Elgin

Fine-grained, glaciofluvial and glaciolacustrine micaceous sands, silts and clays underlie much of this sheet area (Map 1). Although glaciolacustrine deposits have been mapped only to the west of Elgin, similar sediments underlie much of the ground farther east, mapped as moundy glaciofluvial deposits and as raised marine and estuarine alluvium. It is commonly difficult to distinguish between the various categories of fine-grained deposit as they tend to merge one with another. The pattern of deglaciation and sedimentation was notably complicated. The deposits give rise to a particular soil association (Grant, 1960).

Most of the mapped glaciolacustrine deposits occupy kettle-holes within moundy glaciofluvial deposits. Those occurring above about 30 m OD formed while ice remained in the Moray Firth, impeding drainage. Those at lower levels may have been laid down in brackish water at the time when the sea first transgressed into the area during deglaciation. The silts and clays vary in colour, from red to brown, yellowish brown and grey. They are commonly thinly interbedded with sands, gravels and pebbly, clayey diamicton.

Evidence of an earlier period of ponding can be found in the valley of the River Spey immediately upstream of Fochabers, where varved glaciolacustrine clays underlie gravelly river terrace deposits (Peacock et al., 1968). Ponding occurred extensively farther up the Spey valley to higher levels.

Sheets 96W Portsoy, 96E Banff And 97 Fraserburgh

Glaciolacustrine deposits belonging to the Kirk Burn Silt Formation of the Banffshire Coast Drift Group occur along the coast between Portknockie and New Aberdour (Map 2); (Map 3). They generally underlie flat or gently undulating ground within 5 km of the coast and consist of ochreous to dark brown, thinly laminated silts and clays up to about 10 m in thickness. Bands of ferruginous nodules are common locally. The deposits were laid down after the East Grampian ice sheet had retreated from the coast, but when the Moray Firth ice stream remained offshore, forming a barrier against which meltwaters ponded to the south (Figure 42). Meltwaters entering the lakes from the south commonly formed deltas, as for example, in the valley of Burn of Boyne, near Drakemires [NJ 603 613], where sand and gravel overlies dark olive-grey silts and clays.

Most of the ponding took place during the final stages of deglaciation, but there is evidence on Sheet 96W that it also occurred during an earlier episode, before the last glacial advance across the hinterland. For example, 3.5 m of dark yellowish brown, thinly laminated silts and clays with dropstones crop out from beneath 2 m of stiff lodgement till in the valley of the Burn of Deskford near Inaltry [NJ 517 630] (Map 2). Additionally, red clay crops out beneath gravels and till in the valley of the Burn of Fordyce near Ardiecow [NJ 532 616], and laminated sands are overlain by red till in the valley of the Burn of Fishrie at the Mill of Minnonie [NJ 776 604].

Few glaciolacustrine deposits have been mapped on Sheet 97, but many of the moundy deposits there include silts and clays at the base of coarsening-upwards deltaic sequences. Terraced sands and gravels overlie extensive deposits of dark grey sandy silt and clay in the vicinity of Savoch [NK 047 588], and they probably pass beneath the alluvium surrounding the Loch of Strathbeg. The dark coloured deposits were possibly formed in a glacio-estuarine setting and are equivalent in age to the raised glaciomarine St Fergus Silts Formation occurring a few kilometres to the south-east.

Sheets 86E Turriff, 87W Ellon and 87E Peterhead

No silts and clays of glaciolacustrine origin have been mapped on Sheet 86E, but they do occur locally beneath some of the terraced glaciofluvial deposits, as for example, in the vicinity of Turriff. Glaciolacustrine deposits are most widespread on Sheet 87E and on the eastern margin of Sheet 87W, where meltwaters were ponded up against the coastal ice that occupied the North Sea and laid down the Logie-Buchan Drift Group, after the East Grampian ice sheet had receded westwards (Figure 42). Most of the ponding occurred in the valley of the River Ythan downstream of Ythanbank, where red and brown laminated silts and clays up to about 10 m in thickness normally underlie glaciofluvial terraces and the floodplain alluvium (Map 6). Outcrops are rare, however, except on the southern side of the river immediately upstream and downstream of Ellon. The glaciolacustrine deposits extend eastwards beneath glaciofluvial sand and gravel infilling the former valley of the river, which lies to the south of Kirkton of Logie-Buchan (see cross-sections on Sheet 87W Solid-and-Drift). A small patch of red clay lying within a topographical depression in the vicinity of Littlemill of Esslemont [NJ 928 285] was formerly worked for the production of tiles and pipes (Chapter 2).

Ponding also took place in the valleys of the North Ugie and South Ugie waters, where up to about 25 m of interbedded silts, clays and very fine-grained sands are concealed beneath the glaciofluvial terraces, as at Denhead [NJ 998 521], and river floodplains. The deposits are commonly colour-banded, including reds, browns, yellowish browns, greenish greys, dark greys and blacks. Few deposits are exposed, but in stream sections near Ballus Bridge [NK 001 473], south of Mintlaw, 3 m of glaciofluvial gravel with an eastward palaeocurrent overlies over 2 m of yellowish brown sand, gravel and silt thinly interbedded with reddish brown and dark grey plastic clay. The multicolouring indicates that the two coastal ice streams and the East Grampian ice sheet were close by in the area. Furthermore, in a temporary section near the bridge [NK 003 471], the laminated sequence was brown, grey and black towards the base, becoming yellowish brown and vivid reddish brown towards the top, indicating that meltwater derived from ‘Logie-Buchan’ ice entered the lake relatively late (Appendix 1 Ugie valley). All the fine-grained sediments contained good assemblages of reworked Mesozoic palynomorphs, whatever their colour.

No glaciolacustrine deposits have been mapped on Sheet 87E, but they occur extensively beneath the ground mapped as ‘undivided’ red and blue-grey tills of the Logie-Buchan and Banffshire Coast drift groups, respectively. This ground is underlain by up to 25 m or more of thinly interstratified clayey diamictons, clays, silts and fine-grained sands that were deposited in lakes (possibly brackish) at the landward margins of the coastal ice streams as they began to break up during deglaciation (Merritt, 1981). Thick deposits of red, waxy clay lie inland of the Bay of Cruden where they were worked until recently at Errollston [NK 088 370] for making bricks (Appendix 1 Site 17).

Sheets 76E Inverurie And 77 Aberdeen

During deglaciation, as the East Grampian ice sheet retreated westwards across the area covered by these two sheets, several large proglacial lakes formed on the lower lying ground. Some of the present drainage routes were blocked by sediments and residual masses of ice, causing ponding locally, but it is also probable that the regional water table was elevated as a result of the relatively high sea level during deglaciation. Sediments in the vicinity of the Mill of Dyce (Appendix 1) caused ponding in the lower stretches of the Don valley where over 15 m of dark yellowish brown, olive-brown and grey sandy silts and clays underlie the floodplain. They form low-lying terraces near Wester Fintray [NJ 811 164], where a prominent mound of glaciofluvial sand and gravel stands in the centre of the valley. The mound formed as a delta and the sandy deposits there fine downwards into over 12 m of laminated micaceous silts.

Ponding also occurred upstream of Inverurie in the valley of the River Urie and its northern tributaries. Deposits of interlaminated fine-grained sand, silt and clay were laid down in the resulting lake, which stood at about 63 m above OD. Low-lying terraces underlain by fine-grained sands, silts and clays are also present on both sides of the River Urie farther upstream, in the vicinity of Old Rayne, where they lie at elevations of 95 m OD and above. They also form extensive deposits in the valley of the Lochter Burn. The silts and clays are commonly dark yellowish brown to grey, thinly laminated and micaceous. Moundy deposits within the main valley in the vicinity of Portstown [NJ 774 231] comprise of over 13 m of interbedded silts, clayey diamictons, sands and gravels and were probably laid down at the ice margin.

Brown, laminated silts and sands of glaciolacustrine origin are common around Kemnay, where they form low-lying terraces beside the River Don and lie within several topographical depressions. Ponding took place during deglaciation when the gorge of the Don remained blocked between Burnhervie and Port Elphinstone.

Few glaciolacustrine deposits are shown on the drift edition of Sheet 77 published in 1980, but they are commonly interdigitated with the glaciofluvial sand and gravel deposits stretching to the north-east of Belhelvie, which are now placed in the Logie-Buchan Drift Group. They also occur towards the base of many coarsening-upwards deltaic sequences within the moundy glaciofluvial deposits lying between Dyce and the coast. Isolated deposits of reddish brown, stiff, waxy clay and clayey silt occur at Tipperty [NJ 970 268] and Balmedie [NJ 965 166]. These sediments typically give rise to poorly drained, gleyed, brown, forest soils of the Tipperty/Carden soil association (Walker et al., 1982). Both deposits were possibly formed in lakes at the landward margin of ice retreating into the North Sea, but a glacio-estuarine origin is perhaps more likely (Simpson, 1955, p.195) (see below). The deposit at Tipperty was worked for making bricks and tiles (Chapter 2).

Sheets 66E Banchory and 67 Stonehaven

Glaciolacustrine deposits are quite widely distributed on these two sheets (Map 10); (Map 11). Those that were laid down in proglacial lakes at the margin of the East Grampian ice sheet, as it retreated north-westwards across the area, have been assigned to the Glen Dye Silts Formation of the East Grampian Drift Group (Chapter 8). Those formed at the margin of the Strathmore ice stream are assigned to the Ury Silts Formation of the Mearns Drift Group. A considerable thickness of waterlogged, thinly laminated sandy silt and clay is commonly present beneath the alluvium and peat in the elongate ice-scoured basins north of the River Dee (Chapter 7). These deposits, which are commonly grey or greyish brown in colour, have been proved to exceed 7.1 m in thickness in BGS Borehole NO79NW/16, sited on the lacustrine alluvium within the Loch of Park basin. They have been interpreted as ranging between 7.9 and 9.2 m in thickness at nearby resistivity sites.

Many of the moundy and flat-topped spreads of glaciofluvial sand in the Water of Feugh catchment fine downwards into laminated sandy silt and clay. Notable spreads crop out on the western side of the Water of Dye near Bogarn and on the flanks of a small tributary stream at Miller’s Bog [NO 637 861]. Up to 3.0 m of orange-brown and grey, laminated, stiff waxy clay was formerly exposed in the latter area. Similar deposits are present beneath flat-lying ground in the vicinity of Blairydryne [NO 749 926], in the valley of the Burn of Sheeoch, where they underlie deltaic glaciofluvial sands and gravels (Appendix 2). The deltaic sediments, which were formerly well exposed in Lochton Pit (Auton et al., 1988; Brown, 1994) prograded northwards into a glacial lake at an elevation of about 120 m OD (Brown, 1994). Glacial lakes may also have been present farther up the valley, at elevations of about 133 and 126 m OD, though fine-grained sediment has been recognised in association with them only in BGS boreholes and trial pits (NO79SW/6, NO79SW/8 and NO79SW/11).

The only notable outcrop of the Glen Dye Silts Formation on Sheet 67 occurs within a shallow ice-scoured hollow north of Rickarton [NO 816 891]. However, silty lacustrine deposits are probably concealed beneath peat and alluvium in several other ice-scoured basins in the northern part of the sheet. Fine sand with partings of laminated clay underlies glaciofluvial gravel at Bossholes [NO 812 884] indicating the presence of a small ephemeral proglacial lake.

Thinly laminated reddish brown fine sand, silt and clay of the Ury Silts Formation crop out extensively on the low-lying ground of Strathmore, where they also constitute a major part of the concealed succession recorded in boreholes and trial pits. Notable flat-lying spreads crop out east of Fettercairn, west of Fordoun and on the western side of Stonehaven (where they were formerly worked for making bricks and tiles). Smaller outcrops occur on the flanks of sand and gravel mounds north of Ury Home Farm and on the sides of the valley of the Carron Water. Red-brown silts and clays also occur between the ridges of sand and gravel that formed in contact with the southwestward retreating Strathmore ice stream, east of Auchenblae.

Within the valley of the Cowie Water concealed patches of red brown clay and silt extend up to at least 4.5 km inland from the coast. Moderate red, waxy clay, with partings of yellowish brown sand and silt occurs between 6.7 and 12.8 m depth in BGS Borehole NO88NW/12 at Nether Findlayston (Auton et al., 1988). The clay occurs between two units of sand and gravel. The underlying silty gravel, which contains pebbles of quartz and fine-grained, dark igneous (possibly andesitic) rock as well as granitic, psammitic and semipelitic clasts, rests on semipelitic bedrock. The composition of the gravel indicates that it was deposited by meltwater sourced from the Strathmore ice stream. The moderate red colour and waxy nature of the clay are typical of lacustrine sediments within the Mearns Drift Group. It is suggested that a proglacial lake formed, fed principally by meltwaters draining north-westwards (up valley) from Strathmore ice that was retreating towards the coast. The 6.7 m of sand and gravel overlying the red clay fines downwards and becomes more silty and cohesive with depth suggesting an origin as a delta or fan-delta. Pebbles within the sand and gravel are predominantly granitic (with minor amounts of semipelite) indicating that the deltaic sediments were laid down by meltwaters from the East Grampian ice sheet draining south-eastwards down the valley into the lake.

Interbeds of red-brown silt and silty sand, 1 to 2 m thick, with sparse dropstone pebbles, are present within deposits of the Drumlithie Sand and Gravel Formation between Auchenblae and Glenbervie. In these sequences, however, all of the sediments were laid down by meltwater issuing from the Strathmore ice stream as it decayed.

Periglacial deposits

Head

Head deposits are poorly sorted and poorly stratified sediments that have formed mainly as a result of the slow, viscous, downslope flow of waterlogged soils (solifluction and gelifluction), soil creep and hillwash. Solifluction was most active while periglacial conditions existed in the district during the latter part of the Main Late Devensian glaciation and the Loch Lomond Stadial (Galloway, 1958). It occurred during the summer months when the uppermost 0.5 to 1 m of the soil, the so-called ‘active layer’, thawed, while the ground below remained permanently frozen. The thickness and potential mobility of active layers depend very much upon the cohesiveness of the sediments affected, hence the thickest head deposits tend to occur where thoroughly decomposed rocks, clayey tills and fine-grained glaciolacustrine deposits have been remobilised.

Head deposits are ubiquitous across the district (Plate 12a), (Plate 12b), but generally it has not been practical to map them out. They are especially well developed across the Buchan plateau where decomposed rock materials have been affected by periglacial processes (Galloway, 1958; FitzPatrick, 1956, 1963, 1969, 1987). They commonly occur in the bottoms of glacial meltwater channels and valleys, locally concealing beds of gravel that were laid down during the formation of those features (Chapter 7). Head deposits redeposited from weathered granite bedrock typically consist of clayey coarse-grained sand, as, for example, around the Hill of Fare (Sheet 76E). Quartzites tend to produce rubbly deposits, as for example on Mormond Hill (Sheet 97), where angular fragments of quartzite are locally associated with sparse boulders of granite carried from the Strichen pluton to the west.

The outcrop of basic igneous rocks of the Insch intrusion on sheets 76E and 86E (Figure 2) is largely blanketed by head deposits composed of angular gravel derived from the underlying bedrock. Head has been mapped around the village of Old Rayne where it reaches several metres in thickness. A good exposure occurs on the southern side of a glacial drainage channel drained by The Shevock [NJ 664 286] (Map 8), where over 2 m of silty clast-supported angular gravel composed entirely of medium- and fine-grained basic rocks blankets fresh bedrock. Sandier gravel is seen to depths of over 1.5 m in several exposures in the village itself, and similar material mantles the norite bedrock near Oyne. Flatter spreads of clayey head, formed of remobilised till, have accumulated around the floodplains of The Shevock, near Insch, and the Gadie Burn around Buchanstone [NJ 656 261]; comparable head deposits flank lacustrine alluvium west of Myrebird [NO 742 990] on Sheet 66E (Map 10).

Head has been mapped relatively extensively on Sheet 96E Banff, where it applies to thick, weakly coherent masses of weathered bedrock or drift deposit prone to movement on waterlogged slopes, most commonly along steep cliff lines and along glacial melt-water channels.

Old, pre-Late Devensian head deposits have been located at several sites in the district where they provide an important part of the Pleistocene record (Chapter 8; (Table 7)). At least five periglacial episodes are represented at Kirkhill and Leys; other old gelifluctates and cryoturbates occur at Teindland, Crossbrae and the Moss of Cruden (Appendix 1).

Remobilised deposits of till

The remobilisation and solifluction of glacial diamictons following their deposition has been widespread in the lowlands of north-east Scotland (Galloway, 1958). However, it is only apparent at a few sites where solifluction deposits rest on organic sediments dated to the Windermere Interstadial. These sites include the Rothes road cutting (Chapter 2, Landslips) and Garral Hill, near Keith (Godwin and Willis, 1959), both to the south of the district. Others occur at Woodhead on Sheet 86E (Connell and Hall, 1987), Moss-side Farm, near Tarves on Sheet 87W (Clapperton and Sugden, 1977), sites near New Byth and at Glenbervie (Appendix 1). These sites demonstrate the former instability of low-angle till slopes during the Loch Lomond Stadial and the widespread occurrence of slope-foot accumulations of periglacial diamicton. The frequency of former detachment slides involving former ‘active layers’ in the region is unclear, but their identification is important because slopes such as these are liable to be rendered unstable by engineering works.

Frost-shattered rock

Frost-shattered rock is widespread across the district and is particularly well developed on quartzitic rocks such as around Durn Hill (Sheet 96W) and Mormond Hill (Sheet 97). It reaches some 8 m in thickness on Sillyearn Hill [NJ 507 514], near Keith, to the south of the district (Galloway, 1958). Phyllites cropping out around the headwaters of the River Ythan (Sheet 86E) and the Glens of Foudland [NJ 605 348], farther west, also are shattered to depths of up to 5 m locally (Galloway, 1958). Some frost shattering occurred prior to the Main Late Devensian glaciation as shattered rock underlies till locally, such as at Newbigging [NJ 527 591], on Sheet 96W (Galloway, 1958).

Flint-quartzite head

Flint-quartzite head deposits have been mapped out only on Sheet 87W Ellon, whereas they are portrayed as ‘float’ on Sheet 87E Peterhead. They are probably also developed on the other outcrops of the Buchan Gravels Formation on Sheet 86E Turriff (Chapter 4).

Some outstanding examples of head deposits have formed on the flint and quartzite gravels of the Buchan Ridge and Windyhills (Appendix 1 Sites 13, 14). The gravels are invariably concealed beneath up to 1.5 m of ‘cryoturbate’, formed by stirring, churning and other processes of cryoturbation during repeated freezing and thawing cycles in former periglacial conditions. Vertically orientated elongate clasts are common at the top of these deposits, which are typically pale yellow/orangebrown to white in colour and consist of gravelly, sandy kaolinitic clays. The deposits are particularly susceptible to frost heaving and a road across the Moss of Cruden was rendered impassable as a result of this process one particularly cold winter in the late 1970s. Extensive iron pan is commonly developed at the base of the cryoturbated layer, causing poor drainage and the development of a peat cover at the surface.

A unit of moderate brown diamicton up to 1 m thick caps the cryoturbated Windy Hills Gravel Member at its type section near Fyvie (Map 5). The diamicton includes sparse striated clasts of fresh Dalradian metamorphic rocks as well as quartz and quartzite pebbles from the gravel below. The unit is intensely cryoturbated with excellent examples of erect clasts (FitzPatrick, 1987, fig. 13.5; (Plate 4b). The unit is now interpreted to be a till that has subsequently experienced periglacial churning. Some of the flinty head deposits capping the Buchan Ridge probably have a similar origin.

Alluvial and aeolian deposits

The alluvial deposits in the district are of five types: fluvial deposits underlying the floodplains and low-lying terraces of rivers, sediments within enclosed basins, lacustrine alluvium, alluvial fans, and river terrace deposits.

Alluvium of floodplains and low-lying river terraces

Ribbons of alluvium flank the courses of the major rivers in the district, forming low-lying ground potentially liable to flooding. River bank sections commonly reveal clast-supported gravel (shingle) capped by ‘overbank deposits’ consisting of one or two metres of laminated, humic, micaceous, silty sand (loam), locally intercalated with peat. In general, the clasts forming the gravel are subangular to well rounded, reasonably well sorted and exhibit a pronounced imbrication dipping upstream. Only the faster flowing rivers in the district, such as the Spey, Dee, Don and Deveron, have extensive gravelly beds. The River Spey is particularly fast flowing and is braided downstream of Fochabers, where large linear bars of cobble shingle form where river channels bifurcate. Abandoned braid-bars forming the floodplain are generally afforested or covered by rough, boulder-strewn scrubland (Lewin and Weir, 1977). Most other rivers are single thread and have cultivated floodplains. Meander belts are common only along stretches of the Lossie, Ugie, Urie, Don and Feugh (Appendix 1 Nether Daugh).

Thicknesses of alluvium are difficult to generalise as they are very variable, both within, and between catchments. Downstream of Fochabers, the coarse, gravelly alluvium of the Spey is on average about 6 m thick, locally ranging up to 12 m or more. Near Banff, the floodplain of the Deveron is underlain by 2 to 3 m of silt overlying 7 to 15 m of alluvial deposits including sand and gravel. The floodplain of the Ythan is underlain by 4 to 6 m of gravel downstream of Methlick, whereas thicknesses of between 7 and 10 m of alluvial gravel are common in the Dee valley. Alluvial gravels of the Don typically range between 3 and 5 m in thickness downstream of Inverurie, whereas those in the catchment of the Ugie typically range between 1 and 5 m. In the lower reaches of the Ugie, Ythan, Urie and Don, the alluvial gravels commonly overlie thick sequences of fine-grained glaciolacustrine deposits. They commonly infill buried valleys (Chapter 7).

In most valleys, alluvial gravels would have accumulated either in cold climates following deglaciation and during the Loch Lomond Stadial, or, at the beginning of the Windermere Interstadial and Holocene, when little vegetation was present to stabilise soils in the river catchments. Apart from the Spey and Dee, there can be few stretches of river in the district where gravels are currently accreting as opposed to being reworked from older deposits a short distance upstream. The floodplains of larger rivers are generally bounded by bluffs, which in the lower Spey valley are typically steep banks up to 10 m high. In minor valleys, however, and in valleys occupied by ‘misfit’ streams like that of the Idoch Water, south-east of Turriff (Map 5), and the Gadie Burn near Insch (Map 8), the former floodplains are commonly poorly defined because the bluffs have been degraded by periglacial processes since their formation. In these circumstances, clayey solifluction deposits (head) commonly locally overlie alluvial sediments along the edge of the floodplains. In many small valleys, especially in central Buchan, there are no floodplains as such and little, if any, alluvium has been mapped. The floors of these valleys are underlain by up to 3 m or so of head deposits consisting of interbedded clayey sand and gravel and gravelly diamicton, locally overlying older river gravels.

Alluvium of basins

Alluvium has been mapped in numerous small, poorly drained, enclosed, or semi-enclosed basins across the district. Some of the basins are kettleholes, but the majority have been gouged out of rock or have been sub-glacially sculpted in till (Chapter 7). Basins are particularly common on Sheets 76E Inverurie and Sheet 77 Aberdeen, where they are typically lined by up to 3 m or so of thinly bedded, silty, medium- to coarse-grained, micaceous, quartzofeldspathic sand. Most depressions, however, are underlain by a few metres of head deposits like those described in the minor valleys of the district above. Peat formerly occupied most of these depressions before being cut by man and drained artificially.

Lacustrine alluvium

Flat-lying spreads of interbedded humic sand, silt and clay lie within some poorly drained, enclosed basins that contained lochans before being drained artificially. In many instances it is difficult, however, to distinguish between lacustrine alluvium and the alluvium of enclosed basins described above, and the choice of category used varies somewhat from sheet to sheet. This is partly because many of the drainage schemes in the district were carried out following the Crimean War, in the 1860s, and it is no longer clear where the lochans once stood. Other spreads of lacustrine alluvium border existing bodies of fresh water, as, for example, around the Loch of Skene (Map 8). In general mapped spreads of lacustrine alluvium are finer grained than the alluvial deposits filling small basins.

Much of the lacustrine alluvium began to accumulate during Late-glacial times, as for example, that in the Loch of Park, on Sheet 66E Banchory. Organic sediments at this site (Appendix 1) preserve records of vegetational change that has occurred in the district since deglaciation.

Alluvial fan deposits

Fans composed of sand, gravel and gravelly diamicton have accumulated since deglaciation, where tributary streams with relatively steep gradients debouch into more major river valleys. Small, but notable examples occur on Sheet 67 Stonehaven north of Mill of Barras [NO 850 794] and near Greenden [NO 811 776].

River terrace deposits

Dissected remnants of former floodplains flank the alluvium of many of the larger rivers in the district. These form terraces that slope gently down-valley and are typically composed of several metres of stratified clastsupported gravel or sandy gravel. Terrace aggradation probably occurred mainly during Late-glacial and early Flandrian times when the gravel would have been deposited mainly as bars and channel infills by braided streams. The gravel is commonly overlain by spreads of sand and silt up to about 2 m thick, laid down as overbank deposits during the waning stages of periodic flood events. These fine-grained deposits are widespread on the broad terraces of the major rivers where they produce well drained, light sandy soils.

Most river terraces in the district are judged to have formed during ice-sheet deglaciation and consequently have been mapped as glaciofluvial sheet deposits. These terraces are kettled locally and commonly merge into spreads of moundy ice-contact glaciofluvial deposits. Others clearly formed as kame terraces banked up against ice stranded within valleys.

Particularly good examples of low-lying river terraces flank the floodplain of the Water of Feugh, upstream of Strachan (Map 10). They generally rise no more than 3 m above the level of the floodplain and are composed entirely of interbedded sand and pebble gravel, capped by thin spreads of silt and clay. As many as three distinct terraces can be recognised, none standing greater than 1.5 m above the level of its neighbour and they preserve evidence of shallow palaeo-channels on their surfaces. These true river terraces lie adjacent to kettled glaciofluvial terraces, kames and esker ridges that may, in places, stand up to 25 m above the level of the floodplain.

Blown sand

Deposits of wind-blown sand occur in many coastal localities. They are most commonly found next to sandy beaches, from where most of the sand has blown, but sandy glaciofluvial deposits have been a source locally. Blown sand is generally well sorted and fine to medium grained. Its composition depends somewhat on the provenance of the sediment source occurring locally, but is generally quartzose with varying proportions of finely divided shell. Areas mapped as blown sand generally include stabilised dunes, links and cultivated aprons as well as active dunes. Long stabilised deposits have been mapped locally as ‘older blown sand’. A detailed, systematic account of the coastal dune systems of north-east Scotland has been provided by Ritchie et al. (1978).

On Sheet 95 Elgin (Map 1), the largest spreads of blown sand face Burghead Bay where dune ridges tend to be orientated south-west–north-east, parallel with the prevailing wind, and reach up to about 15 m high. Within historical times, the migration of sand along the north and west sides of the Loch of Spynie materially affected local drainage conditions and accounted for the disappearance of several lochs in the area (Peacock et al., 1968). Most of the blown sand on the sheet is probably redistributed beach sand, carried westwards by longshore drift and swept north-eastwards by the prevailing winds, but glacial sands must also have provided a significant source.

Few deposits of blown sand have been mapped on Sheet 96W Portsoy and Sheet 96E Banff, although some of the sandy glaciofluvial deposits occurring along the coast probably have been considerably redistributed by wind in the past when arctic, arid conditions prevailed. Blown sand caps the cliffs at Whyntie Head [NJ 628 661] (Map 2), where the action of blowing sand has beautifully etched rocks in the vicinity (Read, 1923). Modern dunes fringe the bays of Cullen, Sandend and Boyndie, where they rest on deposits forming Flandrian raised beaches.

A continuous, generally 250 to 600 m-wide belt of blown sand backs the coastline for a distance of about 25 km between Fraserburgh and Peterhead on Sheet 97 Fraser-burgh (Map 4) and Sheet 87E Peterhead (Map 7), where dunes reach 30 m in height locally (Peacock, 1983). The sand is generally quite shelly and was formerly used as a source of lime at St Fergus. The dunes are particularly active between Inzie Head and Rattray Head, where the combination of blowing sand and southward longshore drift finally blocked off a tidal inlet to create the Loch of Strathbeg in 1720 (Wilson, 1886). Less extensive deposits face the bays at Rosehearty and Phingask (east of Sand-haven) on the northern coast (Wilson, 1882), and the bays of Peterhead and Sandford on the east coast. The blown sand fringing the Bay of Cruden is relatively more siliceous and locally contains grains of magnetite.

An extensive area of sand dunes lies between Collieston and the Ythan estuary on the boundary of Sheet 87E Peter-head and Sheet 77 Aberdeen. It is mostly designated as the Sands of Forvie Nature Reserve and in contrast to the other deposits described above, the sands there have spread over 2 km inland and reach up to 57 m above OD. Dunes are commonly 12 m or more in height and are particularly active towards the coast and across the peninsula at the southern end of the reserve. The dunes continue to the south of the Ythan estuary, where they form a belt up to 500 m wide that extends 14 km to the estuary of the River Don (Map 9). The dunes reach 15 m in height and back a sandy beach. They are relatively active with blowouts mostly trending from north-west to south-east. Northward expanding deflation plains are common, where a veneer of sand conceals raised Flandrian beach deposits. Sand is banked up against, and largely obscures, the Main Postglacial Cliffline backing those deposits. It also extends inland to about 1 km from the coast, where it mantles till and glaciofluvial deposits; ‘fossil’ sand dunes occur in the vicinity of the coastal look-out at Menie Links [NJ 986 210].

The rugged coastline to the south of Aberdeen is largely devoid of mappable deposits of blown sand. This is mainly due to the high cliff line, but also because the beaches are mainly formed of shingle.

Organic deposits

Peat

Deposits of basin peat in the district occur mostly within the sites of former lochans. Commonly, the peat contains tree boles and other woody fragments as well as the partially decomposed, acidic remains of sedges, reeds, rushes, Sphagnum and heather. Most raised peat mosses are a fraction of the original size because of their exploitation for fuel in historical times. Today, many deposits barely average even one metre in thickness, but they remain waterlogged and would partially regenerate in time. Many peaty hollows have been partly, if not completely infilled with boulders carted off the fields. Others have provided sites for dumping waste materials. Peat resources are discussed in Chapter 2 and the locality and extent of the main deposits are listed in (Table 5).

Many basins are ice-scoured hollows in bedrock, such as Harestone Moss [NJ 932 195], which is underlain by ultra-basic igneous rocks north-east of Belhelvie (Map 9), and those formed on granite around New Pitsligo and Strichen (Map 6), which contain significant resources of peat. Many small basins are kettleholes in glaciofluvial sands and gravels or morainic deposits, such as large parts of Red Moss [NJ 747 014] and Leuchar Moss [NJ 788 047], on Sheet 76E Inverurie. The majority of basins, however, are ice-moulded hollows in till; abundant examples occur on Sheet 76E, for example Skene Moss [NJ 757 107] and Braigies Moss [NJ 758 047]. They are also common on Sheet 87W Ellon (for example the Moss of Belnagoak [NJ 880 428] and Elrick Moss [NJ 954 418]), on Sheet 86E Turriff (for example between Cuminestown and the Ythan gorge) and on Sheet 77 Aberdeen (for example Burreldale Moss [NJ 829 239] and several mosses in the vicinity of Westfield [NJ 946 207]). Deposits of peat rest on raised marine and glaciomarine clays within hollows between Kinloss and Lossiemouth on Sheet 95 Elgin.

Some of the most extensive spreads of low-lying blanket peat in the district rest on particularly clayey and impermeable deposits of till, such as those of the Banffshire Coast Drift Group beneath Rora Moss [NK 045 515] and St Fergus Moss [NK 055 536] (Map 7), and several mosses around New Pitsligo (Map 6). Peat has developed on stiff red till of the Logie-Buchan Drift Group at Lochlundie Moss, 5 km south-west of Cruden Bay on Sheet 87E Peterhead.

Hill peat is most extensive on the granite outcrops in the south of the district, where tree roots and stumps, particularly of Scots pine, are very common. Blanket peat covers parts of Bennachie and the Hill of Fare on Sheet 76E Inverurie and on hilly areas underlain by the Mount Battock granite on Sheet 66E Banchory, where it is relatively thick and widespread. For example, up to 4 m of peat overlying decomposed granite is exposed beside a forestry track on the south-west side of Kerloch [NO 697 879]. Other spreads include those on the flaggy quartzites forming the Hill of Stonyslacks, 8 km south of Buckie on Sheet 95 Elgin, and the Moss of Fishrie, on Old Red Sandstone sandstones and shales between Gardenstown and New Byth, on Sheet 96E Banff. Small spreads occur on Bracklamore and Windyheads hills to the south of Pennan, on Sheet 97 Fraserburgh, which are formed of Old Red Sandstone breccias and conglomerates.

Blanket hill peat is also widespread on the relatively high ground of central Buchan formed by the ‘Buchan Ridge’ and the Hill of Longhaven, which lie to the south-west of Peterhead (Map 7). This ground is either underlain by the clayey, kaolinitic deposits of the Buchan Gravels Formation, or on bedrock that has commonly decomposed to sandy kaolinitic clay (Chapter 3).

Woody peat deposits probably developed extensively around the coasts of north-east Scotland in the early to mid-Holocene, prior to the ‘Main Postglacial’ marine transgression, when most were destroyed by coastal erosion. However, some deposits have survived in sheltered situations, such as beneath the alluvium of the Water of Philorth, 3 km southeast of Fraserburgh ((Map 4); Appendix 1). Remnants of this so-called Boreal forest are exposed on the foreshore in Burghead Bay at low tide (Map 1). Remnants of older peat deposits formed during the Windermere Interstadial are more sparse ((Table 7); (Table 8); Appendix 1 Rothens, Mill of Dyce, Loch of Park, Glenbervie, Howe of Byth). Deposits of early Devensian age or older are even rarer (Appendix 1 Crossbrae, Moss of Cruden, Burn of Benholm).

Shell marl

Thin, discontinuous spreads of freshwater shell marl interbedded with basin peat and fluviatile silt overlie raised marine deposits around Loch Spynie, 4 km northnorth-east of Elgin (Map 1). A 35 cm-thick bed of shell marl resting on Late-glacial marine clay was formerly visible in a brick pit situated in another alluvial depression at Gilston [NJ 206 662], nearby.

Raised marine deposits

In addition to present-day coastal sediments, two distinct sets of raised marine deposits are present in the district. The sets were formed during periods of relatively high sea level, in Late-glacial times and in the mid-Holocene (Chapter 7; (Figure 48)). Sea level was appreciably lower than it is today between these periods. Each set of deposits is capable of being divided lithologically into ‘shoreface and beach deposits’ (mainly shingle and sand) and quiet-water sedimentary facies formed in tidal-flat, brackish lagoon and estuarine environments (mainly fine-grained sand and silt). Only the raised glaciomarine deposits are described here; the raised beaches and associated deposits, rock platforms and other features of coastal erosion are described in Chapter 7.

Late Devensian raised marine and glaciomarine deposits

Raised marine beds, mainly silty clay, are known between Elgin and Lossiemouth, from north of Peterhead, and possibly between Ellon and Stonehaven. In the Elgin area, the Spynie Clay Formation underlies much of the low-lying ground around Loch Spynie [NJ 237 667] (Chapter 8; (Map 1)). It occurs also below Lossiemouth Airfield where it is overlain by sand and gravel forming Late-glacial and Flandrian raised beaches. The maximum recorded thickness is 12.5 m (Peacock et al., 1968). North of the Spynie basin the formation attains a minimum level of 10 m above OD, and possibly reaches over 20 m OD, but on the south side of the basin the mapped level is only a little above present sea level, probably because the sea was partly excluded by stagnant ice from this area when sedimentation had begun elsewhere.

North of Peterhead, on Sheets 87E and 97, the dark grey silts, clays and fine-grained sands of the St Fergus Silt Formation extend up to about 16 m above OD (Chapter 8). Most of the formation is concealed by lacustrine alluvium (peat and silt) or blown sand. It borders red diamicton, clay and sand of the Logie-Buchan Drift Group to the west, whereas seawards, it appears to be banked against, or to have been deformed into, an end moraine (Hall and Jarvis, 1989; Anderson in Scott, 1890; Glentworth and Muir, 1963; (Map 4)). Their lithology suggests that the St Fergus Silts are glaciomarine in origin. Most of the marine bivalves, foraminiferids and ostracods in it belong to boreal or boreal-arctic taxa, and two adjusted radiocarbon dates of about 14.9 and about 14.3 ka BP indicate that it was laid down prior to the Windermere Interstadial (Appendix 1 St Fergus).

Some of the isolated deposits of red clay that have been mapped as glaciolacustrine deposits along the coast between Peterhead and Stonehaven may also be partly glaciomarine. These deposits, the Tullos Clay Member of the Logie-Buchan Drift Group (Chapter 8), contain a macrofauna and microfauna that is predominantly derived from reworked older deposits (Jamieson, 1882b; Bremner, 1916; Simpson, 1948; Munro, 1986). The clays lie between present sea level and about 30 m OD and may correlate with the Errol Clay Formation of eastern Scotland (Peacock, 1999). Early workers recorded the skull of a seal at Westfield [NJ 993 308], near Kirkton of Logie-Buchan on Sheet 87W (Nicol, 1860; Jamieson, 1882b), a fish at Tipperty [NJ 971 268] (Map 9), and remains of Ophiura from the clay at Clayhills [NJ 941 055], in Aberdeen. The red clays on the south side of the city of Aberdeen reach over 18 m above OD at Tullos (Simpson, 1948), where they were worked for making bricks [NJ 949 052] (Chapter 2).

Present-day marine deposits

Shoreface and beach deposits

The physical characteristics of beaches vary considerably over short distances and only broad generalisations can be offered here. The following notes are based on a detailed, systematic account of the beaches of north-east Scotland by Ritchie et al. (1978). The composition of beach deposits is intimately associated with the configuration of the coastline and with available sources of material. Broadly speaking, there are three types of coastline in the district.

1. Long sandy bays of gentle arc backed by extensive sand dunes Typically, these occur between Aberdeen and Fraserburgh, where the beaches have generally accreted outwards by processes of shoreline ‘regularisation’ and northerly longshore drift. The beaches are relatively unstable in that there are marked seasonal changes in profile and in the position of the high tide mark. This instability is also manifested in the presence of dynamic bars and troughs in the intertidal beach zone that cause waves to break farther from the coast than is usual and tend to generate strong rip currents.
2. Long bays of gentle arc like those above, but formed mostly of shingle Beaches of this type back Spey Bay and Burghead Bay on Sheet 95 Elgin. Fluvial inputs have been relatively important in the evolution of this type of beach, together with the dominant westerly longshore drift. The beaches are relatively unstable and are associated with rapidly developing spits and offshore bars.
3. Isolated cliff-foot and bayhead beaches along rugged, indented rocky coastlines These predominantly shingly beaches occur along the northern coast between Buckie and Rose-hearty and less commonly along the eastern coast between Boddam and the mouth of the Ythan (with the exception of the Bay of Cruden) and along the coast south of Aberdeen. The beaches are relatively stable because they occur in sheltered coves and inlets, and because they normally lie on rock platforms. They are rarely backed by sand dunes of any great extent. The size and orientation of the beaches along the northern coast reflect the strike of the rocks and the dominant input of wave energy from the north-east. The largest beaches are superimposed on remnants of raised Flandrian beach deposits.

Saltmarsh deposits

Several small saltmarshes occur within the estuary of the River Ythan downstream of Kirkton of Logie-Buchan on Sheet 87W Ellon. The features are underlain by tenacious silty clay penetrated by decaying rootlets. The only other saltmarsh that has been mapped in the district lies to the west of Kinloss on Sheet 95 Elgin.

Sea-bed sediments

Sea-bed sediments are depicted on three BGS 1:250 000 scale maps of the continental shelf listed in Information sources. The sediments lying off the northern coast of the district are described in the BGS offshore regional report covering the Moray Firth (Andrews et al., 1990), whereas those occurring off the eastern coast are given in another report in the series covering the central North Sea (Gatliff et al., 1994).

Chapter 7 Geomorphological features

A set of generalised maps showing the distribution of drift groups and a selection of deposits, landforms and localities mentioned in the text is provided (Maps 1–11).

Features of glacial erosion

Glacial striae, roches moutonnées and glaciated rock knolls

Evidence of glacial erosion, in the form of scratched and abraded bedrock, indicates the former direction of ice movement. Striated surfaces are not common in northeast Scotland because there are few outcrops of fresh bedrock. Those that do occur are generally restricted to the coasts where glacial erosion has been concentrated (Chapter 3). Most striae are assumed to have been produced by the last major glaciation, namely the Main Late Devensian, but this is not always the case. Some have survived from earlier glacial phases, protected beneath a mantle of till, as, for example, beneath the Craig of Boyne Till Formation at the Boyne Limestone Quarry site (Appendix 1). Furthermore, striae probably only represent particular phases of glaciations when the ice was wet-based and able to glide over rock outcrops. Most of the striae depicted on the geological maps of the district were recorded during the primary survey and few are now visible, but those that are, together with several discovered more recently, generally confirm the earlier observations.

On Sheet 95 Elgin (Map 1), there are two distinct sets of striae trending east-south-east and south-south-east respectively, with a few directed in between these points. The trend of the former set accords with the general direction of ice movement deduced from the distribution of glacial erratics in the area (Figure 45). The direction of travel indicated by the latter set is less clear, but they were possibly formed by ice moving inland from the Moray Firth during the final phase of glacial activity in the area, the ‘Elgin Oscillation’ (Peacock et al., 1968). Glaciated pavements are well preserved on siliceous sandstones between Carden Hill [NJ 142 622] and Quarry Wood [NJ 190 640], where plucked surfaces generally confirm the east-south-east direction indicated by the striae recorded thereabouts.

South-east- to south-south-east-trending striae are dominant on Sheets 96W Portsoy and Sheet 96E Banff, but some striae are directed towards the north-east. There is also a record of east-south-east-orientated striae crossed by northward-directed striae in the south-west corner of Sheet 96E (Read, 1923; (Map 3)). Read (1923) linked the south-east- to south-south-east-directed striae with the glaciation that brought shelly tills and Jurassic erratics from the Moray Firth, and the north to northeast set with a major ice movement towards the coast (see also Bremner, 1934). The former agrees with the views expressed here (Chapter 5), but the latter may be related to more limited movements of ‘inland’ ice at a late stage of the Main Late Devensian glaciation, following retreat of the Moray Firth ice stream towards the coast (Peacock and Merritt, 1997, 2000).

Striae are relatively common along the foreshore between Fraserburgh and Rattray Head (Map 4). Roches moutonnées, grooves and striations are also found immediately inland where there is a gently undulating, glaciated surface developed on gneiss, granite and metabasic igneous rock. Most evidence indicates ice movement towards the east-south-east (Wilson, 1886; Milne, 1892b), but a later set, found on shore sections to the north of a line from the mouth of the Burn of Philorth [NK 028 650] to Inzie Head, suggest a later south-southeast movement (Peacock, 1997, 2000, fig. 27). Inland from these glaciated surfaces there is a transition, over a distance of a few kilometres, from rock that is little weathered at the surface to a wide area where the rocks are generally deeply decomposed with only sparse outcrops of fresh rock. Mormond Hill is one such outcrop, where striae are preserved on ice-smoothed, brecciated quartzite at the entrance to a quarry [NJ 950 568] at about 137 m above OD. Striae have also been observed on a smooth quartzite surface on the north side of the hill [NJ 982 581] at about 155 m OD. When ground as high as this was affected by east-south-eastward-moving ice, it strongly suggests that most, if not all of Buchan is likely to have been glaciated in the Late Devensian. However, large erratic boulders of Strichen granite are buried beneath thick gelifluctate on the northern and western slopes of the hill e.g. [NJ 9585 5690], suggesting that a prolonged period of periglacial conditions followed glaciation.

Striae are rare on Sheet 86E Turriff and Sheet 87W Ellon. Those that do occur are mainly orientated towards the south-east or south-south-east, and were probably associated with the deposition of overlying blue-grey tills of the Banffshire Coast Drift Group relatively early in the Main Late Devensian glaciation, if not earlier ((Figure 4); Wilson, 1886; Read, 1923). However, some north- to north-north-east-trending striae located to the north-east of Huntly, on Sheets 86W and 96E, are associated with a later ice movement, as at two localities, they cross, and are therefore younger than, striae directed toward the east (Read, 1923, fig. 11). Several sets of striae and roches moutonnées in the south of Sheet 87W relate to a later eastward movement of East Grampian ice.

On Sheet 87E Peterhead, striae are mainly restricted to the coast where several north-north-east-orientated markings were reported by Wilson (1882) in former granite quarries around Stirling Hill and near Yoags’ Haven [NK 116 394], some 2 km to the south. These striae were possibly created by East Grampian ice that flowed north-east towards the North Sea prior to the incursion of the coastal ice responsible for laying down the Logie-Buchan Drift Group.

Striae are sparse in the glaciated upland on the western side of Sheet 76E Inverurie. They have been recorded to the north of Torphins and on the northern side of Hill of Fare, where they indicate eastward flow of the East Grampian ice sheet. Glaciated streamlined rock knolls are more numerous, particularly on the lower ground and suggest slightly divergent ice-flow directions (generally towards the north-east and south-east). This divergent flow probably occurred at a relatively late stage when ice movement was influenced by the topography. Such flow directions accord well with retreat positions of the ice front indicated by moraines in cols north-east of the Hill of Fare (see below).

Striae and roches moutonnées are relatively abundant on Sheet 77 Aberdeen reflecting the cumulative amount of glacial erosion throughout the Quaternary (Chapter 3). Inland, striae are mostly orientated towards the east, but towards the coast they swing towards the north-east, especially to the south of the Don (Bremner, 1938; Munro, 1986, fig. 31). Jamieson (1882b) noted that striae directed towards the north-north-east were superimposed on ones orientated towards the east-north-east in a former quarry south-east of Cove Bay and striae of roughly similar orientations are reported to the north of Banchory-Devenick [NJ 915 005] (Munro, 1986). Several swarms of ice-smoothed, elongated, streamlined knolls (rock drumlins) occur on the sheet, notably around Newmacher, Peterculter, to the north-west of Belhelvie, and in the vicinity of Ythan Lodge [NJ 995 270] (Map 9). Most of the features on high ground to the north of the Don are orientated towards the east, but those on the flanks of the Potterton Burn drainage channel are orientated south-east. These orientations accord with that of nearby rock knolls suggesting, as on Sheet 76E, that early ice-flow was towards the east, but as the ice thinned and the higher ground became ice-free, flow directions were influenced by the underlying topography. To the south of Nigg Bay, at Doonies Hill [NJ 960 030], there are several northward-directed roches moutonnées, reinforcing the evidence from striae that ice moved in that direction along the coast.

Few striae have been recorded on Sheet 66E Banchory. To the north of the Dee and the Feugh they show a consistent north-east alignment, but north-north-east-orientated striae have been recorded south of Kirkton of Durris (Map 10). Both sets indicate that the flow of the East Grampian ice sheet was directed parallel to the alignment of the Dee valley. On the watershed between the catchment of the Dee and the Bervie Water, striae trending east-south-east are present on a glaciated surface of granite exposed in a newly constructed forestry track, north-east of the Wild Mare’s Loup drainage channel. In Strathmore, a single roche moutonnée, on the northern side of the Glen of Drumtochty (Map 10), indicates south-east directed ice movement. Crossing south-east- and eastsouth-east-orientated striae have been recorded on an exposure of andesite north-east of Knockbank Farm [NO 746 798]. These observations suggest that the East Grampian ice sheet impinged onto the north-western flank of Strathmore, before the north-east flowing Strathmore ice stream became established (Figure 4).

Striae are much more numerous on Sheet 67 Stone-haven. Those occurring to the south of the Carron Water are consistent with ice moving north-eastwards along Strathmore. It is apparent that the northern margin of the Strathmore ice stream passed offshore just to the north of Muchalls, from where it ‘hugged’ the coastline closely towards Aberdeen. It encroached onshore near Portlethen, where distinctive red-brown deposits of the Mearns Drift Group were laid down.

Most of the striae and glaciated rock knolls developed on the Dalradian outcrop in the northern part of the sheet imply ice movement towards the south-east and accord with the alignment of the majority of those recorded by Bremner (1920a). However, his three records to the north of Muchalls of north-north-east-orientated striae crossed by a later south-east-orientated set, have not been confirmed during this survey. Although the trend of the north-northeast set of striae is coincident with the dominant foliation in the Dalradian rocks, the precision of his quoted measurements leave little reason to doubt Bremner’s original observations. Based on his records of crossing striae, distribution of erratics and the outcrop pattern of glacial deposits, Bremner (1920a) suggested that at some point Strathmore ice moved several kilometres inland onto the upland areas to the north and west of Stonehaven, probably reaching elevations of at least 330 m above OD (Figure 4).

Drumlins, drumlinoid ridges and large-scale glacial gouges

Drumlins are one of the most distinctive features of glacial moulding, yet they are relatively rare in the district. A drumlin is typically a smooth, oval-shaped hillock of glacial drift with a steeper, blunter end pointing up-glacier and a gentler sloping, pointed end in the former down-glacier direction. Most glaciologists agree that drumlins are bedforms streamlined in the direction of ice movement and produced mainly as a result of subglacial deformation of soft sediments (Benn and Evans, 1998). Apart from a few elongate till ridges of uncertain origin occurring in the Elgin district (Peacock et al., 1968; (Map 1)), drumlins appear to be restricted in distribution to the east coast. Even there, however, the features are probably best described as ‘drumlinoid ridges’ because few are perfectly formed.

The general absence of drumlins probably results from the relatively sluggish flow of ice across most of the district, rather than to any lack of deformable substrate, although drift deposits are relatively thin. The East Grampian ice sheet may also have been ‘cold-based’ during most of its existence, in which case little subglacial deformation is likely to have occurred (Benn and Evans, 1998). However, there are local exceptions, as for example to the south-east of Lurg Hill [NJ 506 575] on Sheet 86W Huntly, where there is a swarm of drumlinoid ridges and peat-filled gouges glacial sculpting of Old Red Sandstone bedrock is apparent o the south-western flank of the Hill of Finden [NJ 803 637] (Plate 13), south of Gardenstown (Map 3).

A swarm of drumlinoid ridges and intervening, anastomosing gouges extends from a kilometre or so west of a line from Strichen to Maud, on Sheet 87W Ellon, eastwards towards the coast (Map 6); (Map 7). The ridges typically have a relief of 10 to 25 m, and are formed mostly of deeply decomposed bedrock and till. The trend of the moulding suggests a direction of ice-movement between east and east-north-east, which is slightly at variance with east-south-east- directed striae recorded nearby on Mormond Hill (see above) and two sets of north-east-orientated striae on Sheet 87E Peterhead. The moulding affects, and therefore postdates, most of the deposits at the important glacial/interglacial succession at Kirkhill (Appendix 1), including the blue-grey Corse Diamicton Formation. The patchy uppermost till at Kirkhill, the Hythie Till Formation, has a fabric indicative of ice flow towards the east-north-east, suggesting that it was laid down during the streamlining event. The blue-grey till and rafts of the Banffshire Coastal Drift Group occurring at Oldmill (Appendix 1) has been streamlined similarly. There is little or no evidence of glacial streamlining on the ‘Buchan Ridge’ and across central Buchan, but there are distinct east-south-east to south-east-directed gouges in the vicinity of Ellon, depicted on Sheet 87W (Map 6). These features might have been produced by the ice that pushed inland from the North Sea basin, laying down the deposits of the Logie-Buchan Drift Group, but it seems unlikely that the margins of this ice mass would have had the power to create such features.

Glacial streamlining is evident across large parts of Sheet 76E Inverurie and Sheet 77 Aberdeen, and it generally becomes more pronounced eastwards, especially to the east of a line between Oldmeldrum and Inverurie. Many of the ridges have rock cores, particularly at their stoss (up-glacier) ends and only a thin covering of till. Around Newmachar (Map 9), for example, there is a gradation between small south-east-trending rock-cored drumlinoid till ridges and ice-moulded rock knolls, which is difficult to delineate by surface mapping alone. Limited drilling and trial pitting in the area (Auton and Crofts, 1986) has shown that till on the crests of ridges can vary in thickness between 0.6 and 2.6 m. The orientation of the drumlinoid features, particularly those on higher ground, generally indicates an eastward flow of ice towards the North Sea. West–east-orientated drumlins are well developed between Belhelvie and Udny Station where several examples of crescent-shaped depressions filled with peat occur at the stoss end of the features (Map 9). In the southern part of Sheet 77, the orientation of the features swings around to the eastnorth-east, but again indicates seawards flow of the East Grampian ice sheet. The south-east trend of ridges on the lower ground around Hatton of Fintray, Newmachar and south-west of the Potterton Burn may be preserving evidence of an early south-east directed ice movement (Bremner, 1928). Alternatively, the features may have formed late in the glaciation when flow directions of the thinning East Grampian ice sheet became more influenced by the local topography.

On Sheet 66E and Sheet 67E, drumlinoid ridges are well developed only within the outcrop of the East Grampian Drift Group. Their orientation is compatible with the ice flow directions that can be inferred from nearby striae and glaciated rock knolls (see above).

Ice-scoured depressions

These typically oval-shaped hollows scoured out of bedrock have a broadly similar distribution to the drumlinoid ridges described above and tend to occur in the areas where there has been most cumulative glacial erosion (Chapter 3). They are commonly peat filled because of poor drainage, and are especially common in the area underlain by the Strichen Granite on the margin of Sheet 87W Ellon and Sheet 97 Fraserburgh. Smaller, more irregularly shaped features are common to the west of Maud (Map 6) where highly weathered gabbro and norite crop out. Harestone Moss occupies an ice-scoured hollow on fresher ultrabasic rocks of the Belhelvie Pluton on Sheet 77 Aberdeen. Numerous alluvium-filled depressions lying to the east of Bennachie and the Hill of Fare on Sheet 76E were scoured out by eastward flowing ice, the largest one being occupied by the Loch of Skene (Map 8).

Large elongate elliptical hollows have also been eroded by eastward flowing ice north of the River Don on Sheet 76E. They are linked by west–east-trending drainage channels, and are filled with fine-grained alluvial sediments and peat. Marshes occupy the largest of the hollows, where former lochs, such as Loch of Park and Loch of Leys [NO 702 978] have been drained. South-east-flowing ice has also scoured large irregular hollows in resistant Dalradian bedrock on Sheet 67, north of Stonehaven. Many hollows are occupied by peat mosses, the largest of which is Red Moss [NO 860 940].

Features of glaciofluvial erosion

Glacial meltwater channels

Channels cut by glacial meltwaters are very common throughout the district (Bremner, 1928, 1934; Clapper-ton and Sugden, 1977). Known locally as ‘dens’, many of them are not associated with any significant modern drainage and may be referred to as ‘misfit’ valleys (Plate 14a), (Plate 14b). Three broad generic types of channel have been recognised on some maps: subglacial channels, ice-marginal channels and proglacial spillways. Although the best examples of each type are distinctive, most channel systems formed time-transgressively and distinctions between them are commonly blurred. Furthermore, many have a long and complicated history spanning more than one glaciation. For example, the Tore of Troup on Sheet 97 Fraserburgh (Map 4) and Tom’s Forest channel on Sheet 76E Inverurie (Map 8) are both deeply incised valleys that evidently predate the last glaciation because they are partly filled with till.

It should be born in mind that many of the largest and most impressive drainage channels in the district are not identified as such on the geological maps because they carry modern drainage and contain deposits mapped out within them.

Subglacial, ice-directed channels

Channels of this type, also known as ‘Nye’ channels (Benn and Evans, 1998), formed while most of the district remained buried beneath ice. Subglacial meltwaters were constrained by the regional hydraulic gradient to flow parallel to the direction of ice movement, irrespective of the local subglacial topography. The hydraulic gradient caused meltwaters to flow uphill within some subglacial channels, giving them their characteristic ‘up and down’ long profiles. They are most commonly preserved in cols cutting across topographical barriers that were orientated at an oblique angle to the former direction of ice movement. The channels commonly begin at the crests of such barriers and spurs where the hydraulic gradient and discharge was greatest (Sugden and John, 1976). Where cut into bedrock they are typically steep-sided, winding features that branch and reunite repeatedly. Subglacial channels cut into drift are generally broader, more open features, with a gently undulating long profile.

Ice-directed channels are particularly prominent to the south of Troup Head (Map 3); (Map 4) where they have been incised into the Middle Devonian conglomerates around the Tore of Troup (Merritt and Peacock, 2000, fig. 22). Most channels are straight, like the Den of Muck [NJ 817 607], but some are winding, like one to the south of Pennan Head [NJ 864 643].

Several subglacial channels aligned parallel to the direction of ice movement cut across the high ground linking Bennachie and the Hill of Fare (Map 8) and across relatively high ground to the west of Stuartfield [NJ 973 458] (Map 6). Particularly good examples include ‘My Lord’s Throat’, cut into fresh Bennachie Granite to depths greater than 20 m and the Bandodle channel, of similar depth, cut into the Hill of Fare Granite (both on (Map 8)). The latter channel extends east-north-eastwards from a col near West Bandodle for a distance of 3 km. Inland from Aberdeen, several channels cut across Tyrebagger, Elrick and Brimmond hills (Munro, 1986). Many of the major east-draining valleys such as the Don, Dee and Ythan probably also acted as major conduits for ice-directed drainage. In places they have irregular longitudinal profiles due to both ice and meltwater scour.

Ribbon channels, linking ice-scoured bedrock basins, occur north of the River Dee on Sheet 66E Banchory and numerous ice-directed channels are cut into Old Red Sandstone bedrock in Strathmore. A small, but impressive example of the latter is the Paldy Fair Den, near Glenfarquhar Lodge [NO 722 813] (Map 10); large examples include the group of north-east-trending channels around Woodburnden [NO 766 731]. Similar large features, such as the Glasslaw and Barras channels occur on Sheet 67 Stonehaven. A particularly good example of a subglacially formed ‘up and down’ channel runs north-eastwards from Craig Den [NO 803 831], on the boundary between Sheet 66E Banchory and Sheet 67 Stonehaven, from where it crosses a col and links with an esker in the vicinity of Lindsayfield [NO 821 844] (Map 11). This channel, and others on the southern half of Sheet 97 are depicted on the sand and gravel resource map of Auton et al. (1990). Another impressive example is the Devil’s Kettle, north-west of Stonehaven, which is cut into resistant, gritty psammitic bedrock to depths that exceed 30 m in places.

Ice-marginal and submarginal channel systems

Complex networks of shallow (1 to 5 m deep) channels are encountered across most of the district, even where clear evidence of glacial erosion is absent, such as in central Buchan. Two distinct sets of channels occur that tend to be orientated at right angles to each other, most commonly north–south and east–west respectively, particularly on Sheet 87W Ellon (Map 6). They were created mainly during deglaciation, as the East Grampian ice sheet receded westwards, and they are particularly common where bedrock is decomposed. Most systems exhibit increasing conformity with local topography at decreasing elevation, with ice-directed channels on the cols, passing at lower levels into ones increasingly orientated downslope in the lee of topographical barriers. The higher elements would have been created first, controlled by the overall configuration of the ice sheet, whereas the lower, slope-related channels would have been cut below thinning ice in the later stages of deglaciation (compare with Sissons, 1958, 1961a, 1961b). Good examples of these types of channels with their characteristically curved or crescentic lower courses include the Pitcowdens and Mount Shade channels on Sheet 66E (Map 10) and the Burn of Muchalls and Burn of Elsick channels on Sheet 67 (Map 11). Dendritic networks of lee-side channels occur locally, as for example, to the north-north-east of New Deer and to the west of Auchnagatt (Map 6).

The origin of the ‘lee-side’, slope-directed channel networks is not absolutely clear, but they were probably formed in the submarginal zone of the retreating ice sheet, particularly where stagnant ice was stranded on the lee side of topographical barriers as they became exposed during ice-retreat. The channels were probably cut initially into the surface of the ice (compare with Price, 1983; Clapperton and Sugden, 1977), then into the frozen ground beneath. This assumes that the ice sheet was cold-based, which is likely, especially as the lee-side channels are intimately related to suites of generally north–south-orientated, ‘ice marginal’, stoss-side channels that commonly occur on the west-facing slopes of the topographical barriers. Ice marginal channels such as these are generally thought to be associated with cold-based, polar ice sheets (Benn and Evans, 1998).

North–south-orientated ice marginal channels are particularly well developed along the eastern side of the valley of the Little Water, which joins the Ythan valley at Chapelhaugh [NJ 843 393], on Sheet 87W Ellon (Map 6). At this locality, the marginal channels generally have arched longitudinal profiles, suggesting that the meltwaters that formed them eventually drained westwards beneath the ice margin into a major submarginal channel that is now occupied by the Little Water. Higher ice-marginal channels are commonly truncated by, or feed into, lower ones, indicating that they formed progressively as the ice margin retreated. Other good examples of flights of channels such as these occur in the vicinity of Berryhillocks [NJ 683 597], south of Banff, and around Cook [NJ 800 574], south of Gardenstown (Map 3).

The orientation of the two sets of channels across Buchan indicates general east-north-east to east meltwater flow within the East Grampian ice sheet, at least in its retreat phase, a direction similar to that indicated by the orientation of drumlinoid features (see above), which almost reach the east coast. The pattern thus lends support to the hypothesis that most, if not all, of northeast Scotland was glaciated during the Main Late Devensian Glaciation (Clapperton and Sugden, 1977). However, individual ice-marginal and lee-side channel systems are coupled locally, and both sets of channels formed successively as the ice front retreated westwards, the drainage being continued both through topographically controlled marginal or submarginal channels and along precursors of the present stream system (compare with Bremner, 1928, 1934). The evidence does not support the existence of an integrated, contemporaneous network across the whole region as proposed by Clapperton and Sugden (1977).

Extensive ice-marginal and interlobate channel systems

Large-scale ice-marginal channels, commonly tens of metres deep, occur along the Banffshire coast on Sheet 96W Portsoy and Sheet 96E Banff, where they arc north-east towards the Moray Firth (Map 2); (Map 3). They formed at the margin of the combined central Highland and Moray Firth ice stream as it retreated westwards (Bremner, 1934). Similarly orientated channels occur along the southern margin of Sheet 95 Elgin and on Sheet 85E Knockando to the south, the largest being the Blackhills channel (Peacock et al., 1968) (Map 1). The relatively large scale of these features compared with those associated with the retreating East Grampian ice sheet across Buchan, may result from later formation when the climate was comparatively warmer and there was greater meltwater production. Unlike the subglacial channels described above, they generally have regular long profiles, sloping gently towards the coast.

South-eastwardly orientated channels north of Aberdeen, such as those drained by the Potterton and Blackdog burns, were probably initiated as ice-directed subglacial features (Map 9). However, most of the melt-water drainage along them took place when they emerged from the margin of retreating East Grampian ice sheet. The Culter Burn, Silver Burn and Den of Murtle channels near Peterculter carried drainage around the margin of the East Grampian ice sheet into the Dee valley as the ice-front retreated inland.

Several channels were carved out at the former margins of abutting ice masses. For example, on Sheet 87E Peter-head, a prominent channel system links the upper reaches of the Water of Cruden, west of Hatton, northwards to the valley of the River Ugie, to the east of Longside. The route follows the ‘misfit’ valleys of the Laeca Burn, Den of Aldie, Dens and West Den (Map 7). The system was probably formed close to the landward margin of ice that pushed in from the coast on more than one occasion, when meltwaters were diverted across the high ground linking the ‘Buchan Ridge’ and the Hill of Longhaven. The Den of Boddam [NK 114 415] is partly infilled with gravels of the Buchan Gravels Formation and certainly has a long history. Meltwaters also flowed eastwards at the margin of the Moray Firth ice stream. For example, several major channels ‘loop’ inland and back again along the coast between Buckie and New Aberdour (Map 2) (Map 3) (Map 4). To the east of New Aberdour, on Sheet 97, meltwaters flowed farther inland via several channels, towards Rathven and the Loch of Strathbeg.

The Strathfinella–Glen of Drumtochty channel on Sheet 66E Banchory (Map 10) and the south-west–northeast-trending Ury Home Farm–Limpet Burn channel, on Sheet 67 Stonehaven (Map 11), were eroded between the abutting margins of the East Grampian ice sheet and the Strathmore ice stream (Simpson, 1955). A group of secondary interlobate channels is present on the northern flank of the former channel. These channels drained eastwards into the main channel in the vicinity of Drumtochty Castle [NO 699 801]. The White Hill, Blackhills and Den of Cowie channels on Sheet 67 were formed by drainage at the margin of the Strathmore ice as it retreated (Plate 14a). This meltwater probably included drainage from ice-dammed lakes that developed between the two adjacent ice masses.

Proglacial spillways

In contrast to the subglacial and ice-marginal types of drainage channels described above, proglacial spillways generally have regular longitudinal profiles. When cut into bedrock, they also, typically, have a pronounced V-shaped cross-section, and interlocking spurs. Many formed as ‘overflow’ channels where ice-dammed lakes drained across cols (compare with Charlesworth 1926, 1956; Synge, 1956). Others formed when major drainage routes became blocked by glacier ice. Some would have been cut catastrophically as a result of glacier outburst floods or ‘jökulhlaups’ (Maizels and Aitken, 1991; Brown, 1994).

A good example of a spillway, the ‘Towie Spillway’, occurs on Sheet 86E Turriff between Darra [NJ 744 474] and Towie Castle [NJ 744 439] (Map 5). It formed when the valley of the River Deveron was blocked by ice downstream of Turriff, causing all drainage in the upper catchment of that valley to be diverted southwards into the Ythan valley (Figure 42). The Blackhills channel on Sheet 95 Elgin probably also functioned as an overflow channel, as did the valley of the Black Burn, which joins the River Lossie at Elgin (Map 1). Two prominent spillways draining former water-filled basins have been identified on Sheet 77 Aberdeen, near Blackburn and Kingswells (Munro, 1986).

The Westhills channel, west of Belhelvie, may have acted as a spillway for meltwater draining from the East Grampian ice sheet, before ice-marginal drainage became established in the nearby Potterton Burn channel (Map 9). Although no discrete channel-feature is preserved, it is clear that the extensive spreads of sand and gravel around Corby Loch were laid down by eastward flowing meltwater draining along the Don valley and across the interfluve towards the sea. This drainage route existed before flow was re-established in the lower reaches of the valley of the present river (Aitken, 1998).

A major spillway, Slack Den breaches the watershed between Glen Dye and Strathmore (the ‘Slack of Birnie’ of Synge, 1956) on Sheet 66E (Map 10). The spillway is cut, locally to a depth of more than 100 m, into resistant hornfelsed interbanded psammites and semipelites. The intake of the spillway occurs at about 350 m above OD, about 1.6 km north of the present watershed. The spillway was cut by drainage from a glacial lake that was ponded behind East Grampian ice blocking the lower reaches of Glen Dye. The south-west-directed drainage also dissected ice-contact sands and gravels laid down at the margin of the Strathmore ice stream during its retreat, depositing a fan of sand and gravel (containing clasts predominantly of psammite and semipelite) in the vicinity of Clatterin Brig [NO 664 782]. The surface of the fan stands some 30 m above the level of the present valley floor and has been dissected by the Slack Burn. The south-western end of the spillway also bisects the group of secondary interlobate drainage channels on the northern flank of the Strathfinella–Glen of Drumtochty channel (see above), indicating that meltwater drainage along the spillway commenced after the initial parting of the East Grampian and Strathmore ice streams.

Buried valleys

There are several examples of deep, buried valleys around the coasts of north-east Scotland. These channels were probably formed by subaerial drainage when sea level was relatively low during glacial stages, but when the land was not greatly depressed isostatically. Most are unlikely to date from the early Late-glacial period because sea levels were relatively high during deglaciation. They are commonly capped by glaciofluvial deposits, thus ruling out formation during the low sea-level stand of the early Holocene (Chapter 5). Moreover, many are partly filled with till and clearly predate the Main Late Devensian glaciation. Some may be very old features indeed, like one underlying the lowest reaches of the valley of the River Spey (Map 1), which is partly filled with unusual pale greenish grey sand (Aitken et al., 1979).

Deep channels have been located at the mouths of both the Don and Dee at Aberdeen (Figure 46). Both are largely filled with till and predate the Main Late Devensian glaciation. The River Don flows through a gorge cut in rock immediately upstream of the Bridge of Don [NJ 940 097], yet boreholes in the vicinity of Seaton Park, some 600 m to the south of the gorge, show that rockhead is well below sea level (Lumsden, 1958; Munro, 1986). A buried channel bottoming at about 15 m below OD lies beneath the floodplain of the River Dee between the Bridge of Dee [NJ 928 036] and the Wellington Suspension Bridge [NJ 943 050] (Peacock et al., 1977; Munro, 1986). Downstream of the latter bridge, a drift-filled channel diverges from the present valley and extends eastwards beneath the valley between Torry and Tullos Hill (Simpson, 1948). It reaches the coast at Nigg Bay (Map 9), where it bottoms at a depth of at least 40 m below OD, and extends for more than a kilometre beneath the bay (Munro, 1986). Geophysical data suggests that the feature is cut in bedrock and is narrow and gorge-like (Law, 1962). Another buried channel has been located beneath the centre of Aberdeen (Figure 46).

A further buried valley has been located beneath the valley of the River Ythan at Ellon (Merritt, 1981) (Map 6). It diverges from the present valley downstream of the town, passing south of the Hill of Logie [NJ 978 297] before reuniting with the present course at the estuary (see cross-sections on the Solid and Drift edition of Sheet 87W). Like those at Aberdeen, the buried channel of the Ythan is gorge-like and contains till. Seismic evidence suggests that it descends to some 40 m below OD at the Snub [NK 002 282] (Quaternary Research Association, 1975).

It is likely that many buried channels remain to be discovered, particularly along the coast. For example, the River Ugie flows through a gorge cut in bedrock near Inverugie [NK 102 482] (Map 7), yet boreholes drilled upstream failed to reach bedrock at 2 m above OD (McMillan and Aitken, 1981). This suggests that a buried channel occurs somewhere between Inverugie and St Fergus village [NK 098 520]. Inland, segments of old channels cut in rock and locally up to 10 m deep are known, for example, at Kirkhill Quarry, where the protection from glacial erosion provided by the features has helped preserve the complex Middle to Late Devensian succession there (Connell et al., 1982; Connell, 1984a; Appendix 1 Site 7).

Features of glacial or glaciofluvial deposition

Eskers, kames and moraines

Descriptions of individual eskers and kames are given in Chapter 6 in the notes on glaciofluvial ice-contact deposits, and their resource potential is discussed in Appendix 2. The most prominent eskers in the district are located on Maps 1 to 11. Perhaps the most studied and photogenic esker is that forming the Kippet Hills beside Meikle Loch at Slains ((Plate 15); Appendix 1).

Moraines are not common within the district (Charlesworth, 1956; Synge, 1956). Those that do occur have been described in Chapter 6 in the description of ‘hummocky glacial deposits’. Most moraines are of a rather chaotic nature and formed when large masses of ice stagnated in the lee of major topographical barriers. There are some examples of recessional moraines formed during ‘active’ glacial retreat, particularly on Sheet 66E Banchory and Sheet 67E Stonehaven, and a significant terminal moraine close to the limit of the Logie-Buchan Drift on Sheet 87W Ellon.

Erratics

It has long been recognised that the occurrence of boulders in areas far removed from their source is good evidence of the former direction of glacial transport (Plate 16a), (Plate 16b). At times too much emphasis has been placed on the distribution of distinctive ‘indicator erratics’ in early reconstructions of former ice sheets and transport pathways (e.g. Jamieson, 1906; Read, 1923; Bremner, 1934a, 1938; Synge, 1956). In addition, some of the observations made in the older literature are questionable. However, with caution, the known distribution of erratics still provides valuable evidence for determining the extent and interrelationships of former ice sheets in the district (Sissons, 1967). They also form an important element in the lithostratigraphical division of the Quaternary sequence at group level (Chapter 8). In more modern work, the complete description of clast types is made as ‘stone counts’ rather than relying on single far travelled clasts that may have complicated transport histories.

The paths determined for the transport of erratics in the Elgin area are reasonably consistent, revealing a predominantly easterly flow of ice, although there is also evidence of an older movement towards the south-east (Figure 45). The distribution of erratics of the granitic ‘Inchbae’ augen gneiss from central Ross-shire is particularly important, indicating the dominant ice flow into the Moray Firth from the north-west Highlands. The picture that emerges in the rest of north-east Scotland is more difficult to understand (Figure 47). The south-easterly movement of erratics both from the Boyndie metagabbro and the ‘lower zone’ cumulate basic and ultrabasic rocks of the West Huntly and Barra Hill intrusions, is consistent with the distribution of other erratics and rafts of sediment and fossils derived from the bed of the Moray Firth (see below and Whitehills Glacigenic Formation, Chapter 8). The radial movement of erratics away from the Netherly Diorite and West Huntly intrusion is more difficult to explain. One explanation could be that the outcrops of indicator rocks may be more extensive than thought hitherto because of poor exposure. For example, concealed masses of gabbro have been identified by geomagnetic surveys between Huntly and the coast (Munro, 1970; Munro and Gallagher, 1984). Another explanation involves the probable cold-based nature of former ice sheets over the interior of north-east Scotland, and their inability to scour away the products of previous glaciations. The interrelationship of East Grampian ice with more powerful ice streams occupying the Moray Firth basin may also be responsible for such patterns (Peacock and Merritt, 2000).

The analysis of ice-movement direction is further complicated by the reported south-westward carry of erratics of the Peterhead granite (Bremner, 1928, 1943; Synge, 1956; (Figure 47)). However, the occurrence of little known sheets and small bodies of granite, including red granite, between Fraserburgh and Peterhead on Sheet 97 Fraser-burgh and Sheet 87E Peterhead casts some doubt on the observation. The boulders do, however, lie along the margin of the ice that moved onshore from the North Sea basin giving rise to the Logie-Buchan Drift Group (Chapter 8). It is difficult to account for this direction of movement unless ice crossed the North Sea basin from Scandinavia. Some foundation to this argument has been derived from the discovery of erratics of rhomb-porphyries and larvikites also thought to have been carried from southern Norway (Read et al., 1923; Campbell, 1934; Bremner, 1939, 1943; Ehlers, 1988; Hall and Connell, 1991, fig. 74). These include a boulder on the beach at Portsoy, and several findings around Ellon and Aberdeen ((Figure 47); Sissons, 1967), including Nigg Bay (Appendix 1). However, the majority of confirmed Scandinavian erratics have been found inland of where deposits of the Logie-Buchan Drift Group occur at surface and are thus likely to have been transported onshore during a pre-Devensian glaciation.

Large rectangular blocks of hornfelsed sillimanite-bearing pelitic gneiss, many weighing several tons, form an erratic train eastward from its outcrop around Green Hill [NO 638 162], on the western margin of Sheet 76E Inverurie (Map 8). The erratic train extends as far as Monymusk, and isolated erratics many metres in diameter have been found farther east, in the vicinity of ‘The Horner’ [NO 748 156], east of Kemnay.

A south-eastward movement of erratics of andalusite schist and an eastward carry of mafic and ultramafic rocks from the Insch intrusion have been recorded in the northwest part of Sheet 77. The Insch erratics result from the final eastward flow of the East Grampian ice sheet, but the schists were probably carried by an earlier movement. At the coast, erratics of sandstone and conglomerate have been carried north-westwards by the onshore movement of ice that laid down deposits of the Logie-Buchan Drift Group. This movement possibly also carried erratics of the mafic and ultramafic Belhelvie intrusion, which extends offshore, rather than the north-eastward movement depicted on (Figure 47). An earlier south-east movement is required to account for the Belhelvie erratics found in the lowest exposed gravels at Nigg Bay (Bremner, 1928).

Bremner (1920a) reported the presence of erratics of quartzite and sandstone that had been carried north-east from Strathmore onto the watershed between the Dee and Cowie Water (Map 10). In particular, a quartzite boulder was found in the col between Craigbeg [NO 769 911] and Mongour [NO 757 900], over 330 m above OD. This north-eastward transport has been confirmed by the recent discovery of a rounded erratic of porphyritic andesite, many metres in diameter, ‘wedged’ in the bottom of a small glacial drainage channel occupied by the Burn of Anaquhat. The erratic was found [at NO 746 881], 260 m above OD and almost 5 km inland of the nearest outcrop of andesite in Strathmore. Bremner (1920a) postulated that the Old Red Sandstone erratics had been transported by an advance of Strathmore ice, prior to the arrival of ‘Dee Valley Ice’ (East Grampian ice sheet), which ‘completely cleared out or covered up all traces of Strathmore drift, in places where one is safe to infer that it was once present’. The age of the early advance of Strathmore ice is not known, but it possibly occurred during the maximum development of the Main Late Devensian glaciation of the district, prior to 22 ka BP (Appendix 1 Balnakettle).

A distinctive suite of erratics occurs in the deposits of the Logie-Buchan Drift Group, including Permian and Mesozoic limestones and dolomites in addition to Old Red Sandstone and conglomerates, indicating a source to the south and offshore (Appendix 1 Kippet Hills and Sandford Bay)

Glacial rafts

Glacial rafts or ‘megablocks’ are dislocated slabs of rock and unconsolidated strata that have been transported from their original position by glacial action (Aber et al., 1989). Erratics and large glacial rafts of Mesozoic rocks derived from the floor of the Moray Firth and the western North Sea are widespread in coastal areas, where they occur in deposits of the Banffshire Coast Drift Group and Logie-Buchan Drift Group respectively (Figure 4). Rafts are particularly common in the White-hills Glacigenic Formation of the former group, where they include reddish brown Permo–Triassic marls, dark grey to black Jurassic mudstones, Cretaceous glauconitic sandstones and Quaternary marine clays, silts and sands (Plate 17). The rafts range upwards in dimension from fist-size fragments to slabs that are probably several tens of metres thick and a kilometre or more in width. Several very large rafts of mudstone, silt and clay were worked for making bricks and tiles (Chapter 2).

The known localities of rafts are concentrated along the northern coast of the district where glacial striae also indicate that ice moved onshore towards the south-east at one stage. This suggests that the rafts were emplaced in zones of longitudinal compression, either proglacially or subglacially, as ice occupying the Moray Firth passed from deformable marine sediments on to bedrock, and was forced to override the high coastline and interior valleys (Peacock and Merritt, 1997, 2000). The rafts were detached in the Moray Firth basin, probably as high pore-water pressures built up in confined aquifers (sands interbedded with clays), a situation that has been invoked for the emplacement of similar rafts at Clava, near Inverness (Merritt, 1992b). Rafts were once assumed to have been moved while ‘frozen on’ to the base of cold-based ice sheets, but their intimate association with deformation tills suggests that failure may also have occurred along décollement surfaces in a subglacial deforming layer (Benn and Evans, 1998).

Rafts have been described at several localities in the district (Appendix 1 Boyne Limestone Quarry, Gardenstown, Oldmill and Moss of Cruden). At the time of writing, several rafts could be seen in a quarry at Ardglassie [NK 012 617], 5 km south-south-east of Fraser-burgh on Sheet 97 (Merritt and Connell, 2000). The rafts include sheared pale yellowish brown fine-grained sands with traces of ripple cross-lamination and dark grey clay. One raft of black lignitic clay contained a rich Late Jurassic palynoflora (information from Dr L A Riley, 1998). The rafts rest on relatively fresh bedrock and are capped locally by a thin unit of brown till. The quarry is situated on a south-west–north-eastorientated ridge, and most of the shear planes preserved in the rafts dip towards the north-north-west, suggesting ice push from that direction against the ridge. Many characteristic features of rafts are displayed, including shear zones, conjugate microfaulting, brecciation and folding.

Periglacial phenomena and features of mass wastage

Periglacial features

A wide range of phenomena that formed in former cold, periglacial environments has been recognised across the district (Galloway, 1958; FitzPatrick, 1956, 1963, 1969, 1987). Many elements of the present landscape are essentially relict periglacial landforms. The main effect of periglacial mass wasting has involved the downslope movement of till deposits and regolith and the subsequent accumulation of ‘head’ deposits as spreads on lower slopes and valley floors (Chapter 6). These processes have created the ubiquitous rounded hill crests of the region with valley sides that are commonly convexo-concave in profile (Ballantyne and Harris, 1994). Solifluction terraces are very common, as is frost shattered bedrock and structures indicative of the former presence of ground ice, notably cryoturbation structures, ice-wedge casts and tundra polygons (Plate 18). Although mostly occurring at the surface, features such as these have been identified at several stratigraphical positions within the Pleistocene sequence (Chapter 8). Aeolian periglacial features such as ventifacts, ‘cover sands’ and loess are likely to occur, but apparently have not been reported.

Intense periglacial activity last occurred during the Loch Lomond Stadial. Marked slope instability, with destruction of soils and remobilisation of diamictons by gelifluction, led to slope-foot accumulations of significant thickness, locally burying organic sediments of Windermere Interstadial age (Chapter 8). It is very likely that permafrost had been established previously during ice-sheet deglaciation, because ice-wedge casts and polygon networks affect the sediments of low river terraces in the lower Ythan and Ugie valleys (Clapperton and Sugden, 1977; Gemmell and Ralston, 1984; Armstrong and Paterson, 1985). These sediments were formed after the retreat of the coastal ice that laid down the Logie-Buchan Drift Group, possibly during a subsequent intensely cold period late in the Main Late Devensian glaciation (see Appendix 1 Ugie Valley).

Earth-pillars

Earth-pillars have developed on the slopes of several valleys in the district from spurs and ribs of material containing durable clasts in a friable, easily eroded matrix. Large boulders cap these features, which have been eroded mainly by rain-wash on steep slopes. Good examples occur on the eastern side of the valley of the River Spey near Fochabers at Craigs of Cuildell [NJ 332 553] (Map 1), where they have been carved out of weathered conglomerate (Plate 19).

Features of marine erosion

Two distinct sets of raised beaches, clifflines and rock platforms are common around the coasts of north-east Scotland. The sets were formed during periods of relatively high sea level, in Late-glacial times and in the mid-Holocene (Figure 48). Sea level was appreciably lower than it is today between these periods. Raised glaciomarine deposits have been described in Chapter 6.

During the past twenty thousand years or so, sea levels in north-east Scotland have been influenced by both the isostatic depression of the land under the ice load, and by global (eustatic) sea levels (Chapter 5). The largest amount of isostatic depression occurred in the western Highlands where the ice load had been greatest. That region has consequently been the centre of uplift since the ice melted. The rate of uplift was greatest during and immediately after deglaciation and has fallen exponentially since. The result of this pattern of isostatic recovery is twofold. Firstly, all raised marine deposits and associated erosional features in north-east Scotland are tilted eastwards and northwards away from a centre of uplift positioned in the vicinity of Rannoch Moor. Secondly, the younger, lower lying deposits and features are tilted less than the older ones. It follows that beaches of the same height in different places are not necessarily of the same age.

The older, Late Devensian set of raised features formed during Late-glacial times, that is, between the initial decay of the Main Late Devensian ice sheet in this region and the end of the Loch Lomond Stadial (Gray and Lowe, 1977). These remnants of shorelines lie up to about 30 m above OD, and are backed by discontinuous, degraded clifflines. The deposits locally merge into glaciofluvial sediments that were laid down contemporaneously. The Late-glacial set of shorelines and deposits are commonly truncated by an extensive cliffline, the Main Postglacial Cliffline. This feature formed during the mid-Holocene, at the climax of a major marine transgression, the Main Postglacial Transgression, which culminated between 6300 and 5700 BP in the Fraserburgh area (Appendix 1 Philorth valley) and at about 4750 BP in the lower Ythan valley (Smith et al., 1999). The Main Postglacial Cliffline commonly backs the younger set of Flandrian (Postglacial) raised marine deposits, which lie up to about 10 m above OD.

Late Devensian raised beaches

Late Devensian raised beach deposits are present on the Moray Firth coast between Forres and Banff, but there is little or no record of such features between the latter locality and Fraserburgh. They are best developed between the mouths of the River Findhorn and the River Spey on Sheet 95 Elgin, particularly on the seaward side of the Covesea–Roseisle ridge, which would at one time have been an island (Peacock et al., 1968, fig. 18; (Map 1)). They are less evident on the landward side of the ridge, where they are replaced in part by undulating, kettled glaciofluvial gravels, probably because the sea was excluded by stagnant glacier ice left behind during the westward retreat of the Moray Firth ice stream. Stagnant ice may also have been present on the southern side of the Spynie basin where Late-glacial beaches and raised marine sediments are again absent. The highest beaches occur as short stretches to the west of Burghhead (about 24 m above OD), Hopeman (about 24 and 19.8 m OD), and Easter Covesea (about 24 and 21 m OD). The hill on which Lossiemouth stands formed an island, and beach features at 22.5, 19.5 and 14.6 m OD were formerly seen in the western suburbs. These stretches of beach are formed of gravel, bouldery in places. At Greenbrae Quarry [NJ 137 692], the ‘19.8 m’ beach overlies a striated glaciated pavement. Detailed levelling has shown that the back feature of another beach descends eastwards from about 14.3 m OD south of Grange Hall [NJ 064 606] to about 13.7 m OD by Milton Brodie House [NJ 092 629] over a distance of 3 km.

There is no certain evidence for Late-glacial beaches between Lossiemouth and the mouth of the Spey, but farther east a raised beach at 14 m OD merges southwards into the Mosstodloch Terraces of the River Spey (Map 1). Eastwards, there is a gravel beach up to 2.4 m thick (top surface about 15 m OD) resting on till on the low quartzite cliffs at a locality [NJ 450 675] immediately south of Craig Head, west of Findochty (Map 2). A Late-glacial beach feature at about 22 m OD is seen south-east of Whitehills, and again in Banff, where it is followed by the A96 trunk road (Map 3). These beaches are generally formed of a metre or so of poorly sorted gravel and boulders. Spreads of glaciofluvial sand and gravel appear to merge with raised beaches (at about 25 m OD) at the mouth of the Boyndie Burn, forming a raised delta. Other degraded Late-glacial shoreline features were identified between Cullen and Tochieneal [NJ 522 653] (Map 2) by Read (1923), but they have not been confirmed by the recent revision. At Rosehearty (Map 4), an area of bare rock below 9 m OD, which extends for about a kilometre west of the village, may have been scoured by the Late-glacial sea, but no deposits are known.

At St Fergus, a raised beach has been identified at about 15 m OD (Appendix 1), but apart from this no Late-glacial raised beaches have been identified with certainty between Fraserburgh and Aberdeen. Farther south, Ritchie et al. (1978) reported a possible raised beach terrace at 10 to 12 m OD on the north-east side of the estuary of the River Ythan, above the level of Flandrian raised beach terraces at 3 to 6 m OD. However, terraces around 10 to 12 m OD are glaciofluvial (Merritt, 1981). Raised beaches may be expected adjacent to the low coastline north of Aberdeen, because raised marine clays are known at Aberdeen itself (see below) and the trend of isobases of Late-glacial beaches to the south suggests that the contemporary sea level would have been above 15 m OD (Sissons, 1983, fig. 4). The absence of Late-glacial beaches suggests that the Late-glacial sea here extended into an area of sediment-covered ice (compare with the Elgin area, above), the subsequent decay of which destroyed any evidence of beaches.

Bremner (1920a) was the first to question the extent of raised beach deposits shown as underlying much of the town of Stonehaven on early editions of Sheet 67. He recorded the presence of red till and ‘brick clay’ on the surface of the supposed ‘100 foot raised beach’ and an ‘absence of all appearance of sifting and arrangement by sea waves’ of the supposed beach gravels. These observations have been confirmed during the resurvey of Sheet 67. The gravels underlying the town that were formerly assigned to the ‘Second Raised Beach’ are shown as glaciofluvial sheet deposits on the current geological map.

Fragmentary Late Devensian raised shorelines have been recognised at six localities between Stonehaven and Inverbervie (Cullingford and Smith, 1980, table 3, fig. 2) and have been grouped with more extensive fragments that extend as far south as St Cyrus. The Stonehaven–Inverbervie shorelines occur at elevations ranging from 7.5 to

38.8 m OD between Garron Point and the mouth of the Cowie Water, at Downie Point, Bowdun Head, north of Crooked Haven, north of Upper Mill and at Kinghornie near Inverbervie (Map 11). Raised beach deposits have been mapped south-westward from Garron Point, at Downie Point and at Kinghornie. The deposits appear to be heterogeneous. A resistivity sounding at Garron Point indicated that the deposit is composed mainly of clayey silt, but in a degraded bluff 350 m south-west of Mains of Cowie [NO 877 868], 3 m of well-rounded cobbly beach gravel was formerly exposed, overlying red-brown till (Auton et al., 1988).

The shoreline fragments occur at distinctly higher elevations than the majority of those found farther south, between Dundee and Montrose (Cullingford and Smith, 1980, fig. 7) and have not been directly correlated with these well established sequences. Because of their high elevation and proximity to spreads of kettled glaciofluvial outwash, Cullingford and Smith (1980) postulated that the Stonehaven–Inverbervie shorelines developed early in the deglaciation of Strathmore, while ice was still present in the vicinity. These Late-glacial shorelines are therefore probably older than any of the features in the Montrose Basin and in the Forth and Tay valleys.

Flandrian raised beaches

The highest and best developed Flandrian raised beaches in the district were formed about six thousand years ago. They generally equate with the ‘Main Post-glacial Shoreline’ of eastern Scotland, which is tilted gently north-eastwards and drops to its lowest level between Fraserburgh and Peterhead (Figure 49). Along the coast of the Moray Firth it falls from about 9 m OD at Inverness, to 7.5 m OD near Elgin, and to near present high-tide level at Fraserburgh (Peacock et al., 1968; Smith et al., 1982, 1983; Firth and Haggart, 1989). Along the North Sea coast, it rises southwards to about 4 m OD in the Aberdeen area, and to between 5 and 5.5 m at Milton Ness [NO 772 647], 8 km north of Montrose, and 6.4 m at Montrose (Smith and Cullingford, 1985; Cullingford et al., 1991). At Fraserburgh, however, the youngest Flandrian shoreline is about one metre higher than the Main Postglacial Shoreline, indicating that sea level is continuing to rise hereabouts (Smith et al., 1982).

The most extensive Flandrian raised beach deposits in the district occur on Sheet 95 Elgin (Map 1), where they form storm beaches and shingle ridges that stretch from Kinloss to Burghead and from Lossiemouth to Portgordon (Ogilvie, 1923; Steers, 1937; Peacock et al., 1968). The western belt ranges from 180 m to 1.6 km in width and comprises shingle ridges that are more or less parallel to the beach and that are separated from one another by strips of peat. The ridges are locally buried beneath blown sand. The eastern belt ranges from 700 m to 2 km in width and contains similar features, except that the ridges bend southwards away from the shore at the mouth of the River Lossie (Plate 20). The ridges are formed of openwork, well-sorted, well-rounded, pebble to cobble grade gravel with sparse beds of sand. Although generally spherical, wave action has created local accumulations of tabular or bladed clasts. Shell fragments are common locally in the lower ridges. The gravel is generally very durable and has been exploited widely for aggregate (Appendix 2). The highest ridges occur at the back of the two belts, where they locally reach about 9 m OD. These features probably formed during the creation of the Main Postglacial Shoreline, whereas those at lower levels towards the sea formed subsequently, while sea level was falling.

There is a belt of peaty and sandy flats that stretches between Kinloss and Loch Spynie, behind the beach ridges (Map 1). The flats have been variously mapped as raised marine deposits, alluvium and peat, depending on what sediment occurs at the surface. They mark the sites of former brackish lagoons and tidal flats that existed when Roseisle and Branderburgh were islands in the mid-Holocene. The marine deposits generally consist of olive-grey or brown silty clay and yellowish brown micaceous sand. The River Lossie seems at various times to have found outlets northwards near its present mouth at Lossiemouth, and westwards south of Burghead. In historical times, there was a seaport beside the Palace of Spynie [NJ 231 658] until the late 15th century.

Flandrian raised beaches, formed mainly of shingle, occur discontinuously between the mouth of the River Spey and Troup Head, mainly at the heads of inlets and coves along this rugged, indented stretch of coast. The back feature of the raised beaches locally coincides with the base of the cliffline backing the pre-Late Devensian rock platform.

Raised Flandrian estuarine deposits are widespread in the Fraserburgh area, where they consist mainly of brown silty clay. The clay locally overlies peat, within which two laterally persistent layers of minerogenic material occur (Appendix 1 Philorth valley). The lower layer consists of grey sand and lies at about 1 m below OD. It has been dated to about 7 ka BP, and is attributed to a regionally significant tsunami (‘tidal wave’). The upper layer consists of micaceous sandy silt that was laid down between 6.3 and 5.7 ka BP during the Main Postglacial Transgression. The brown clay at the surface began to accumulate after about 4.8 ka BP during a subsequent transgression that probably continues today.

Shingle ridges and sand dunes, some of which were formed in historical times, extend along the coast between St Combs and Rattray Head, on Sheet 97 Fraserburgh. They separate the Loch of Strathbeg (formerly a tidal lagoon) and the historical port of Old Rattray from the sea (Peacock, 1983). Between Rattray Head and Peterhead (Map 7) the highest Flandrian beach is only a little above high tide level, and is backed by a low cliff cut in glacial deposits. A borehole drilled on the raised beach 350 m south-east of Annachie [NK 1085 5280] proved 2.6 m of sandy gravel resting on till at an elevation of 1.1 m below OD (McMillan and Aitken, 1981). The gravel is well rounded with a matrix of medium- to coarse-grained shelly sand. However, this raised beach, and others facing Sandford Bay and the Bay of Cruden, is largely concealed by blown sand. A borehole positioned amongst sand dunes behind the beach at the Bay of Cruden [NK 0858 3555] proved 8.1 m of grey sand and sandy gravel with shell fragments resting on till at 9.5 m below OD.

Several small deposits of grey silty clay flank the estuary of the River Ythan downstream of Kirkton of Logie-Buchan on Sheet 87W Ellon. These deposits resemble estuarine carse clays of the Forth estuary and are associated with the highest Flandrian shoreline at about 4.5 m OD (Smith et al., 1999). At Waterside [NK 003 273], a layer of grey sand attributed to a major tsunami (see above) has been found lying at about 0.5 m OD (Long et al., 1989). The main raised Flandrian shoreline is extensive between the Ythan estuary and Aberdeen, but it is generally buried beneath blown sand, like the low cliffline backing the feature. However, information from boreholes and trial pits in the vicinity of Menie Links indicates that the shoreline is mainly underlain by over 2 m of dark grey, micaceous fine-grained sand and sandy silt (Auton and Crofts, 1986). The deposit is peaty at the top and becomes gravelly towards the base. The peaty top is probably equivalent to several beds of peat that have been recorded below present-day beach deposits in the vicinity (Jamieson, 1858; Bremner, 1943). Extensive building development has taken place on the raised beach deposits between the mouths of the Don and the Dee, which are mainly composed of sand and gravel (Munro, 1986).

A broad raised beach, at about 5 m OD occurs immediately inland of the present high-water mark around Stone-haven Bay where it forms the flat-lying ground beneath Stonehaven sea front and shopping centre. The raised beach is generally about 200 m wide and is backed by a degraded cliffline, trending north–south, cut in bedrock capped by thin till and head (Barclay Street follows the break of slope at the base of the cliff). The raised beach reaches some 350 m in width near the mouth of the Cowie Water, where the flat ground has been developed as holiday caravan parks. Records from site investigations near Stonehaven harbour indicate that the raised beach deposits comprise pebbly sand and gravel, containing small amounts of shell debris locally. The deposits generally range in thickness from 2 to 5 m and overlie either red-brown laminated clay or bedrock. The nature of the stratification within the raised beach is unknown, but surface topography suggests that Allardice Street follows a north–south-trending beach ridge (Carroll, 1995a).

Flat-lying raised beach deposits have been mapped, inland of the high water mark between Bervie Bay (Inverbervie) and the southern margin of Sheet 67. They extend to between 5 and 10 m OD. Some of the highest deposits may be of Late-glacial age, but most of the lower spreads are products of marine transgression during the Holocene. The deposits consist of coarse-grained pebbly sand and gravel, with minor amounts of shell debris (Carroll, 1995b). The raised beaches are backed by the steep Main Postglacial Cliffline, which is cut into till and bedrock. Widespread landslides are developed along the cliff face and in many places, the raised beach deposits are concealed beneath slipped material at the foot of the cliff.

Raised wave-cut platforms and associated features

Wave-cut rock platforms with back-features a few metres above present high water mark are common along the rocky northern coast of the district on sheets 95, 96W and 96E. Between Burghead and Lossiemouth, at Cullen and at Gardenstown, a prominent rock platform and cliff is cut into the relatively weak Permo–Triassic and Devonian sandstones. There are ‘fossil’ sea-stacks on the platform above high tide level to the east of Covesea (Map 1), and also at Cullen. These features are probably partly post-glacial in age, but between Portsoy and Troup Head, the platforms probably occur at more than one level and some of them are partially overlain by glacial deposits of the Whitehills Glacigenic Formation (Chapter 8). The main platform reaches some 300 m in width between Whyntie Head and Whitehills (Plate 21), but where the cliffs are particularly high, such as at Troup Head, it is either missing or is represented only by rock reefs with concordant levels. There is little evidence of correlative platforms along the eastern coast of the district. One does occur at the ‘Bridge of One Hair’, south of the Bay of Nigg, on Sheet 77, where a sea-stack is joined to the mainland by a plug of red till some 30 m thick (Synge, 1956).

A raised rock platform standing at an elevation of about 23 m OD extends between Ruthery Head [NO 888 872] and the Glen Ury distillery [NO 871 869], north of Stonehaven. Bremner (1920) suggested that the platform is a ‘preglacial’ feature because it is capped by till. It is more likely to have been formed following a major pre-Late Devensian glaciation(s) when sea level was relatively high as a result of glacio-isostatic depression of the land. The till-covered feature is, after all, capped by Late Devensian raised beach deposits. Other fragmentary rock platforms occur to the south of Stonehaven at Downie Point, Bowdun Head, near Crooked Haven by Kinneff, and at Kinghornie, just north of Inverbervie.

Submerged wave-cut platforms

Along the eastern coast, the sea cliffs locally descend to a topographical break, possibly a wave-cut platform, at 55 to 60 m below OD (Crofts, 1975). A series of till-covered platforms have been detected from seismic records offshore between Stonehaven and Inverbervie (Stoker and Graham, 1985). These occur at about 30, 45 to 50 and 60 to 70 m below sea level. Wave-cut platforms such as these have been attributed to severe marine erosion during periods of lowered sea level coincident with periods of glaciation (Sutherland, 1984a). Submerged geos and stacks partially enclosed by till occur to the north of Newburgh and south of Aberdeen (Synge, 1956; Walton, 1959).

‘Pre-glacial’ raised beaches

Deposits of angular and rounded boulders, gravel, and sand lie locally between bedrock and till along the Banffshire coast. They were ascribed by Read (1923) to a ‘pre-Glacial raised beach’. However, re-examination of these sediments at Portessie [NJ 4510 6722], in a gully [NJ 4710 6815] between Findochty and Portknockie, and at Portnockie harbour [NJ 4882 6858] (Map 2), suggests that most, if not all, are of glacial or glaciofluvial origin (Peacock and Merritt, 2000). The deposit at the first locality is interpreted as ‘immature’ till whereas the gravel at the other two localities is almost certainly glaciofluvial, laid down or reworked immediately prior to the advance of the Moray Firth ice stream of the Main Late Devensian ice sheet across the area.

Chapter 8 Quaternary lithostratigraphy and correlation

Introduction

Some of the more recently published drift editions of sheets covering the district accommodate lithostratigraphical units of glacigenic deposits in addition to the conventional morpho-lithogenetic categories (Figure 5). For example, several glacigenic formations have been mapped on Sheet 66E Banchory and Sheet 67 Stone-haven. On other sheets, such as 87W Ellon and 76E Inverurie, conventional morpho-lithogenetic categories of deposit have been assigned to one or more lithostratigraphical group on the basis of gross lithological characteristics. It is beyond the scope of this publication to define and describe every lithostratigraphical unit, but some discussion of stratigraphical correlation is included, especially of the tills, because it is important for deciphering the history of events that have taken place. A set of generalised maps showing the distribution of drift groups and a selection of deposits, landforms and localities mentioned in the text is provided at the back of this publication (Maps 1 to 11).

Definition of terms

Lithostratigraphy involves the description, definition and naming of rock units (Whittaker et al., 1991). It is fundamental to stratigraphy because accuracy in biostratigraphy (based on fossils), geochronometry (based on measurements of age in years) or chronostratigraphy (apportioning units of strata to intervals of time) relies on the correct recognition of the relative spatial relationships of rock units, as does subsequent palaeogeographical reconstruction (see Chapter 1). Individual units are normally described and defined using their gross lithological characteristics and by their interrelationships with adjacent units, but this is not always possible for Quaternary glacigenic sequences, especially when they have been glacitectonised and/or occur as large glacial rafts. Such units are defined primarily on the basis of their unconformable bounding surfaces, rather than on lithology alone (e.g. Whitehills Glacigenic Formation, see below). Strictly speaking, the result is a hybrid of lithostratigraphy and allostratigraphy (Hedberg, 1976; Owen, 1987), the latter being based on the recognition of units bounded by discontinuities and their correlative conformities, such as in offshore seismic sequence stratigraphy.

In the Quaternary succession of north-east Scotland, discrete lithological units are named, as far as possible, after localities lying close to characteristic sections. They are then ranked in a formal hierarchy of Bed, Member, Formation and Group (Table 7). The Formation is commonly defined as the basic mappable unit in the hierarchy. Type sections (stratotypes) are chosen to illustrate typical lithologies, characteristics and relationships to adjacent units, but in complex and commonly poorly exposed areas a single section or borehole sequence rarely suffices and several partial type sections and reference sections may be required. Where formations and groups embrace numerous lesser ranked units, or there is poor exposure, a type area is given instead of a type section. Many type sections and type areas are defined in the descriptions of important localities in Appendix 1. All units established by BGS have been entered into the BGS Lexicon of named rock units and the index of computer codes (BGS, 2002).

Numerous units have been named in the literature; most are informal and as far as possible each one has been allocated here to one of the groups and assigned formation, member or bed status (Table 7). Names generally have not been changed unless ambiguity exists, where they have been used already for other lithostratigraphical units, or cardinal rules have been broken (e.g. the use of ‘lower’ and ‘upper’ qualifiers).

Since setting up the scheme outlined here, a revised correlation of Quaternary deposits in Scotland has been published by the Geological Society of London as part of a wider review of the British Isles (Sutherland, in Bowen, 1999). No groups are recognised in that publication, the ‘member’ becomes the basic mappable unit and formations are really ‘type sequences’. For example, the ‘Kirkhill Formation’ embraces all units at Kirkhill irrespective of lithology, provenance, genesis and age. Sutherland’s scheme is not adopted here for several reasons. The main objection is that most of his formations are not mappable units. Some include all drift deposits occurring in an area (e.g. Kirkhill) whereas others include only minor elements of the local drift sequence (e.g. Castleton and Camp Fauld). The omission of lithological qualifiers makes it difficult to comprehend what the units are and how they interrelate. There are also numerous spelling mistakes (e.g. Elton instead of Ellon Member) and inconsistencies (e.g. similar locality names appear at both formation and member level). Another problem is that several important papers have been published since the Geological Society Report was in preparation and recent mapping has helped clarify spatial relationships of many deposits away from type localities. For example, new information at Boyne Bay and Gardenstown indicates that the ‘Castleton Formation’ is almost certainly a collection of glacial rafts and not in situ, and the organic deposits at Camp Fauld and Crossbrae are now placed in a different stratigraphical context (Appendix 1). In general, Sutherland’s nomenclature has not been adopted here unless new names are required for formerly inappropriately named, or unnamed units.

The groups

Five lithostratigraphical groups of predominantly glacigenic (glacial, glaciofluvial, glaciolacustrine) deposits have been established in north-east Scotland. They reflect lithology and provenance rather than age and may encompass units formed during several glaciations. Although some nonglacigenic materials such as palaeosols and periglacial deposits are included, these units are closely associated with the glacigenic sequence and represent important events in the geological record of Quaternary events in the district. A ‘top down’ approach has been adopted in order to allow sufficient room in the classification for complex successions and to accommodate more information in the future. It also allows traditional morpho-lithogenetic units to be embraced on sheets where formal lithostratigraphical division has not been attempted.

Traditionally, it has been the practice in north-east Scotland to relate the glacigenic deposits to one of three distinct bodies of ice that are thought to have existed in the region (Figure 4). Three ‘series’ have been informally recognised (Sutherland, 1984a; Hall and Connell, 1991). Ice moving north-eastwards along the east coast led to the deposition of the Red Series (Jamieson, 1906; Synge, 1956), which include a variety of materials of a typically vivid reddish brown colour. Ice moving eastwards along the Moray Firth impinged onto the coastal lowlands and was responsible for laying down a suite of dark grey deposits assigned to the Bluegrey Series (Synge, 1956). In order to complete the picture, the typically sandy, yellowish brown diamictons laid down in the interior by ice flowing from the Grampian Highlands were assigned to an Inland Series (Hall, 1984a).

The tripartite division is retained in the present scheme, but the series are renamed as groups (‘series’ is a chronostratigraphical term). The Blue-grey series becomes the Banffshire Coast Drift Group and the Inland series becomes the East Grampian Drift Group. The Red series warrants division in order to separate deposits that were laid down by ice that moved onshore from the North Sea basin (Logie-Buchan Drift Group) from those that were derived entirely from ice flowing within Strathmore (Mearns Drift Group). Additionally, the deposits laid down by ice emanating from the Central Highlands and entering the area via the Spey valley are assigned to the Central Grampian Drift Group. Clast composition is the most important attribute in deciding from which ice stream a deposit was derived. Matrix colour remains a useful parameter for general classification. However, both methods require care in their application because distinctive clasts may have been reworked from older deposits and incorporation of local rock or sediment can locally change the typical colours of matrices.

The distribution of the five drift groups is shown approximately in (Figure 4) and on Maps 1 to 11. The relative positions of the ice streams changed with time during the last glaciation, and the ice streams of earlier glaciations probably did not cover exactly the same ground as in the Late Devensian (Chapter 5). There is consequently local, but stratigraphically significant interdigitation of deposits belonging to different groups.

Organic deposits, buried soils and periglacial sediments and structures

The lower relief areas of north-east Scotland are notable for the occurrence of a significant number and variety of Middle and Late Pleistocene nonglacial, terrestrial sediments. These serve as important stratigraphical marker horizons and provide evidence for ice-free episodes within the Pleistocene succession. Additionally, they provide important, and in some cases unique, evidence from which to reconstruct the climate of these periods.

Organic deposits

Peat, organic mud and organic-rich sand occur at numerous sites buried by later sediments. Most commonly found are Late-glacial organic deposits (Gunson, 1975), chiefly lake or pond muds and thin peat beds, organic muds and sands ((Table 8) Garral Hill, Byth, Woodhead, Mill of Dyce, Loch of Park and Glenbervie). The pollen record from these sites reveals that in the Windermere Interstadial, opening at around 13 000 BP, an initial phase of open habitat vegetation was succeeded by closed heath and scrub vegetation, with local growth of tree birch and pine. After around 11 500 BP, the temperature began to decline and tundra vegetation had become established by the start of the Loch Lomond Stadial at 11 000 BP.

No deposits have yet been found in north-east Scotland representing the last nonglacial period preceding the Main Late Devensian Glaciation. This ‘Sourlie Interstadial’ (Jardine et al., 1988), dated to around 30 000 BP, has been recognised from sites in western and central Scotland, where good evidence of tundra flora and fauna is preserved. The buried soil at Teindland and the peat at Crossbrae ((Table 7); Appendix 1) were formerly thought to date, at least in part, from this period, but the original relatively young radiocarbon dates from the sites probably result from contamination.

Evidence of older organic sediments has been found, but these deposits are beyond the range of radiocarbon dating. As only a few uranium series and luminescence dates have been obtained, the ages of these deposits are poorly known. However, detailed pollen investigations, and an increasing number of studies of Coleoptera allow tentative biostratigraphic correlation. Evidence of Early Devensian interstadial conditions is found at Camp Fauld, Crossbrae and Burn of Benholm ((Table 7); Appendix 1). The organic materials at these sites appear to be equivalent in age to Oxygen Isotope Stages (OIS) 5c and 5a, but correlation with one or other interstadial remains tentative. Birch and pine woodland was probably extensive in the region during the warmest parts of these interstadials. Episodes with dominant dwarf birch and willow scrub mark cooler episodes and the terminations of the interstadials are presaged by a decline into tundra conditions.

Organic muds and sands associated with podzolic buried soils at Teindland and Kirkhill contain pollen of temperate woodland. At Teindland, the pollen record shows an early phase of pine, alder and hazel woodland with grassland clearings, followed by vegetation indicative of increasingly colder conditions. This appears to mark cooling at the close of the last interglacial, OIS 5e. At Kirkhill, an older interglacial (possibly OIS 7) is represented by thermophilous pollen of pine and alder and very rare grains of birch, lime and elm (Appendix 1). This pollen is contained within a unit of organic-rich sand (Swineden Sand Bed), which is believed to have been redeposited from soil horizons in the vicinity. The pollen spectra are dominated by open grassland communities, which were probably present during the redeposition (information from J J Lowe, Royal Holloway University, London, 1990). This unit of sand is conformably overlain by sandy and gravelly gelifluction deposits (Corse Gelifluctate Bed) that formed in a periglacial environment ((Table 7); Appendix 1). As at Teindland, the organic sand contains evidence of climatic deterioration.

Interglacial and interstadial buried soils

Buried soils and associated weathering profiles occur at several sites and record significant intervals of temperate conditions (Plate 22); (Plate 23). Truncated podzolic soil profiles of interglacial status, developed in sand and gravel parent materials, occur in association with organic muds and sands at both Teindland (OIS 5e) and Kirkhill (possibly OIS 7) ((Table 7); Appendix 1). A second, younger, truncated soil profile at Kirkhill (Fernieslack Palaeosol Bed) is considered to represent a gleyed ‘brown earth’ of interglacial status (OIS 5e) developed in diamicton parent material (Connell and Romans, 1984). Weathering profiles, with disintegration of clasts and mobilisation and movement of iron and manganese, are developed on the surface of older tills, gravels and sands of the East Grampian Drift Group at Teindland, Boyne Limestone Quarry, Howe of Byth, Moreseat, Tillybrex (Ellon), and Kirkhill ((Table 7); Appendix 1). Around this last site, weathered till and gravel can be traced over an extensive area and represents both OIS 5e and earlier warm intervals.

Periglacial sediments and structures

A range of periglacial features have been recognised in units at several positions within the Pleistocene sequence. These include mass movement deposits, frost shattered bedrock, fluvial sands and gravels and structures indicative of the former presence of ground ice, notably cryoturbation structures and ice-wedge casts. It is clear that periglacial processes have had a significant impact on the landscape of north-east Scotland during most of the Quaternary, particularly across central Buchan (Chapter 7). Intense periglacial activity last occurred during the Loch Lomond Stadial when marked slope instability, with destruction of soils and remobilisation of diamictons by gelifluction, led to slope-foot accumulations of significant thickness, locally burying organic sediments of Windermere Interstadial age.

Evidence of older periods of periglacial activity includes a bed of sandy gelifluctate with ice-wedge casts that lies beneath rafts and diamicton of the Whitehills Glacigenic Formation, and on slightly weathered gravel at Oldmill (Appendix 1). These features probably represent one of the cold periods of the Middle or Late Devensian. Organic sediments of probable OIS 5e age at Teindland and either 5c or 5a age at Camp Fauld and Crossbrae (Appendix 1) record a decline to tundra conditions and the increasing slope instability is manifest as increased accumulation of sand, gravel and diamicton.

Undoubtedly the clearest record of recurring events is from Kirkhill, where quarrying activities over a period of 10 years allowed detailed resolution of multiple periglacial phases ((Table 7); Appendix 1). Extensive mass movement occurred towards the end of the Main Late Devensian glaciation and possibly also in the Loch Lomond Stadial (Manse Gelifluctate Bed). Cryoturbation, solifluction and ice wedge growth is also recorded in the Corsend Gelifluctate Bed lower in the sequence. The latter bed contains elements that probably formed both at the close of the last interglacial (OIS 5e) and prior to the complex phase of glaciation that deposited the Corse Diamicton and overlying Hythie Till ((Table 7)). ‘Silt-droplet fabrics’ have been revealed by micromorphological examination and probably record the development of an immature soil, or soils, close to a periglacial landsurface (J C C Romans, personal commnication in Connell, 1984c). These events may also be represented in an equivalent bed at Oldmill (Appendix 1).

Another sandy gelifluctate bed in the Kirkhill/Leys sequence (Corse Gelifluctate Bed) is associated with an ice-wedge cast network. It has been dated by luminescence to 142 ± 19 ka yr BP (Duller et al., 1995) and if correct, marks an important periglacial period within OIS 6. The oldest periglacial sediment recognised at Kirkhill is the angular felsite rubble of the Kirkton Gelifluctate Bed, which must date to OIS 8 or an older cold period.

Banffshire Coast Drift Group

This group of deposits contains clasts derived from the Moray Firth basin in addition to more locally occurring rock types. They occur from Inverness eastwards towards Peterhead (Figure 4). The diamictons, silts and clays are typically dark grey, calcareous and contain ice-worn Quaternary shell fragments in addition to abundant reworked Early Jurassic to Early Cretaceous microfossils. Permo–Triassic and Lower Jurassic to Upper Cretaceous strata crop out within the Moray Firth basin (BGS, 1977) and both clasts and large glacial rafts of these rock types are common in the tills of this group (Chapter 7). Locally, both the colour and composition of the tills stongly reflect the lithology of the underlying bedrock. The sands and gravels generally contain similar suites of clasts within the tills, and shell fragments are common.

Till

The most distinctive tills of the Banffshire Coast Drift Group are bluish grey in colour, clayey and contain erratics and rafts of Mesozoic rocks and Quaternary sea floor sediments derived from the Moray Firth basin, together with reworked fossils and shell fragments. They have a discontinuous distribution at the surface and some older units occur at depth interbedded with deposits of the other groups ((Figure 4); (Table 7)). Though it has long been concluded that most of these bluish grey diamictons were deposited by ice moving landwards from the Moray Firth basin (Jamieson 1906; Read, 1923; Bremner, 1928, 1934), direct evidence for such ice movement, such as striations or clast fabric measurements, are rare (Chapter 7). Furthermore, there has been little consensus concerning when, and how many times, ice moved onshore to produce these distinctive deposits (Chapter 5).

Whitehills Glacigenic Formation

Blue-grey diamictons occur widely, but not extensively near, or at, the surface between Elgin, Peterhead and Ellon. Most of them probably belong to one laterally discontinuous, allostratigraphical unit, the Whitehills Glacigenic Formation (Peacock and Merritt, 1997, 2000). The formation is generally capped by brown, red or grey till containing clasts derived from the west or west-northwest (Jamieson, 1906; Read, 1923; Peacock et al., 1968; Aitken et al., 1979; Hall et al., 1995a) and is almost certainly of Late Devensian age (Peacock and Merritt, 1997, 2000). It is named after the village lying 4 km to the west of the type section in the Boyne Limestone Quarry [NJ 612 658] on Sheet 96W Portsoy ((Map 2); Appendix 1 Boyne Bay, Gardenstown and King Edward).

The Whitehills Glacigenic Formation is notable for its inclusion of a relatively large proportion of glacial rafts of Mesozoic strata and Quaternary sediments derived from the Moray Firth basin (Plate 17b). The rafts include mudstones, friable sandstones, clays, silts, sands and gravels (Chapter 7). Jurassic material predominates in the west, whereas some Lower Cretaceous rocks and fossils are also present in the east. Some rafts crop out at the surface and are large enough to be mapped out individually on Sheet 96W and Sheet 96E, such as the one at Whitehills itself (Map 3), which has been largely removed for making bricks and tiles (Chapter 2). A ‘blue oolite clay’ with striated stones, probably a till including rafts of clay, was formerly worked at Lumbs [NK 028 578] (Milne, 1892a), on Sheet 97 (Map 4), and further occurrences of locally shelly ‘blue clay’ have been reported to the south and west of Fraserburgh (Jamieson 1906). Large Jurassic erratics were formerly exposed at Atherb and Brucklay Castle, to the south-west of Strichen ((Map 6); Jamieson, 1906) and around Plaidy and Turriff ((Map 5); Read, 1923). Dark grey to black till, locally shelly, has been recorded at depth at several localities in the catchment of the River Ugie (McMillan and Aitken, 1981). There are few exposures, but at the time of writing several rafts could be seen in a quarry at Ardglassie [NK 012 617], 5 km south-south-east of Fraser-burgh (Chapter 7), and others were exposed at the Oldmill site ((Plate 24); Appendix 1).

Several other units of blue-grey till that have been recognised in the district may correlate with the White-hills Glacigenic Formation (Table 7). For example, in the valley of the Red Burn, south of Fochabers, Hall et al. (1995a) reported a dark till (Altonside Till) with Mesozoic erratics and a fabric indicating derivation from the northwest ((Map 1); Appendix 1 Teindland). They correlated the Altonside Till with other dark tills recorded by Aitken et al. (1979) in the lower Spey valley. At the King Edward site ((Map 5); Appendix 1), south-west of Gardenstown, stones in an unnamed unit of dark till retain striations suggesting ice movement from the west or north-west (Jamieson, 1906). Although that unit of till is apparently absent at the Howe of Byth site to the east ((Map 4); Appendix 1), the Howe of Byth Gravel Formation there contains some Mesozoic erratics and might have been laid down contemporaneously with the Whitehills Glacigenic Formation. The East Leys Till that occurs between Fraserburgh and Peterhead is another correlative, although it is not overlain by younger tills (Appendix 1 Kirkhill). It is typically decalcified to a depth of about 4 m and includes red granite, probably from the Peterhead area, and local Inzie Head Gneiss, in addition to chalk and flint. The dark grey silty clay matrix of the till contains dinoflagellate cysts of Late Jurassic to Early Cretaceous age (Hall and Jarvis, 1993a). The East Leys Till is generally thought to have been laid down during the Devensian because it lies stratigraphically above the Fernieslack Palaeosol Bed (formerly Kirkhill Upper Buried Soil, which is likely to be Ipswichian in age or older (Duller et al., 1995). The Pitlurg Till, which occurs between Ellon, Cruden Bay and the Buchan Ridge (Appendix 1 Ellon) also includes materials that were apparently carried westwards from the North Sea basin in addition to Jurassic erratics and fossils derived from the Moray Firth (see below).

Tills overlying the Whitehills Glaciogenic Formation

At the type section of the Whitehills Glacigenic Formation in the upper part of the Boyne Limestone Quarry (Appendix 1) the shelly till and rafts are generally overlain by a unit of brown stony till containing clasts derived mainly from the west. This unit, named as the Old Hythe Till Formation (Peacock and Merritt, 2000) is correlated with the Crovie Till Formation (Peacock and Merritt, 1997) occurring at Gardenstown to the east (Appendix 1). At both localities the contact of the brown till with deposits of the underlying Whitehills Glacigenic Formation is gradational. Furthermore, at both localities the brown tills are judged to have been laid down by ice flowing towards the east following a change of direction from the south to south-easterly movement that brought about the emplacement of the Whitehills Glacigenic Formation and its included rafts. It is apparent that the Crovie Till Formation and Old Hythe Till Formation are likely to have been laid down during the same event, but by East Grampian ice and ice from the central Grampians respectively ((Figure 4); (Table 7)).

A similar sequence occurs at the Oldmill site (Appendix 1; (Plate 24)), on Sheet 87E Peterhead, where an un-named unit of brown till is tentatively correlated with the Hythie Till Formation at Kirkhill. The brown till has erratics derived from farther west and a weak ‘west to east’ fabric (Merritt and Connell, 2000). It rests on dark grey till containing rafts of Jurassic shale, Quaternary marine clay and sand (Table 7). At Kirkhill, dark clay with Mesozoic erratics (Corse Diamicton) also locally underlies brown till with clasts derived from the west and southwest, the Hythie Till (formerly Kirkhill Upper Till). The former quarry at Kirkhill is situated within a drumlinoid ridge trending west–east suggesting that ice last flowed from the west. Other ice-moulded features in the area are similarly orientated ((Map 6)). The history of events in Buchan thus appears to match that elucidated at the Boyne Limestone Quarry and Gardenstown, with ice first flowing across the area from the north-west, transporting rafts from the Moray Firth, before changing towards an easterly direction.

Essie Till Formation

Calcareous blue-grey tills occur at the surface between Rosehearty and Peterhead on Sheet 97 and Sheet 87E. Most glacial striae, ice moulded features and erratics (Map 4); (Map 7) indicate that the last movement of ice was towards the east-south-east or south-east across the area. There is also evidence of a minor re-advance of ice from the north and east that affects a narrow coastal zone between Cairnbulg Point and St Combs (Peacock, 1997). The tills are commonly mottled grey and brown at the top and contain a variety of materials eroded from the bed of the Moray Firth. They cannot easily be distinguished lithologically from those of the Whitehills Glacigenic Formation described above, but they do not appear to have been affected by any subsequent ice movement from the west as at Boyne Bay, Gardenstown, Kirkhill and Oldmill (Appendix 1). In the absence of overlying tills derived from inland, it is uncertain whether these coastal blue-grey tills, assigned here to the Essie Till Formation, are strictly equivalent to those of the Whitehills Glacigenic Formation (or the East Leys Till); they probably relate to a later event.

The type locality of the Essie Till Formation is around South Essie farm, near St Fergus (Map 7) where several pipeline trench exposures revealed more than 2.5 m of dark grey, calcareous silty clayey diamicton with shell fragments and erratics of pink granite, quartz and metasedimentary rock (M. Munro in British Geological Survey Records). The unit generally overlies a reddish brown, calcareous silty clayey diamicton that probably correlates with the Hatton Till of the Logie-Buchan Drift Group. The Essie Till Formation was also exposed in a BGS Registered trial pit NK05NW/6 [NK 0280 5778] dug in the vicinity of the former Lumbs claypit (Map 4) where a ‘blue oolite clay’ with striated stones was formerly worked (Milne, 1892a). The pit exposed over 3.8 m of stiff, dark grey clay and pebbly clay diamicton with lenses of reddish brown diamicton, small masses of red clay and very sparse fragments of shell and chalk. The dark diamicton contained sparse pebbles and cobbles of marl and red sandstone whereas the red diamicton contained granules of red sandstone and mudstone. The red materials and black mudstones and chalk are derived from outcrops of Permo–Triassic and Cretaceous strata lying within the Moray Firth basin to the north (Figure 2). The trial pit was dug into an isolated, drumlinoid hill trending northwest to south-east. Both the glacial streamlining of the feature and the composition of the till indicate ice flow towards the south-east.

Several boreholes in an area of some 15 km² to the north of the River Ugie, on Sheets 97 and 87E, proved dark shelly diamictons that are probably equivalent to the Essie Till Formation (McMillan and Aitken, 1981). For example, Borehole NK05SE/9 drilled at Kinloch Farm [NK 0989 5093], 4 km north-west of Peterhead, proved 5.4 m of stiff, dark bluish grey pebbly clay diamicton with clasts of pink granite, quartz and schistose metasedimentary rock. As in the pipeline trenches, the underlying unit (3.2 m thick) is similar in lithology, but is dark reddish brown in colour and contains sparse boulders of granite. It rests on 9.8 m of crudely laminated, red, reddish brown and yellowish brown, micaceous silt of probable glaciolacustrine origin and equivalent to the Ugie Clay Formation of the Logie-Buchan Drift Group. Both diamictons are probably deformation tills and result from an advance of coastal ice across glaciolacustrine deposits ((Figure 42)a). The latter were laid down in lakes ponded against coastal ice after ice from inland had retreated westwards.

The clasts of pink granite in the Essie Till Formation are probably derived from a small outcrop of coarse-grained pink granite to the north-west of Crimond (Map 4) described by Wilson (1882). Commercial records indicate that pink granite also underlies part of the gas terminal at St Fergus.

Arnhash Till Member

There are several localities along the northern coast of the district where glaciofluvial deposits have been disturbed by, or locally overridden by a late-stage re-advance of the Moray Firth ice stream (Peacock and Merritt, 2000). The readvancing ice locally laid down a thin deposit of reddish brown to bluish grey gravelly diamicton, for example on the sands and gravels at Arnhash (now Pitnacalder) gravel pit [NJ 872 628], south of New Aberdour (Map 4). This unit is named here as the Arnhash Till Member of the Essie Till Formation. Other sites include the former Gallows Hill pit [NJ 512 665] near Cullen (Bremner, 1928), on Sheet 96W, Troup Head on Sheet 96E (Peacock, 1971) and Broomhead pit [NJ 983 640] southsouth-west of Fraserburgh on Sheet 97. At Pitnacalder and Broomhead, deltaic deposits of the Blackhills Sand and Gravel Formation have been pushed by ice nudging inland from the north-west and north-north-east respectively (Peacock and Merritt, 2000e, fig. 26).

Older blue-grey tills

At Camp Fauld, on the Moss of Cruden (Appendix 1), excavations revealed greenish grey diamicton (Aldie Till) stratigraphically overlying a dark grey diamicton (Camp Fauld Till). The Aldie Till contains clasts of flint, quartzite, fresh red granite, red sandstone, mica-schist and weathered basic rocks (Hall and Jarvis, 1994; BGS records) whereas the Camp Fauld Till contains quartzite, flint, red granite and Lower Cretaceous sandstone; the last-named rock was probably derived from the nearby Moreseat outcrop (compare with Hall and Connell, 1982; Hall and Jarvis, 1994), rather than from offshore. Elsewhere on the Moss, black clay with clasts of granite and quartzite was seen near Moreseat [NK 053 403], where it underlies rafts of Lower Cretaceous sandstone and clay (Jamieson et al., 1897). Dinoflagellates of possible Late Jurassic to Cretaceous age have been recovered from black clayey till, probably also Camp Fauld Till, at a locality about 500 m west of Moreseat (Harland in Hall and Connell, 1982).

Whittington et al. (1993) concluded that both the Aldie and Camp Fauld tills were deposited by ice moving from west to east or north-west to south-east. The sequence therefore seems at first sight to be similar to that occurring at Kirkhill, Oldmill, Gardenstown and the Boyne Limestone Quarry. However, an Early Devensian or latest Eemian age for units of peat overlying the Camp Fauld Till at Camp Fauld (Whittington et al., 1993; Whittington, 1994; Duller et al., 1995) indicates that the till was laid down in a pre-Devensian glaciation predating the deposition of the Whitehills Glaciogenic Formation (Table 7). The possibility of two ‘dark coloured boulder clays’ in the Ellon district was first suggested by Bremner (1928, p.163).

Further complications occur to the south of the Buchan Ridge where three and locally four till units are separated by deposits of sand and gravel (Merritt, 1981; Hall and Jarvis, 1995). From Ellon north-eastwards towards the Buchan Ridge, dark grey till (the ‘indigo’ till of Jamieson, 1906 and/or the Pitlurg Till of Hall and Jarvis, 1995) underlies red diamictons of the Logie-Buchan Drift Group. At Ellon, the red deposits rest on grey till of western provenance whereas at Bearnie, they rest on a till of westerly derivation that incorporates material derived from the Pitlurg till (Appendix 1 Bellscamphie; (Figure A1.20)). The igneous and metamorphic clasts in both the Pitlurg Till and ‘indigo’ tills are likely to have been derived from the gneisses, gabbros and norites that crop out to the north-west of Ellon. The ammonite in dark indigo clay recorded by Jamieson (1906) at Ellon, and dinoflagellate cysts found in the Pitlurg Till by Hall and Jarvis (1995), indicate derivation ultimately from the Moray Firth, as rocks of this age are exposed only on the sea floor in an narrow outcrop extending westwards from Fraserburgh (Gatliff et al., 1994). A weak north–south fabric has been detected in the Pitlurg Till by Hall and Jarvis (1995), but direct transport from the north-west through to the north-east is difficult to accept because the dark grey tills contain few, if any, erratics of flint and white quartzite from the Buchan Ridge. An amino-acid ratio derived from a shell fragment in the Pitlurg Till (Appendix 1 Bellscamphie) suggests that it is the equivalent of the Whitehills Glacigenic Formation rather than the older Camp Fauld Till. However, most ratios obtained indicated older ages.

Similar diamictons to the ‘indigo’ till have been reported at South Anderson Drive in Aberdeen (Bremner, 1934, 1943; Hart, 1941) close to the Bridge of Dee [NJ 927 036]. Bremner reported that ‘dark shelly boulder clay’, named here as the Anderson Drive Diamicton Formation, was overlain by a ‘grey boulder clay’ with characteristic erratics derived from Belhelvie and Barra Hill to the north-west (Appendix 1 Nigg Bay). The latter till is probably equivalent to the Den Burn Member of the Banchory Till Formation of the East Grampian Drift Group (Table 7).

Several other units of grey diamicton containing Mesozoic erratics that have been recorded in the district are probably significantly older than the Whitehills Glacigenic Formation. For example, the lowest stratigraphical unit recognised by Hall et al. (1995a) in the vicinity of the Teindland site (Appendix 1) is a reddish brown till that includes a few erratics of Mesozoic siltstone and sandstone. The presence of these erratics derived from the Moray Firth basin suggests that this diamicton, named as the Red Burn Till, should be included in the Banffshire Coast Drift Group despite its colour. The age of the till is not known, but it is likely to predate the Teindland Buried Soil Bed, for which an Ipswichian age is likely (Hall et al., 1995; (Table 7)).

Origin of the blue-grey tills

The Essie Till Formation rests on red diamictons of the Logie-Buchan Drift Group in the vicinity of Peterhead. It occurs at the surface and is most probably Late Devensian in age. Most other blue-grey tills, with their erratics and rafts of Mesozoic rocks, are buried beneath thin brown tills of inland derivation and are conveniently regarded as part of a single lithostratigraphical unit, the Whitehills Glacigenic Formation. As the brown tills generally overlie the blue-grey ones with little evidence of a break, both were probably laid down during a single glacial episode, during which the direction of ice movement across Buchan changed from south-east to east. This episode either occurred in the Middle Devensian or, more likely, the early part of the Late Devensian. On entering the North Sea basin to the east of Fraser-burgh, part of the Moray Firth ice stream may have been deflected westwards again onto the coast south of Peter-head, probably by Scandinavian ice lying offshore.

It appears also that ice moved onshore from the Moray Firth during at least one major glaciation before the Ipswichian to lay down the Camp Fauld Till near the crest of the Buchan Ridge and the Red Burn Till at Teindland (Table 7). The bluish grey Benholm Clay Formation at the Burn of Benholm site was possibly laid down by Moray Firth ice deflected onshore to the south of Inverbervie, but its distinctive palynoflora suggests derivation from the south (Appendix 1).

Sands and gravels

Typically, these glaciofluvial deposits include shell fragments and clasts of sedimentary rocks derived from the Moray Firth basin in addition to locally occurring rocks. They occur mainly at the surface and are considered to be of Late Devensian age, although older deposits are known at depth locally (Table 7). The surficial deposits lie between Elgin and Fraserburgh and extend inland for no more than about 10 km. They are assigned here to the Blackhills Sand and Gravel Formation (Peacock and Merritt, 1997). Two distinct suites of deposits can be distinguished on the basis of landform: moundy ‘ice-contact deposits’ and terraced ‘sheet deposits’. The two may be regarded as informal members of the formation.

Other units include the Auchmedden Gravel of Hall et al. (1995b), which forms the upper part of the outwash fan being exploited at the Howe of Byth gravel pit (Appendix 1). It probably correlates with the Blackhills Sand and Gravel Formation and overlies the Byth Till, also probably of Late Devensian age. The latter rests on a lower unit of gravel, the Byth Gravel, which is of Mid-Devensian age or older (Hall et al., 1995b; (Table 7)). A unit of the Kirkhill sequence (Appendix 1) is assigned tentatively to the group, the Kirkhill Church Sand Formation (formerly Kirkhill Upper Sands). This unit is probably Late Devensian in age.

Glaciolacustrine deposits

Spreads of clays, silts, sands and gravels extend along the coastal hinterland between Portknockie and New Aberdour on Sheets 96W, 96E and 97. They were first described in detail by Read (1923), who named them the ‘Coastal Deposits’. He also included dark grey shelly tills and clays in the unit, but these lithologies are now known to belong to a separate, underlying formation composed mainly of till (Peacock, 1971), since named as the White-hills Glacigenic Formation (Peacock and Merritt, 1997). The main bodies of sand and gravel within the ‘Coastal Deposits’ are now included in the Blackhills Sand and Gravel Formation. The remaining glaciolacustrine silts and clays are defined as the Kirk Burn Silt Formation, the type section of which occurs at Castle Hill, Gardenstown (Peacock and Merritt, 1997; (Map 3); Appendix 1).

The main outcrops of the Kirk Burn Silt Formation lie between Portknockie and Portsoy, at Macduff, between Gardenstown and Pennan, and to the north of New Aberdour (Map 2); (Map 3); (Map 4). For the most part the deposits form flat or gently undulating ground within 5 km or so of the coast (Peacock, 1971). The sediments consist mainly of ochreous to dark olive brown, laminated to thin-bedded silts and clays up to about 10 m in thickness. Bands of ferruginous nodules occur locally. Pebbles are not common, but where they do occur, they include clasts of white decalcified Cretaceous limestone. The sediments show signs of settlement and minor movement, including slumping, but there is no evidence that they have been affected by a significant postdepositional glacial advance from inland as deduced by Read (1923), or that they are widely covered by solifluction deposits as described by Synge (1956). They have been deformed apparently by ice moving from the sea at Troup Head (Peacock and Merritt, 2000; (Map 3)).

The deposits of the Kirk Burn Silt Formation were laid down during deglaciation of the Main Late Devensian ice sheet. East Grampian ice had retreated inland but an active ice stream remained in the Moray Firth, forming a barrier a short distance to the north of the present coastline and causing lakes to form against the high ground to the south (Figure 42). It seems that lakes first formed to the north of New Aberdour, draining eastwards along the coast. A larger lake then formed between Gardenstown and Pennan, draining via the Afforsk Spillway into the valley of the Deveron via the valley of the Idoch Water. Much larger lakes formed subsequently between Portknockie and Macduff.

The deposits of the Kirk Burn Silt Formation are prone to landslipping (Chapter 2; (Plate 3)).

Glaciomarine deposits

Glaciomarine deposits of the Banffshire Coast Drift Group were laid down during the overall retreat of the Moray Firth ice stream at two distinct localities within the district. The older of the two deposits, the St Fergus Silts Formation, crops out around the Loch of Strathbeg on Sheet 96 and in the vicinity of the St Fergus gas terminal on Sheet 87E. The younger deposit, the Spynie Clay Formation, is associated with Late-glacial raised beaches in the Elgin area.

St Fergus Silt Formation

At the type locality, the St Fergus Silt Formation underlies alluvium of a former freshwater loch that lies behind the belt of coastal sand dunes forming St Fergus Links [NK 124 434], to the north of Peterhead ((Map 7)). The type section is taken to be BGS Borehole NK15SW/1 [NK 1030 5344], in which some 16 m of laminated silty clay and clayey silt with sand laminae and shell fragments were recorded, resting on till (McMillan and Aitken, 1981; Peacock, 1999). The deposits vary from dark grey to dark greyish brown, and are both unbedded and laminated, calcareous and contain dispersed clasts up to cobble size that are probably dropstones. The lithology and fauna suggests that the St Fergus Silt Formation is glaciomarine (Hall and Jarvis, 1989; Peacock, 1999), and two adjusted radiocarbon dates of about 14.9 BP and 14.3 BP (Hall and Jarvis, 1989) on marine bivalves (Hiatella arctica) indicate that it was laid down in a period prior to the Windermere Interstadial. The general surface level of the St Fergus Silt Formation and of an associated raised beach imply a marine transgression to at least 12 m OD (see Appendix 1).

Spynie Clay Formation

Late-glacial raised marine beds composed mainly of silty clay, are associated with raised beaches between Elgin and Lossiemouth on Sheet 95 Elgin (Chapter 7). These beds, named the Spynie Clay Formation by Peacock (1999), underlie low-lying ground around Loch Spynie [NJ 237 667] and Lossiemouth Airfield (Map 1). The maximum recorded thickness is 12.5 m (Peacock et al.,

1968). North of the Spynie basin the formation attains a minimum level of 10 m above OD, and possibly reaches over 20 m OD. On the south side of the basin the mapped level is only a little above OD, probably because the sea was excluded by stagnant ice from this area when sedimentation had begun elsewhere.

The type locality of the Spynie Clay Formation is the former Spynie Claypit [NJ 232 672] (Peacock, 1999), where up to 6 m of sand is underlain by thin beds of shell debris and peat, resting on some 9.5 m of ‘blue’, yellowish red and black clay with scattered boulders, locally interbedded with sand and diamicton (Buchan, 1935; Eyles et al., 1946). Fossils, chiefly the brittle star Ophiocten sericium (formerly Ophiolepis gracilis) (Plate 25), numerous ‘Leda pygmaea’ (probably Yoldiella lenticula) and rare Portlandia arctica, have been found between 5 and 6.5 m below the surface of the clay (Buchan, 1935). All are decomposed. The regular lamination, the presence of well-dispersed, probably ice-rafted boulders, and the apparent absence of bioturbation support the view that much of the sediment was deposited rapidly in a glaciomarine environment. The low-diversity, high-arctic fauna is similar to that in the Errol Clay Formation of eastern Scotland, which has yielded radiocarbon ages between 12.8 and 14.3 ka BP (Peacock and Browne, 1998; Peacock, 1999).

Mearns Drift Group

The Mearns Drift Group comprises interbedded diamictons, glaciolacustrine silts and clays, and glaciofluvial sands and gravels that are all typically vivid reddish brown in colour and contain clasts that are derived mostly from the andesitic volcanic rocks and red Devonian conglomerates, sandstones and siltstones of Strathmore. The group broadly equates with the southern outcrop of the ‘Red Series’ of Bremner (1916), Sutherland (1984a), Sutherland and Gordon (1993), and Hall and Connell (1991). The deposits were laid down by, or at the margins of, ice that flowed north-eastwards into the North Sea basin from Strathmore (Figure 4). The formations in the group are depicted on Sheet 66E Banchory and Sheet 67 Stonehaven. Deposits occur at least as far north as Nigg Bay (Appendix 1).

Three formations have been identified within the Mearns Drift Group to include the known glacial, glaciofluvial and glaciolacustrine deposits. They are the Mill of Forest Till, Drumlithie Sand and Gravel, and Ury Silts formations, respectively. The deposits are the product of the Main Late Devensian glaciation. Locally, some older units may be present at depth, but unless some intervening materials of a different origin and colour occur, it is difficult to distinguish them. No examples of older units of red till are known, but two distinct phases of movement of the Strathmore ice stream have been deduced by Armstrong et al. (1985) from the Dundee district, an earlier movement directed towards the south-east and a later one towards the east-north-east. This reorientation of ice flow is presumed to have occurred during the Main Late Devensian glaciation, either as the result of an expansion of the Southern Upland ice cap as the glaciation progressed (Sutherland, 1984a), or as a response to Scandinavian ice pushing into the North Sea basin. The presence of red till at the mouth of most of the valleys that drain the southern flank of the eastern Grampian Highlands, such as Glen Clova and Glen Esk (Bremner, 1934b, 1936; Synge, 1956), indicate that in general the Strathmore ice either forcibly pushed back ice occupying the valleys to the north, or that the ice within them had already begun to retreat enabling the Strathmore ice to advance towards the mountains. A more complicated sequence of events appears to have occurred at the Balnakettle site, near Fettercairn (Appendix 1).

There are several notable occurrences of dark bluish grey shelly diamicton (Benholm Clay Formation) underlying red till along the coast to the south of Aberdeen (Campbell, 1934; Auton et al., 2000), for example at the Burn of Benholm site (Appendix 1). It is interpreted as having been deposited by ice moving onshore from the North Sea basin, and hence the Benholm Clay has been assigned tentatively to the Logie-Buchan Drift Group (see below).

Mill of Forest Till Formation

Reddish brown tills laid down by the Strathmore ice stream are assigned to the Mill of Forest Till Formation. These typically cohesive, silty and clayey diamictons contain a mixture of clasts derived from Old Red Sandstone strata and Devonian volcanic rocks, together with some clasts of granitic and Dalradian metasedimentary rocks. Well-rounded clasts of igneous and metamorphic rocks are mostly recycled from Old Red Sandstone conglomerates, whereas more angular clasts are derived from outcrops of Dalradian metamorphic and Caledonian igneous rocks forming the adjacent Grampian Highlands. Some of this ‘Highland’ material may have also been recycled from pre-existing glacigenic sequences during the Main Late Devensian glaciation.

The type section of the Mill of Forest Till Formation occurs in a river cliff [NO 8630 8538], 150 m downstream of Mill of Forest Farm, Stonehaven, where 6 to 7 m of stiff, reddish brown, matrix-supported, silty sandy diamicton crops out beneath 3 m of cobble gravel of the Drumlithie Sand and Gravel Formation (Map 11). The till contains rounded pebbles and cobbles of quartzite, psammite, andesite and feldspathic microgranite, derived mostly from adjacent outcrops of conglomerate (Plate 8b). During the revision survey of Sheet 67 in 1926, Campbell observed a metre-thick unit of grey till at the base of the river cliff.

The Mill of Forest Till Formation is generally less than 5 m thick. Grossly overconsolidated lodgement tills are commonly developed in its basal part whereas friable, sandy to gravelly diamictons (flow tills) up to about 2 m thick commonly occur in its upper part, and locally interdigitate with the overlying Drumlithie Sand and Gravel Formation. One such flow till was recorded in a reference section (BGS trial pit NO77NW/1) near Drumelzie [NO 711 790], on Sheet 66E. In another reference section on that sheet (BGS trial pit NO77NE/11), at East Mondynes [NO 780 797], a flow till rests on red-brown lodgement till (Auton et al., 1990).

Drumlithie Sand and Gravel Formation

Deposits of the Drumlithie Sand and Gravel Formation were laid down principally as coarsening-upward sequences of sand and gravel at, or near, the margin of the actively retreating Strathmore ice stream, which impinged locally on the coast between the mouth of the River Dee and Stonehaven. Mounds, plateaux and ridges were formed at the retreating ice margin as fans and deltas that were subjected to minor re-advances as the ice withdrew. Eskers are common. Deposits commonly display evidence of slumping, small-scale normal faulting and cryoturbation, and are capped by up to a metre of red-brown sandy flow till.

The sands and gravels contain a mixture of clasts derived from the Silurian and Devonian clastic, volcaniclastic and volcanic rocks of Strathmore together with metamorphic and granitic clasts from the Grampian Highlands. Some of the clasts of crystalline rock have been recycled from Old Red Sandstone conglomerates. For example, the most characteristic durable clasts are well-rounded boulders and pebbles of quartzite from conglomerates. Clasts of soft red-brown sandstone, siltstone, mudstone and decomposed purple andesite are also common, but their relative proportions decrease rapidly away from the Old Red Sandstone outcrops from which they were derived. In its type area, between Auchenblae and Temple of Fiddes [NO 817 818], on Sheet 66E, the Drumlithie Sand and Gravel is typically 5 to 10 m thick. Thickly bedded sands with lenses of gravel are exposed in the partial type section in the Meikle Fiddes esker at Kaim of Clearymuir [NO 798 815] (Map 10).

Ury Silts Formation

The glaciolacustrine Ury Silts Formation consists of characteristic red-brown interlaminated micaceous fine-grained sand, silt and clay. These sediments were laid down in ice-marginal lakes that developed as the Strathmore ice stream retreated south-westwards. Exposures in the Ury Silts Formation are sparse. A partial type section is taken from 14.5 to 18.5 m depth (unbottomed) in BGS Borehole NO88NE/4, drilled near the Houff of Ury (Auton et al., 1988), on Sheet 67 (Map 11). This reddish brown silty deposit, which contains sparse isolated pebbles (interpreted as dropstones), is overlain by 1.7 m of moderate reddish brown, friable silty diamicton (flow till). The flow till is overlain by 12.4 m of Drumlithie Sand and Gravel Formation. More typically, there is a gradational contact between the silt and clay unit and the overlying sand and gravel, as in a reference section taken in BGS Borehole NO88NW/13, drilled in Craigies Wood [NO 8389 8514], near Stonehaven. The base of the Ury Silts Formation is generally in sharp contact with underlying diamictons of the Mill of Forest Till Formation, or with bedrock.

Relationship between Mearns and East Grampian Drift Groups

The Mill of Forest Till Formation generally rests directly on bedrock, but in some coastal localities, notably at Nigg Bay (Appendix 1), Cove (Jamieson, 1882b) and Findon (Synge, 1963) red till rests on grey till (Map 9), (Map 11). Both Jamieson and Synge observed gradational contacts between the tills and concluded that they were laid down during the same (Late Devensian) glaciation. The relationship suggests that the direction of ice flow changed from east-north-east to north-northeast as the Strathmore ice stream became dominant. The grey tills contain clasts derived from the west and are correlated here with the Banchory Till Formation of the East Grampian Drift Group (Table 7). At the Burn of Benholm site (Appendix 1), the Mill of Forest Till Formation rests with a sheared (glacitectonic) contact on the Benholm Clay Formation of the Logie-Buchan Drift Group.

Temporary sections at Ury Home Farm [NO 858 881], on Sheet 67, revealed up to 2.0 m of red-brown clayey lodgement till (Mill of Forest Till Formation) overlying up to 2.2 m of moderate yellowish brown sandy bouldery diamicton with clasts of Dalradian metasedimentary rocks and pink granite (Banchory Till Formation). The relationship between the tills suggests that East Grampian ice advanced into the lower reaches of the valley of the Cowie Water prior to advance of the Strathmore ice (Figure 4). In contrast, at Balnakettle (Map 10), near Fettercairn (Appendix 1) and Cantlayhills [NO 874 905], north of Stonehaven (Map 11), sandy diamictons assigned to the Banchory Till Formation overlie the Mill of Forest Till Formation. At both sites, the upper unit is interpreted as flow till, indicating that retreat of the East Grampian ice sheet and Strathmore ice stream was contemporaneous.

Deposits of the Drumlithie Sand and Gravel Formation are locally concealed beneath thin diamictons of the East Grampian Drift Group along parts of the Highland boundary. For example, at Balnakettle (Appendix 1) and Cantlayhills, sheared and contorted beds of reddish brown sand and gravel occur beneath yellow-brown sandy diamicton. At Cantlayhills, the diamicton was formed either from a debris flow directly from the East Grampian ice sheet, or as a landslide following deglaciation. In contrast, at Balnakettle (Plate 26) the sand and gravel has been disturbed by a local re-advance of the East Grampian ice sheet following the withdrawal of the Strathmore ice stream in the area (Appendix 1).

Logie-Buchan Drift Group

It has been widely believed for over a century that during the last (Main Late Devensian) glaciation ice flowing from Strathmore turned northwards between Stonehaven and Aberdeen and then north-westwards on to the coast of Logie-Buchan, between Aberdeen and Peterhead ((Figure 4); Jamieson, 1906; Bremner, 1916; Synge, 1956; Clapperton and Sugden, 1977; Merritt, 1981; Hall, 1984; Munro, 1986; Hall and Connell, 1991). Glacial striae provide evidence of the northerly ice flow (Jamieson, 1882a, 1906; Bremner, 1916; (Map 7)). The onshore movement is confirmed by the clay mineralogy of the red deposits (Glentworth et al., 1964) together with the unique suite of glacial erratics derived from the floor of the North Sea basin immediately off the coast of Aberdeenshire. The cause of the onshore movement is controversial, but the simplest explanation is that it was constrained to do so by Scandinavian–North Sea ice lying offshore (Chapter 5). The diversion led to the deposition of a distinctive suite of typically vivid reddish brown, interbedded materials that form the hummocky topography and prime agricultural land to the north of Aberdeen on Sheet 77 Aberdeen, Sheet 87W Ellon and Sheet 87E Peterhead. The Logie-Buchan Drift Group has been established here to include these glacigenic deposits. It correlates in part with the seismostratigraphical Wee Bankie Formation (Stoker et al., 1985; Gatliff et al., 1994), which extends for some 25 to 40 km offshore, where it is bounded to the south of Stonehaven by the prominent Wee Bankie Moraine (Figure 44).

The Logie-Buchan Drift Group (Figure 50) is composed mainly of a complex, interbedded sequence of clayey diamicton, clay, silt, mud, sand and gravel in which individual units can rarely be traced laterally for more than a few tens of metres. Most deposits are vivid reddish brown in colour and calcareous. The sands are typically fine to medium grained, silty and micaceous. Laminae and thin beds of yellowish brown, medium-grained sand and fine gravel commonly contain shell fragments. In addition to locally occurring rock types such as amphibolite, feldspathic psammite, quartzite and meta-greywacke, the deposits contain appreciable proportions of rocks derived from the sea bed to the east, including limestone, dolomite, calcareous siltstone, white and red sandstone of Devonian, Permo–Triassic, possible Jurassic, Cretaceous and Tertiary age (Figure 2). The group underlies the coastal lowlands to the north of Aberdeen, east of Ellon and both south and north of Peterhead, where it forms distinctive, fresh-looking, hummocky topography comprising kettleholes, mounds, plateaux, esker ridges and narrow, winding, steep-sided valleys.

The deposits of the Logie-Buchan Drift Group were first described by Jamieson (1858, 1882a, 1906) as part of his Red Clay Series. They have been described as ‘red drift’ by Merritt (1981) and the Red Series by Hall (1984) and Hall and Connell (1991). The type area is the parish of Slains, between Collieston and Cruden Bay, on Sheet 87E (Map 7). Typical sediments include silty and sandy diamictons that appear to have formed as flow tills and subaqueous debris flows, and muds formed in a glaciolacustrine, or possibly glacio-estuarine environment. These deposits, interbedded with glaciofluvial sands, fill hollows in the bedrock surface and collectively reach over 25 m in thickness (Merritt, 1981). Topography bears little relationship to the subdrift surface. The sequence is locally dominated by stiff, stony clayey diamicton, especially over bedrock ‘highs’. The uppermost metre or so generally consists of firm to stiff, pebbly, silty, clayey diamicton that has probably been formed by solifluction and cryogenic mixing. Thick diamicton-dominated sequences probably include deformation tills, as at Sandford Bay and Errollston (Appendix 1). Laminated silts and clays overlying buff-coloured, pebbly sands with shell fragments commonly occur at the base of the sequence within hollows. Mappable units of glaciofluvial sand and gravel, till and glaciolacustrine deposits are assigned to the Hatton Till, Kippet Hills Sand and Gravel and Ugie Clay formations, respectively.

Hatton Till Formation

Hall and Jarvis (1995) describe several sites around Bellscamphie where a stiff, reddish brown, relatively stony diamicton that they named the Hatton Till rests on the dark grey Pitlurg Till (of the Banffshire Coast Drift Group). Elsewhere to the north and east of Ellon similar red tills rest on grey and yellowish brown sandy tills of the East Grampian Drift Group. The Hatton Till, which is raised to formational status here, is typical of the more stony red diamictons in the Logie-Buchan Drift Group described above (Merritt, 1981). Its type section is at Bellscamphie (Appendix 1).

Thick units of reddish brown deformation till equivalent to the Hatton Till Formation are exposed at Sandford Bay, near Peterhead, where they have been derived from the south or south-east, and locally incorporate rafts of blue-grey diamicton (Appendix 1). To the north of Peterhead, red tills interdigitate with the Essie Till Formation of the Banffshire Coast Drift Group (see above). In contrast, red tills are not known to interdigitate with yellowish brown, sandy tills of the East Grampian Drift Group to the north of the River Ythan. Furthermore, at Cross Stone, to the south of Ellon, the Hatton Till Formation abuts two morainic ridges indicating that Logie-Buchan ice was free to expand into the Ythan valley following the retreat of East Grampian ice (see below).

Kippet Hills Sand and Gravel Formation

In addition to the thin, laterally impersistent beds of sand and gravel that form part of the sequence, several larger bodies have been identified on Sheet 87E. For example, a sinuous ridge, the Kippet Hills Esker, extends northwards from Cotehill Loch [NK 028 294], near Collieston, past Meikle Loch to Ladies’s Brig [NK 029 318], where it widens north-eastwards into a flat-topped, fan-shaped mound at Whitehills ((Figure 50); Appendix 1 Kippet Hills). At the partial type section near Knapsleask [NK 0327 3206], the gravel is distinctive in that it contains 40 per cent or more of calcareous material including Palaeozoic dolomite, Mesozoic limestone and calcareous siltstone, Pliocene shelly sandstone (Crag) and shell fragments of early Quaternary and younger age derived from the offshore Aberdeen Ground Formation. Hall and Jarvis (1995) referred to this distinctive glaciofluvial deposit as the Kippet Hills Gravels and Sands and it is named here formally as the Kippet Hills Sand and Gravel Formation of the Logie-Buchan Drift Group (Table 7). Hall and Jarvis describe similar material in an abandoned railway cutting at Bellscamphie [NK 0184 3369], which is taken here as another partial type section (Appendix 1 Ellon).

Other deposits of sand and gravel

A distinct suite of pebbly sand deposits lies along the boundary between the Logie-Buchan and East Grampian drift groups to the south of the Buchan Ridge, on Sheet 87W Ellon and Sheet 87E Peterhead. In the vicinity of the Hill of Auchleuchries [NK 006 365] (Map 6), the moundy deposits are capped by yellowish brown till of presumed western derivation, whereas some of the mounds to the north and west of Hatton (Map 7) are capped by red till of the Logie-Buchan Drift Group (Merritt, 1981). The sands contain clasts of a variety of lithologies, but include conspicuous amounts of pink (possibly Peterhead) granite. Surprisingly, they contain very little flint and quartzite from the Buchan Gravels Formation, which crops out just 2 km to the north. The sands overlie dark grey till with shell fragments, which is correlated with the Pitlurg Till of the Banffshire Coast Drift Group at Bellscamphie (Appendix 1). This till unit appears to have been deposited by ice moving onshore from the east-north-east (Bremner, 1928; Hall and Jarvis, 1995). As the pink granite clasts in the pebbly sands are likely also to have been derived from this direction, these glaciofluvial deposits probably relate to the same body of ice that laid down the Pitlurg Till. This association is supported by the presence in the pebbly sands of seams of fissile, chocolate-brown to olive-grey clay (Merritt, 1981). Hall and Jarvis (1995) conclude that the Pitlurg Till was laid down in the Early or Middle Devensian (OIS 4–3) and it follows that the pebbly sands are probably of similar age. However, if the Pitlurg Till correlates with the Whitehills Glacigenic Formation, the sands would have been laid down early in OIS 2 (Table 7).

It is proposed here that the pebbly sands described above be named as the Auchleuchries Sand and Gravel Formation, the type section being BGS Borehole NK03NW/1 (Merritt, 1981) on the Hill of Auchleuchries [NK 0057 3649]. The Bellscamphie site (Appendix 1 (Figure A1.20)) provides a reference section. The deposits of the Auchleuchries Sand and Gravel Formation have been placed in the East Grampian Drift Group on Sheet 87W Ellon, because they do not contain a significant proportion of clasts derived from offshore. However, the distribution of the deposits at, or within, the western boundary of the Logie-Buchan Drift Group and the rich shell fauna that was recovered from the Hill of Auchleuchries by Jamieson (1882b, p.172) both suggest that the unit is better placed in the Logie-Buchan Drift Group.

Ugie Clay Formation

Following the withdrawal of the East Grampian ice sheet, the coastal ice that laid down the Logie-Buchan Drift Group dammed the lower reaches of the valleys of the Ythan and Ugie causing extensive lakes to form (Hall, 1984; Hall and Connell, 1991; (Figure 42)). The fine-grained deposits laid down in these lakes are assigned here to the Ugie Clay Formation. Laminated silts and clays of the formation, mostly reddish brown in colour, occur extensively beneath the floodplains and glaciofluvial terraces of the North and South Ugie waters on Sheet 87E (McMillan and Aitken, 1981) and beneath similar features in the valley of the River Ythan on Sheet 87E (Merritt, 1981; Chapter 6 Glaciolacustrine deposits). Lakes also formed in the Don valley upstream of the Mill of Dyce (Appendix 1), but as the sediments there are dominated by materials derived from the retreating East Grampian ice sheet, they have been assigned to the Glen Dye Silts Formation of the East Grampian Drift Group (see below).

The type section of the Ugie Clay Formation is a stream section [NK 0050 4732] near Baluss Bridge, south of Mintlaw, where over 2 m of reddish brown and dark grey plastic clay is thinly interbedded with yellowish brown sand, gravel and silt. A BGS Borehole NK04NW/2 drilled 190 m to the south-south-west of the section proved at least 4 m of the sequence overlying orange-brown, very sandy till, resting in turn on weathered psammitic bedrock. Till exposed nearby has a macrofabric indicating that it was deposited by the East Grampian ice sheet flowing eastwards (Hall and Connell, 1991). The Ugie Clay Formation is capped locally by terraced glaciofluvial sand and gravel. Organic muds sampled from temporary sections in the formation at Baluss Bridge itself, yielded rich Palaeozoic and Mesozoic palynomorph assemblages and unreliable radiocarbon dates (Appendix 1 Ugie valley). Similar laminated organic deposits occur at Errollston clay pit [NK 088 368], near Cruden Bay (Appendix 1).

Several deposits of red and brown silt and clay in the Aberdeen area are named here as the Tullos Clay Member and assigned tentatively to the Ugie Clay Formation (rather than the Ury Silts Formation). They include those worked in clay pits at Tipperty [NJ 970 268] (Bremner, 1943; Munro, 1986) and Blackdog [NJ 962 139] (Jamieson, 1906; Bremner, 1916; Peacock, 1975). They also occur within the valley between Torry and Tullos Hill in Aberdeen (Simpson, 1948), and have been exposed at the Nigg Bay site (Appendix 1). The deposits are locally fossiliferous and those occurring at lower levels may be glaciomarine (up to about +30 m OD) (Chapter 6 Raised marine deposits)

Relationship between Logie-Buchan and East Grampian Drift Groups

Apart from at the Hill of Auchleuchries (see above), deposits of the Logie-Buchan Drift Group generally occur at the top of the glacial sequence to the north of Aberdeen (Figure 50). This suggests that they were the youngest glacigenic deposits to be laid down in the area. The interdigitation of bluish grey and red units to the north of Peterhead indicates that the Moray Firth ice stream laid down deposits of the Banffshire Coast Drift Group contemporaneously (see below). To the north of Ellon, red tills generally overlie tills of the East Grampian Drift Group.

In the Aberdeen area, the stratigraphical relationship between deposits of the Logie-Buchan Drift Group and the brown and grey tills derived from the East Grampian ice sheet is more complex. In the north-eastern part of Sheet 77, red-brown till is generally about 1 m thick and underlain by grey/brown till resting on bedrock (Munro, 1986). Isolated, ill-defined rafts of red till, up to 1.5 m in diameter, occur within the grey-brown till beyond the western margin of the Logie-Buchan Drift Group near Ardo House [NJ 929 208] (Map 9). Near Tipperty [NJ 969 277], isolated patches of grey-brown till are incorporated in red-brown till, and vice versa. Where a discrete unit of red till overlies grey-brown till, as in a pipeline trench in the vicinity of Mill of Ardo (Munro, 1986, fig. 30 A-B), the boundary between the two is generally regular and sharp. Where there are more complex field relationships between the two till units, or where the grey-brown till overlies the red, contacts are commonly ill-defined and gradational. Similar stratigraphical relationships also occur in several BGS boreholes and trial pits north of Aberdeen (Auton and Crofts, 1986). In most instances (e.g. Borehole NJ91NE/14, near Blackdog Rifle Ranges), red-brown silts and clays, overlie greyish brown till, but rarely, in Borehole NJ81SW/2, near Little Clinterty, brown sandy till overlies red-brown clayey till.

The local evidence of interdigitation of red tills and grey tills of inland provenance led Clapperton and Sugden (1977) to conclude that the ice responsible for laying down the Logie-Buchan Drift Group was in contact with ice flowing westwards from the East Grampian ice sheet. More commonly, red tills overlie tills of the East Grampian Drift Group (Bremner, 1916; Auton and Crofts, 1986) suggesting that East Grampian ice expanded to the present position off the coast, if not beyond, before retreating sufficiently to allow the encroachment of ice from offshore (Hall and Connell, 1991). Clapperton and Sugden (1977) concluded that the early expansion of East Grampian ice was the result of fluctuating flow strength between ‘Cairngorm–Grampian’ and ‘Strathmore’ ice in the Late Devensian. In contrast, Connell and Hall (1987) concluded that the tills were laid down in separate glaciations, in the Early and Late Devensian, respectively. This was based mainly on evidence from the area north of the River Ythan, where diamictons of the two groups are not seen to interdigitate but locally are observed to be separated by the Ugie Clay Formation. However, as noted above to the north of Ellon, yellowish brown till, derived from inland, caps shelly deposits of the Auchleuchries Sand and Gravel Formation. This indicates the former close proximity of Logie-Buchan Drift Group and East Grampian Drift Group ice at this location, a situation that apparently has not been previously described north of the Ythan (see Jamieson, 1906, p.26–27; Bremner, 1916, p.337).

A pair of crescentic, asymmetrical end moraines have been identified during recent fieldwork on Sheet 87W in the vicinity of Cross Stone [NJ 954 278], near Ellon, at the western boundary of the Logie-Buchan Drift Group ((Figure 50); (Map 6)). These features indicate quite clearly that a lobe of ‘Logie-Buchan’ ice advanced directly onshore up the valley of the River Ythan following retreat of the East Grampian ice sheet. The latter retreated sufficiently to allow a substantial glacial lake to form in the Ythan valley upstream of Ellon (Merritt, 1981; Hall, 1984). The inland margin of the Logie-Buchan Drift Group can be traced north-eastwards towards the Buchan Ridge, where it is quite distinct in the vicinity of the Den of Boddam [NK 102 411]. It reaches about 80 m above OD at the Den before descending to about 50 m OD to the west of Peterhead (Map 7).

Relationship between Logie-Buchan and Banffshire Coast Drift Groups

Red tills overlie yellowish brown and grey tills derived from the west between Ellon and Peterhead (Wilson, 1886; Jamieson, 1906). Immediately to the south of the River Ugie, they overlie greenish grey till that was probably also laid down by East Grampian ice (McMillan and Aitken, 1981). West of Peterhead, at Downiehills [NK 089 471] (Map 7), Jamieson (1858) recorded the occurrence of large blocks of Peterhead Granite in reddish brown till together with possible shell fragments. Later in the same area, he noted the occurrence of dark blue clay intermingled with deposits of the ‘Red Series’ (Jamieson, 1906). The latter description refers to exposures in the former claypit at Ednie [NK 085 500], where the red deposits were seen to enclose irregular masses described as ‘dark grey till’ and ‘dark blue clay and sand’ (Wilson, 1886; BGS records). These observations have caused confusion. On one hand they have been taken to support the view that the deposits of the ‘Blue Grey’ and ‘Red’ series are intermingled and are of approximately the same age (Jamieson, 1906; Hall and Connell, 1991). The alternative view is that older ‘Blue Grey Series’ (Banffshire Coast Drift Group) sediments have been incorporated into the red glaciotectonically by a later, northward movement of ice (Sutherland, 1984a; see Appendix 1 Sandford Bay).

The situation is reversed to the north of the River Ugie, however, where red sediments are locally overlain by the dark grey, shelly Essie Till Formation, which contains erratics of pink granite, quartz and metasedimentary rocks. This diamicton, assigned to the Banffshire Coast Drift Group above, was probably laid down by a late re-advance of the Moray Firth ice stream. Red deposits of the Logie-Buchan Drift Group locally overlie blue-grey tills in the vicinity of Ellon and Aberdeen, but the age of the older tills is uncertain (see Banffshire Coast Drift Group above).

Relationship between the Logie-Buchan and Mearns Drift Groups and the Benholm Clay Formation

Deposits of the two groups do not appear to abut one another at the surface on land (Figure 4), and offshore no contact has been recognised within the Wee Bankie Formation. However, there are several notable occurrences of dark bluish grey, shelly diamicton underlying red till along the coast to the south of Aberdeen (Campbell, 1934). The best known locality is at the Burn of Benholm ((Map 11); Appendix 1). These shelly diamictons have been assigned to the Benholm Clay Formation (Auton et al., 2000) and they were almost certainly laid down by ice moving onshore from the North Sea basin during a pre-Devensian glaciation (Table 7). Despite its colour, the Benholm Clay Formation is placed tentatively in the Logie-Buchan Drift Group because of its derivation by ice from the North Sea basin (Plate 27).

East Grampian Drift Group

This group is broadly equivalent to the sediments that were previously referred informally to the ‘Inland Series’ (Hall, 1984a; Sutherland and Gordon, 1993). The deposits contain clasts that have been carried by ice flowing from the eastern Grampian Highlands during several glaciations (Figure 4). Although erratics from farther afield do occur (Figure 47), the colour and clast composition of the tills closely reflect the nature of underlying bedrock (commonly deeply weathered), or of rocks cropping out within a few kilometres to the west. The tills are generally sandy, thin (less than 2 m) and patchy, especially across central Buchan, where they are normally pale yellowish brown in colour. Thicker and more widespread tills occur in the valleys of the Dee and Don, where colour ranges from brown to grey. Glaciofluvial and glaciolacustrine deposits are relatively uncommon across large tracts of countryside, especially in central Buchan. The uppermost metre or so of all materials is commonly severely disturbed by periglacial activity (Connell and Hall, 1987).

As explained in Chapter 5, the sandy nature and pale colour of the tills occurring in central Buchan have led some authors to conclude that they have been weathered since their deposition. These attributes, together with inferred relationships to incorrectly dated organic sediments (Appendix 1 Crossbrae), led some to conclude that they were laid down either during an early Devensian glaciation (Hall, 1984; Sutherland, 1984a; Hall and Connell, 1991) or during a pre-Devensian one (Charlesworth, 1956; Synge, 1956, Fitzpatrick, 1958, 1972; Galloway, 1961a, b, c). The sandiness and colour are equally likely, however, to result from the incorporation of significant proportions of deeply weathered bedrock or previously weathered deposits, in which case the tills need be no older than Late Devensian in age (Clapperton and Sugden, 1977; Hall and Bent, 1990). For example, pedological studies of a till (possibly Hythie Till Formation) near Mintlaw suggest that the soil-forming processes occurred during the Holocene (Van Amerongen, 1976).

Glacigenic deposits of inland provenance are known to occur at several stratigraphical levels (Table 7) and there are possibly representatives of glaciations going back to OIS 8, if not the Anglian (Appendix 1 Kirkhill). The older deposits have been identified through the discovery of intervening units of peat, soil and gelifluctate, but in the absence of firm dating control, their chronostratigraphical context is not known for certain.

Most deposits of inland provenance that have been named in the literature occur at widely dispersed sites where relatively long sequences have been described, for example at the Teindland and Kirkhill localities (Appendix 1). Although formal lithostratigraphical units have been established locally at Boyne Bay and Gardenstown (Appendix 1), only on Sheet 66E Banchory and Sheet 67 Stonehaven have units of this group been mapped out systematically. On these sheets, the glaciofluvial, glaciolacustrine and glacial deposits associated with the former East Grampian ice sheet are placed in the Lochton Sand and Gravel, the Glen Dye Silts and the Banchory Till formations, respectively. These Late Devensian deposits extend northwards onto Sheets 76E and 77, if not farther north, and they form part of the important sequence established at Nigg Bay ((Table 7); Appendix 1).

Till

Banchory Till Formation

Glacial sediments assigned to the Late Devensian Banchory Till Formation are typically sandy, gravelly or bouldery diamictons containing clasts of Dalradian psammite, pelite and semipelite, and Caledonian granitic rocks (Plate 8a). Like most tills related to ice emanating from the East Grampian sheet, their composition closely reflects the nature of the underlying bedrock. For example, a lime-rich fertile soil has developed on the till on the flanks of Rhinbuckie Wood [NO 716 933], where calcareous strata crop out (Map 10). The Banchory Till is generally less than 5 m thick, free-draining and notably overconsolidated. Typically it is pale to yellowish brown, although darker colours occur beneath the water table, ranging from greyish brown to olive-grey. Red and brown colours occur where the tills are developed on decomposed (reddened) granite bedrock.

The type area of the Banchory Till Formation is the upland on the southern flank of the Dee valley, between Banchory and Strachan, on Sheet 66E. A reference section occurs [NO 607 920] at a 5 m-high river cliff of the Burn of Granney (a tributary of the Water of Feugh), just to the west of Sheet 66E. The till is sandy, moderate brown in colour and contains abundant angular clasts of the underlying granite, together with psammite and microgranite. A further reference section [NO 6073 9199] occurs in a small working 400 m westsouth-west of Finzean, also just to the west of the Sheet 66E. Bouldery subglacial diamictons of the Banchory Till Formation commonly occur at the base of the East Grampian Drift Group. More sandy and gravelly diamictons are locally interstratified with the overlying Lochton Sand and Gravel Formation and, in places, along the contact between the East Grampian and Mearns drift groups, they overlie the latter deposits (see above).

Three probably Late Devensian members of the Banchory Till Formation have been established in the Aberdeen area (Map 9), but only one is visible at the Nigg Bay site ((Table 7); Appendix 1). The Nigg Till Member is dark greyish brown, contains clasts derived locally from the west to west-south-west and is typical of surficial tills in the lower Dee valley and to the south. The brown Kingswells Till Member is probably a local variant, having a west-south-west to east-north-east fabric and containing much ‘Hill of Fare’ granite. It is equivalent to Till B of Murdoch (1977) and its type area is the Kingswells–Culter–Aberdeen city area. A third unit, the dark grey Den Burn Member with locally derived clasts underlies the above-mentioned till in the vicinity of the Kingswells Roundabout [NJ 878 061]. It has a north-west to south-east fabric and is equivalent to Till A of Murdoch (1977). It is possibly equivalent to the ‘grey boulder clay’ with clasts derived from the north-west, which Bremner (1934a, 1943) recorded as overlying ‘dark shelly boulder clay’ in excavations for the Anderson Drive Ring Road. The shelly unit is assigned here to the Banffshire Coast Drift Group (see above).

A fourth member of the Banchory Till Formation occurs in the Ellon–Bellscamphie area. This till, named here as the Bearnie Till Member, contains clasts of local provenance but is typically dark grey and contains a rich Jurassic palynoflora (Appendix 1 Site 15). The last two attributes are probably the result of glacial reworking of the older Pitlurg Till Formation (Banffshire Coast Drift Group). The Bearnie Till is overlain by the red, Hatton Till Formation ((Table 7); (Figure A1.20)).

Other units

Most surficial tills of the Late Devensian East Grampian ice sheet occurring to the north and west of Sheet 76E Inverurie and Sheet 77 Aberdeen have not yet been assigned to lithostratigraphical formations. However, the Hythie Till Formation (Kirkhill Upper Till) has been identified at several sites in central Buchan, including Oldmill (Appendix 1 Site 8). The Hythie Till probably correlates with the Byth Till Formation at the Howe of Byth site and the Crovie Till Formation at Gardenstown ((Table 7); Appendix 1 Sites 6, 3). The relationship of the Hythie Till Formation to the Aldie Till Formation at the Camp Fauld site is uncertain, but the latter unit may be of an earlier Devensian age ((Table 7); Appendix 1 Site 14 Moss of Cruden). The granite-rich till at the base of the Sandford Bay sequence is named here as the Sandford Bay Till Member of the Hythie Till Formation (Appendix 1).

Pre-Devensian tills derived from the west, north-west and possibly also north have been identified at several localities. They include the Rottenhill Till Formation (Kirkhill Lower Till) at Kirkhill assigned to OIS 6, together with two other units of possibly similar age, the Crossbrae Till Formation at Crossbrae and the Bellscamphie Till Formation in the Ellon area ((Table 7); Appendix 1 Site 15). The group also includes the oldest known till in the district, the Leys Till Formation of the Kirkhill sequence, which is probably at least as old as OIS 8 (Table 7). The Leys Till is also the oldest Pleistocene glacial deposit known onshore in Scotland.

Sands and gravels

Lochton Sand and Gravel Formation

Coarse-grained glaciofluvial deposits of the Lochton Sand and Gravel Formation are characterised by clasts of Dalradian metamorphic rocks and Caledonian granitic rocks, both of relatively local provenance. Sandier units are typically pale yellowish brown in colour. Coarse-, medium- and fine-grained quartzo-feldspathic sands are common, normally forming upward-coarsening units. The formation includes moundy (ice-contact) and terraced (sheet) deposits that both commonly pass downwards into laminated silt and clay of the Glen Dye Silts Formation. Sharp erosional contacts on to the underlying Banchory Till Formation are evident, especially at the base of terraced spreads.

The type locality of the Lochton Sand and Gravel Formation is Lochton Pit [NO 749 926] in the valley of the Burn of Sheeoch, on Sheet 66E (Auton et al., 1988; Brown, 1994) (Map 10). Here, a coarsening-upward deltaic sequence of sand and gravel up to 5.4 m thick overlies interlaminated clay and silt of the Glen Dye Silts Formation. Reference sections occur between 0.4 and 5.2 m depth in BGS Borehole NO69SE/15, sited on a kame terrace in the valley of the Water of Feugh, near Heugh Head, and in a trial pit (NO69SW/5), dug into an esker ridge south-west of Easter Clune (Auton et al., 1990).

The ‘lower sands and gravels’ of the Nigg Bay section contain clasts that are predominantly derived from the Dee valley to the west (Appendix 1). Importantly however, they contain very sparse Scandinavian erratics and flint (Chapter 7 Erratics; (Figure 47)). The unit is named here as the Ness Sand and Gravel Member of the Lochton Sand and Gravel Formation.

Other units

Several other units of sand and gravel have been assigned to the East Grampian Drift Group, most predating the Late Devensian (Table 7). The Pishlinn Burn Gravel Bed, which underlies the Whitehills Glacigenic Formation at Gardenstown, is tentatively placed within the group, although its age and origin is problematic (Peacock and Merritt, 1997; Appendix 1 Site 3). At the former Tillybrex gravel pit [NK 001 347], 6 km northeast of Ellon (Map 6), over 11 m of weathered gravels with no morphological expression are overlain by brown till of westerly derivation, capped in turn by red diamicton and clay of the Logie-Buchan Drift Group (Merritt, 1981). The gravels at Tillybrex are here named the Tillybrex Gravel Formation. Although no contact has been recorded, the gravels appear to occur stratigraphically below patchy deposits of dark grey till correlated with the Pitlurg Till Formation of the Banffshire Coast Drift Group (Hall and Jarvis, 1995; Appendix 1 Ellon). The gravels are underlain by yellowish brown sandy till containing clasts of psammite from the west, together with some well-rounded quartz, quartzite and flint pebbles derived from the Buchan Gravels Formation. This till, correlated with the pre-Devensian Bellscamphie Till by Hall and Jarvis (1995), overlies weathered psammite bedrock. Like the Bellscamphie Till, the Tillybrex Gravel most probably predates the Devensian (Table 7).

The westerly palaeocurrent of the Tillybrex Gravel Formation is broadly similar to the gravels at Oldmill [NK 023 440] and the Denend Gravel Formation (Leys Lower Gravel) of the Kirkhill sequence. It suggests that ice advanced into eastern Buchan during a pre-Devensian cold stage(s), perhaps as old as OIS 8 (Appendix 1 Oldmill, Kirkhill and Ellon). Other pre-Devensian units at Kirkhill assigned to the East Grampian Drift Group include the Pitscow Sand and Gravel Formation (Kirkhill Lower Sand and Gravel) and the Swineden Sand Bed ((Table 7); Appendix 1). The Birnie Gravel Formation, which underlies the Benholm Clay at the Burn of Benholm site (Appendix 1), is also assigned to the East Grampian Drift Group.

Glaciolacustrine deposits

Glen Dye Silts Formation

Thinly interlaminated sandy silts and clays, typically pale olive to greenish grey in colour, were laid down in ice-marginal lakes as the East Grampian ice sheet decayed and retreated in the Late Devensian. They are formally assigned here to the Glen Dye Silts Formation. In the type area of Glen Dye, on Sheet 66E (Map 10), these deposits occupy flat-lying ground near the confluence of the Mill Burn and the Water of Dye, at Miller’s Bog [NO 636 861]. The formation is exposed in several sections in the southern bank of the Mill Burn, 600 to 800 m upstream of its confluence with the main river. Here a unit, 0.8 to 3.0 m thick, of micaceous sandy silt and clay with well developed horizontal lamination overlies clayey diamicton of the Banchory Till Formation. The exposed silts and clays show pronounced iron staining and are typically mottled orange and olive-grey in colour.

Deposits of the Glen Dye Silts Formation typically rest with a sharp planar contact on the underlying till or bedrock, but they pass upwards gradationally into deposits of the Lochton Sand and Gravel Formation. This upward transition is recorded in a trial pit (NO68NW/2) sited on the floor of a small gravel working at Miller’s Bog and a trial pit at Lochton (NO79SE/1). The silts are locally interbedded with deposits of the Lochton Sand and Gravel Formation. As well as underlying the Lochton Sand and Gravel Formation in Glen Dye, the Glen Dye Silts Formation is widely developed beneath floodplains, alluvial flats and basin peat elsewhere on Sheet 66E Banchory. Laminated silts and clays are also common in similar settings on Sheet 76E Inverurie and Sheet 77 Aberdeen. They are particularly thick and extensive beneath the floodplain and glaciofluvial terraces of the River Don, between Dyce and Inverurie (Map 8), and beneath similar features within the valley of the River Urie upstream of Inverurie (Auton and Crofts, 1986; Auton et al., 1988; see also Appendix 1 Mill of Dyce and Nigg Bay).

Central Grampian Drift Group

Deposits of the Central Grampian Drift Group occur on Sheet 95 Elgin and Sheet 96W Portsoy. They were laid down by ice that radiated outwards from a centre over Rannoch Moor in the western Highlands, carrying rock fragments from the Central Highland Migmatite Complex and Caledonian igneous rocks as well as the local, dominantly psammitic Grampian Group rocks (Figure 45); (Figure 47). The ice that entered the district did so by two routes (Figure 4). Some ice flowed down the Spey valley and merged with ice from the East Grampians. Other ice flowed northwards towards the Moray Firth where it abutted, and merged with a more powerful stream emanating from the north-west Highlands. These combined ice streams flowed into the Moray Firth and along its southern shores. The relative power of the merging ice streams varied through time, resulting in the interdigitation of deposits of the three drift groups. As in the other groups, some formations have been identified that predate the Devensian.

The interplay between ice streams has resulted in complicated glacigenic sequences in the Inverness area, on Sheet 84W (Fletcher et al., 1996). There, psammite-rich tills laid down by central Grampian ice overlie tills dominated by Old Red Sandstone lithologies. The latter were laid down by ice that flowed from the Central Highlands via the Great Glen towards Inverness, where it crossed Old Red Sandstone strata. The more powerful flow of ice from the north-west Highlands forced the ice from the Great Glen eastwards across high ground towards the Elgin area, where it laid down sandstone-rich tills and associated glaciofluvial deposits. Some Mesozoic erratics occur in the surficial tills around Burghead and Lossiemouth, indicating that some ice crossed the southern margin of the Moray Firth (Figure 45). Deposits have been assigned to the Central Grampian Drift Group on Sheet 96W Portsoy, but Sheet 95 Elgin and Sheet 84W Fortrose were published before the group was formally established.

Till

These tills have been described in general terms in Chapter 6. No formations have been mapped out individually, but a local stratigraphy has been established in the vicinity of the Teindland site (Hall et al., 1995; (Table 7); Appendix 1). In the Elgin area, brown sandy tills of the Central Grampian Drift Group contain many well-rounded clasts derived from Old Red Sandstone conglomerates cropping out to the west. They overlie dark grey clayey tills containing shell fragments and Mesozoic erratics derived from the bed of the Moray Firth to the north-west (Peacock et al., 1968; Aitken et al., 1979). Hall et al. (1995a) name the former unit the Tofthead Till and the latter the Alton-side Till (Banffshire Coast Drift Group). They also recognise a younger sandy diamicton, the Waterworks Till. They conclude that all three tills, together with locally intervening units of sand and gravel, were laid down in different phases of the Main Late Devensian glaciation, the Waterworks Till being the result of a minor re-advance. The Tofthead Till probably correlates with the Old Hythe Till at the Boyne Limestone Quarry (Peacock and Merritt, 2000).

None of the three Late Devensian tills occurring in the vicinity of Teindland appears to be present at the main site. Instead, another sandy till unit occurs that Hall et al. (1995a) name as the Teindland Till and assign to the Early Devensian. It overlies a package of units that includes the Teindland Buried Soil, which probably dates from the Ipswichian (OIS 5e). The soil lies stratigraphically above a unit that Hall et al. (1995a) name as the Teindland Gravel. It is composed predominantly of rounded clasts of quartzite and psammite that are probably largely derived from Old Red Sandstone conglomerates. Nearby, it overlies the Red Burn Till, the oldest known till in this area (Appendix 1). Although reddish brown in colour and containing clasts mostly of quartzite and psammite, this diamicton, like the Alton-side Till, contains some Mesozoic erratics and hence it is also assigned here to the Banffshire Coast Drift Group, albeit tentatively.

Another old till unit assigned here to the Central Grampian Drift Group occurs at the Boyne Limestone Quarry (Appendix 1). This very sandy diamicton, the Craig of Boyne Till Formation, contains much decomposed, easily weathered calc-silicate rock and therefore appears to be weathered to a greater extent than it is (Peacock and Merritt, 2000). Nonetheless, it contains some clasts that have been considerably weathered in situ and is therefore likely to be pre-Devensian in age, like the Red Burn Till at Teindland.

Other deposits

Although the sand and gravel occurring on Sheet 95 Elgin was mainly laid down at the margin of the Moray Firth ice stream as it retreated (Chapter 6), its composition suggests that much of the ice crossing the area was probably sourced in the central Highlands. Therefore, these glaciofluvial deposits should be assigned to the Central Grampian Drift Group, but no formal lithostratigraphical units have yet been set up. The only current named unit is the pre-Devensian Teindland Gravel.

Information sources

Further geological information held by the British Geological Survey relevant to the Quaternary geology of north-east Scotland is listed below. It includes published maps, memoirs and reports and open-file maps and reports. Other sources include borehole records, site investigation reports, hydrogeological, geochemical and geophysical data and photographs.

Searches of indexes to some collections relevant to the Quaternary of the district can be made on the Geoscience Data Index system in British Geological Survey libraries. This is a developing computer-based system that carries out searches of indexes to digital and nondigital databases for specified geographical areas. It is based on a geographical information system (GIS) linked to a relational database management system. Results of the searches are displayed on maps on screen. At the present time (2003) the databases indexes that are particularly applicable to the Quaternary geology of northeast Scotland are:

Old Hythe Till Formation (Central Grampian Drift Group)

This unit corresponds to the upper till (Unit 4) of Peacock (1966) and is up to 5 m thick. It is a matrix-supported diamiction, which is very stiff, massive to crudely laminated (sheared), inhomogeneous, silty, stony clayey and with local sandy lenses. It is dark grey to dark yellowish brown (10 YR 4/2), but locally dark greenish grey (5GY 4/1 to 5G 4/1). Clasts are generally angular to well rounded, mainly less than 35 cm, but locally up to 1.5 m across near the top. Apart from the boulders, which are mostly gabbro, the clasts are of psammite and quartzite (locally dominant), with some pink granite, red-brown and brown sandstone, pelite, calc-silicate rock and amphibolite. There are also sparse boulders of troctolite. Sparse shell fragments occur throughout at the northern end of the quarry, but are apparently absent at the southern end. The till is weathered to a depth of 1.5 to 2.0 m at the top where it is decalcified, mottled and contains partially decomposed clasts of gabbro and calc-silicate rock. Vertical fissures are prominent in the weathered zone.

Site description

The total length of the exposure (Figure A1.5) is some 300 m; the thickness of superficial deposits diminishes from about 12 m at the north-east end of the quarry, south-westwards to between 5 and 6 m. The basal Craig of Boyne Till is found throughout, whereas the overlying Whitehills Glacigenic Formation is absent from part of the southern face. Its contact with the Craig of Boyne Till Formation is locally sharp, but more commonly gradational over about 0.3 m, being marked in places by concentrations of bleached pebbles of quartzite and psammite derived from the Craig of Boyne Till Formation (Peacock and Merritt, 2000b, fig.10). The lithology suggests that the underlying bedrock was weathered prior to deposition of the Craig of Boyne Till Formation, which was itself weathered before emplacement of the Whitehills Glacigenic Formation. The Whitehills Glacigenic Formation includes shell fragments, particularly at the north-eastern end of the quarry, and examination of the clay has revealed Mesozoic palynomorphs (Callovian to possibly Portlandian/Volgian) and small fragments of lignite.

A section 10 m from the north-eastern end of the quarry (Figure A1.5) locality 1 is illustrated in (Figure A1.6), where the glacigenic sequence below the upper till, the Old Hythe Till Formation, is made up of six units (A to F). Here, apparently undisturbed Craig of Boyne Till Formation (A) is separated by a gently undulating, concavo-convex, subhorizontal sand-lined discontinuity (T1) from an overlying leaf of till (B). The latter is similar to till (A), but is slightly less weathered, and is coloured brown to greyish red. It is overlain southwards by a deformed wedge of moderate brown brecciated clay (C). Units B and C are separated from moderate yellowish brown and orange mottled sandy clay (D) by a sharp, planar, sheared contact (T2). This contact, together with another (T3), have been folded locally, and clay (D) and olive black clay (E) squeezed upwards into the overlying Old Hythe Till Formation (unit G). Clays of units D and E both contain sparse rounded to well-rounded pebbles, including red sandstone (possibly Triassic). Clay E also includes comminuted shell fragments. Units A to E reach about 5 m in thickness, and are overlain by some 5 m of the Old Hythe Till Formation. Unit F is matrix-supported sandy gravel with shell fragments and laminae of dark clay, possibly glaciotectonically intercalated.

At locality 6, some 210 and 240 m from the north-eastern corner of the quarry, a 5 m-thick body of sand was recorded in June 1997. It lay above dark clay (unit E of locality 1) to reach the present-day soil over part of the exposure, wedging out at the northern end below the upper till of locality 1 (unit G). In October 1997, however, after the face had been taken back, it appeared to be entirely overlain by till of the Old Hythe Till Formation. Towards the northern end of the sand body, the boundary of the sand with the underlying dark clay (E) was folded by steep, almost vertically plunging folds, and the sand body itself frayed out northwards immediately below the Old Hythe Till Formation in a zone of shearing. At its southern end, the sand body merges into streaks and bands of sand within unit E, with complex folds and faults. These streaks and bands are locally crossed by numerous subhorizontal shears to give the appearance of brittle fracture cleavage (Plate 17b). The faulting and shearing at the southern end of the sand body has also led to the occurrence of angular blocks of clay up to 20 cm across surrounded by sand, a feature that is also found farther south along the face near the south-western corner of the quarry. The upper part of the sand body locally retains trough cross-bedding in sets some 0.5 m thick, but the sedimentary structures are generally obscured or destroyed by small-scale normal and reversed faults and shears (including some conjugate systems), subhorizontal shears, and shears dipping 25° to 30° to the west or north-west (Peacock and Merritt, 2000b, fig.14).

Details of other localities sited on (Figure A1.5) are given in Peacock and Merritt (2000a) and Merritt et al. (2000).

Genesis and age

The clasts in the Craig of Boyne Till, including numerous gabbro erratics and well-rounded clasts of quartzite and psammite derived from outcrops of gabbro and Middle Old Red Sandstone a few kilometres west of the quarry, imply a former ice movement to the east or north-east. The weathering implies that the till probably predates an interglacial period, either the last (Ipswichian), or an older one.

The lithology of the Whitehills Glacigenic Formation suggests derivation particularly from Mesozoic strata that crop out on the floor of the Moray Firth (Andrews et al., 1990), as well as Quaternary marine deposits. This is supported by structural evidence for ice movement from a northerly direction, such as folds overturned to the south, and a number of shears dipping gently between north-east and north-west. North–south or south–north directed striae were seen by J D Peacock and F M Synge (unpublished) in 1971 on a ‘high’ of fresh bedrock where the basal weathered till was absent, and they probably relate to this ice movement. Amino-acid epimerisation ratios from hinges of Arctica islandica at Gardenstown and King Edward (Miller et al., 1987; Peacock and Merritt, 1997) suggest that the Whitehills Glacigenic Formation was emplaced after early OIS 3 (about 60–45 ka cal BP), possibly in the Late Devensian. The Whitehills Glacigenic Formation thus seems to have been emplaced by a glacier flowing eastwards along the Moray Firth (the Moray Firth glacier), with its southern flank extending south or south-east on to inland parts of coastal Banffshire.

The derivation of the Old Hythe Till based on its content of erratics is ambiguous since it includes material derived from seawards (shell fragments) as well as from the south (troctolite) and west or south-west (e.g. well-rounded pebbles of quartzite). It is suggested that the Old Hythe Till Formation at the Boyne Limestone Quarry was deposited by ‘inland’ ice that had been directed to flow south-eastwards by the still-active Moray Firth ice stream (Peacock and Merritt, 2000a).

Site 3 Castle Hill, Gardenstown

At this key site, glaciolacustrine silts and sands of the Kirk Burn Silt Formation (formerly ‘Coastal Deposits’) overlie a glacigenic sequence that comprises glacio-tectonites (Whitehills Glacigenic Formation) overlain by a thin till (Crovie Till Formation) and underlain by gravel (Figure A1.7). The glacitectonites, which were apparently emplaced by ice moving from seawards, include rafts of marine sand in which sedimentary structures are preserved, the rafts being separated by shears. Amino-acid epimerisation ratios from specimens of Arctica islandica collected from one raft suggest that the emplacement of the glacitectonites post-dated the early part of Oxygen Isotope Stage 3 (about 60–45 ka BP).

Castle Hill [NJ 795 642] has long been a focus of geological interest, particularly as it yielded cold-water marine molluscan remains some 25 to 50 m above sea level (Prestwich, 1837; Chambers, 1857; Miller, 1859; Jamieson, 1906; Read, 1923). More recent work (Peacock, 1971; Sutherland, 1984b, 1993b) has shown that the upper part of the succession comprises the sediments of a glacial lake that possibly drained southwards through the Afforsk channel [NJ 788 630], on Sheet 96E (Map 3). Following re-examination, the lower part of the succession is now recognised as dominantly glacigenic (Peacock and Merritt, 1997).

The succession of superficial deposits on part of the north face of Castle Hill is shown in (Table A1.3) after Peacock and Merritt (1997).

Pishlinn Burn Gravel Bed

The Pishlinn Burn Gravel Bed (Figure A1.7), sections S1, S2 and localities D, E, F1 and F2 is a poorly sorted, matrix- to clast-supported gravel consisting predominantly of angular to subangular fragments of Dalradian slaty, turbiditic arenite and semipelite, with sparse well-rounded pebbles of red sandstone. It is interpreted as a water laid or mass movement deposit in which the Dalradian clasts have been reworked from former talus slopes to the west and redeposited by glacial or glaciofluvial action, possibly including rafting.

Whitehills Glacigenic Formation

The Whitehills Glacigenic Formation at sections S1 and S2 contains a basal unit of sand a little over 2 m thick with dispersed pebbles and subsidiary lenses and seams of clay ((Figure A1.8) unit 4). Unit 4 is characterised by pervasive shearing and attenuation of both sandy and clayey lithologies, with the truncation of folds by a glacitectonic lamination in sand, and the formation of anastomosing seams of diamicton due to subhorizontal shearing and rafting. A raft of the Pishlinn Burn Gravel occurs in section S2. These are the features of a ‘Type A’ glacitectonite in which the passage of overriding ice has resulted in penetrative deformation and high strain (compare with Benn and Evans, 1996).

The overlying units in Section S1 (Figure A1.8) are thickly interbedded sand and silty clay, bounded in places by sand-lined shears. Some of the sand beds retain sedimentary structures in the form of ripple lamination, and unit 7 is a fining-upward succession in which a pebble lag at the base is overlain by 10 to 20 cm of coarse sand with rounded shell fragments. Strata similar to units 5 to 7 apparently occur in the poorly exposed ground between the top of unit 7 and the base of the Crovie Till Formation, and there is another probable lag deposit of gravelly sand with shell fragments at Locality C1 (Figure A1.7). This has yielded a molluscan fauna of boreo-arctic aspect including the bivalve Arctica islandica and the temperate foraminifer Elphidium crispum. The characteristics of the sediments from unit 5 to near the base of the Crovie Till Formation (shears bounding some of the units, but the retention of sedimentary structures) suggests that these are paraautochthonous rafts of marine strata in which deformation is nonpenetrative (Type B glacitectonites of Benn and Evans, 1996). However, there is a thin (0.6 m) band of dark diamicton immediately below the Crovie Till Formation, which suggests a return to conditions of high strain.

Crovie Till Formation

The reddish brown till of the Crovie Till Formation itself is a gravelly sandy clay diamicton, up to 0.8 m thick, with clasts of Dalradian rocks and sparse, well-rounded pebbles probably derived from Old Red Sandstone conglomerate. It is only present locally, and where absent, the Whitehills Glacigenic Formation is directly overlain by the Kirk Burn Silt Formation.

Kirk Burn Silt Formation

The Kirk Burn Silt Formation on Castle Hill comprises up to 30 m of nonfossiliferous deposits, the lower beds being predominantly fine-grained sand, the upper being silt, clay and fine-grained sand. Some of the sand is coarsely laminated and interbedded on a scale of 0.1 to 0.3 m (Peacock, 1971; Sutherland, 1984b; Merritt and Peacock, 1997). Ferruginous concretions occur in places. Ripple bedding, slump folds and dewatering structures have been seen near the summit of the hill, and Sutherland (1984b) reported cross-lamination in the lower sands with foresets dipping south-eastwards. He also reported that a specimen from near the base of the deposit contained 14 per cent organic matter and a sparse pollen assemblage dominated by grains of Pinus. Peacock (1971) suggested that the deposits of the Kirk Burn Silt Formation here, and along a long stretch of the Banffshire coast, were deposited in a glacial lake that was dammed by ice in the Moray Firth, and which drained southwards through the Afforsk meltwater channel ((Figure 42), (Map 3)). This channel, however, is one of a suite, some of which are now known to predate the deposition of the Whitehills Glacigenic Formation (see Site 4 King Edward). The possibility must therefore be considered that the lakes were held up by ice to the south as well as in the Moray Firth, and that they drained generally eastwards to the North Sea.

Site 4 King Edward

King Edward [NJ 722 561], midway between Turriff and Banff (Map 5), is a long recognised locality for the presence there of a Quaternary marine shell bed containing whole arctic marine shells beneath dark grey shelly till (Jamieson, 1866; Sutherland, 1984c). Opinion has been divided over the origin of the shell bed. The state of preservation of the shells and the extent of the shell bed led to the view that this was an in situ marine deposit (Jamieson, 1866; Sutherland, 1981). Alternatively, the intimate association of the shell bed with shelly till led others to consider it to be a glacial raft or rafts (Read, 1923; Peacock and Merritt, 1997).

Jamieson gave details of a section beside the Burn of King Edward about 100 m south-west of the old bridge over the Banff–Turriff road in a series of papers spanning almost 50 years (Jamieson, 1858, 1865, 1866, 1906). Today the section is obscured and forested. The upper part comprised up to 8 m of coarse glaciofluvial gravel penetrated by ice-wedge casts. It cropped out along the sides of the burn and its tributaries below high terraces (Read, 1923; Sutherland, 1984c). Below the gravel lay up to 9 m of dark grey pebbly mud, with striated shells towards its base. This diamicton, here named the Castleton Memberof the Whitehills Glacigenic Formation (Castleton Formation of Sutherland, 1999), is typical of the shelly till found widely in the King Edward area (Read, 1923). The base of the section revealed a thin (60 cm thick) layer of brown shelly sand interstratified with more than 3 m of stoneless dark grey silt. The silt contained arctic shells in a crushed and decayed state, but apparently in situ. Jamieson regarded the lowermost shelly silt as representing a marine submergence under arctic conditions that occurred prior to glaciation of the area.

Recent excavation of a river bank 200 m south-east of the original locality [NJ 7236 5604] has confirmed the general succession of terrace gravel resting on dark grey muddy diamicton. The latter rested on over 6 m of intercalated brown sand, grey silt and mud, and dark grey muddy diamicton. Shell fragments occur in varying concentrations and states of preservation throughout these layers. Whole shells, including specimens of Lunatia pallida and valves of Arctica islandica and Macoma balthica were recovered from a sand layer at a depth of 12 m (Table A1.4). The base of the mud sequence was not seen, but a pit beneath the adjacent floodplain of the Burn at [NJ 7234 5602] showed that it rests on coarse glaciofluvial gravel and on bedrock. Although not conclusive evidence, the presence of disturbed contorted and steeply dipping beds is strongly suggestive of glacial disturbance, possibly rafting.

Jamieson (1865) noted that the faunas from King Edward and Gardenstown (see below) are similar. He tabulated the then known modern distributions in terms of those living:

Item (3) unfortunately gives a misleading ‘cold’ impression because ‘within the Arctic Circle’ includes the coast of Norway with its boreal fauna (Zenkevitch, 1963). Of the shells in Jamieson’s list, only two (Tachyrhynchus reticulata and Serripes greenlandicus) can be classed as truly arctic to subarctic. Another (Yoldia limatula) is an American species that may be confused with the arctic to subarctic Y. hyperborea (Ockelmann, 1954). Excepting these arctic species, a deep-water taxon (Yoldiella lucida) and two boreal taxa (Polinices nanus and Turritella communis), all the molluscs listed in (Table A1.4) have been recorded in the Late-glacial (Windermere) Interstadial (13 000–11 000 BP) Clyde Beds of western Scotland (Smith et al., 1904). The fauna may thus be taken as generally of non-arctic, interstadial, offshore aspect.

It seems likely that the shell-bearing sands and muds at King Edward are glacially transported rafts within a sequence of glacial deposits derived largely from the bed of the Moray Firth (see Chapter 8; Whitehills Glacigenic Formation). King Edward lies only 1 km north-west of Plaidy (Map 5) where a large erratic of Oxfordian mudstone was worked in the 19th century (Jamieson, 1859). Other shell-bearing marine deposits previously thought to be in situ (Sutherland, 1981) have since been shown to be erratic masses, for example at Clava (Merritt, 1992b), Gardenstown (Peacock and Merritt, 1997) and the Boyne Limestone Quarry (Peacock and Merritt, 2000a).

At King Edward, amino-acid ratios between 0.073 and 0.095 (mean value 0.078 + 0.010) have been obtained from five Arctica shells collected from till at a site 200 m north-east of Jamieson’s section. Uncalibrated AMS radiocarbon ages of greater than 44 200 BP (AA–1323) and greater than 41 500 BP (AA–1324) are reported on two of the analysed shells (Miller et al., 1987). On this basis the shells in the Castleton Member of the Whitehills Glacigenic Formation have been assigned to the interval between 40 and 80 ka BP (Miller et al., 1987). This age is consistent with the faunal evidence at King Edward of interstadial conditions. It appears on current evidence that these marine muds and sands were originally deposited on the floor of the Inner Moray Firth during OIS 4 or 3.

The timing of the glacial phase or phases that emplaced the rafts of marine sediment is uncertain. While some tills and rafts derived from the Moray Firth are thought to have been deposited during the Late Devensian (Merritt, 1992b), the possibility remains that some were transported during a Middle Devensian glacial phase equivalent to the Norwegian Skjonghelleren glaciation (Figure 43). The dark grey shelly tills of the Whitehills Glacigenic Formation around King Edward are covered in places, or merge upwards into, red-brown sandy till (Read, 1923). The latter is probably laterally equivalent to the Crovie Till Formation at Gardenstown and the Old Hythe Till Formation at the Boyne Limestone Quarry, but the exact age of all these units is unclear (Table 7).

Site 5 Crossbrae Farm, Turriff

The Late Pleistocene sequence uncovered in a drainage ditch at Crossbrae Farm [NJ 753 512], 3 km north-east of Turriff ((Figure A1.9)a; (Map 5)), in 1980 is important for three main reasons. Firstly, the interstadial organic deposits found there predate the last major cold stage and such sediments are rare in Scotland (Lowe, 1984). Secondly, the organic sediments yielded radiocarbon ages of 26 400 ± 170 and 22 380 ± 250 yr BP (SRR–2041) (Hall, 1984), which apparently indicate that they formed during an interstadial episode immediately before the build up of the last ice sheet in Buchan, thus constraining the timing of its expansion. Thirdly, the organic deposits were thought to be overlain only by solifluction deposits (Hall, 1984). The site was therefore cited as part of a body of evidence that this part of Buchan had escaped glaciation during the Late Devensian (Sutherland, 1984a). The results of further excavations at Crossbrae in 1992 ((Figure A1.9)b) and a multidisciplinary investigation of the sediments have been reported by Whittington et al. (1998).

The Crossbrae Farm Peat Bed locally rests on weathered Devonian pebbly sandstone ((Figure A1.9)c). The peat may also rest on till, as the drainage contractor for the 1980 excavation reported a ‘hard, reddish-coloured boulder clay’ (Crossbrae Till Formation of (Table 7)), at least 20 cm thick, beneath the peat, but only weathered bedrock was encountered in the 1992 excavations.

The Crossbrae Farm Peat Bed reaches a maximum known thickness of 55 cm and comprises sandy peat with interbedded silty sands and sand laminae. Pollen analysis has revealed a former dwarf shrub tundra vegetation, with Betula nana and Salix herbacea. Bruckenthalia spiculifolia was also present in the flora. Supporting evidence of an interstadial environment is provided by a range of plant macrofossil remains and by a total of 40 coleoptera taxa, including Olophnum boreale (Payk.), Acidota quadrata (Zett.) and Boreaphilus henningianus (Sahlb.). None of these beetles live today in the British Isles, but each is found in the present-day fauna of northern Fennoscandia. Based on the overlap of the climatic envelopes of 23 coleoptera species, the average temperature at the time of the formation of the Crossbrae Farm Peat Bed is estimated as:

Mean temperature of the warmest month 10°C ± 1°C

Mean temperature of the coldest month -9°C ± -3°C

Further radiocarbon age determinations for the Peat are as follows:

These dates must be seen as minima for the Crossbrae Farm Peat Bed. The dates obtained earlier therefore appear to be anomolously young, probably owing to contamination by younger carbon in groundwater.

The Crossbrae Farm Peat Bed clearly predates the Late-glacial and the Sourlie Interstadial around 30 ka (possibly equivalent to the Ålesund Interstadial of western Norway, (Figure 43)). Crossbrae is one of only five sites in Scotland from which pollen of the Balkan heath Bruckenthalia has been recovered, the others being Camp Fauld (Whittington et al., 1993) and Burn of Benholm (Auton et al., 2000) in north-east Scotland (see below), and Sel Ayre (Birks and Peglar, 1979) and Fugla Ness (Birks and Ransom, 1969) on Shetland (Whittington, 1994). The organic deposits at Allt Odhar (Walker et al., 1992), Sel Ayre and Camp Fauld (Hall, 1993c) all appear to relate to an Early Devensian interstadial in which an early warm phase is succeeded by significantly colder conditions. Available dating evidence suggests correlation of organic sediments at Allt Odhar, Sel Ayre and Burn of Benholm with the Brørup Interstadial, equivalent to OIS 5c (Table 7). The flora and coleoptera at Crossbrae preclude correlation with the warm stages of OIS 5a or 5c, but may represent the later colder phases. The remarkable similarity between the coleopteran assemblages at Crossbrae and Allt Odhar provides support for a common age. The Crossbrae Farm Peat Bed is tentatively regarded as being of OIS 5c age (Whittington et al., 1998).

Excavations in 1992 showed that the Crossbrae Farm Peat Bed is directly overlain by coarse gravel up to 1.2 m in thickness. This clast-supported pebble and cobble gravel unit has a strongly erosive base and is dominated by quartzite and quartzose psammite clasts. It is succeeded by crudely stratified, clast-rich diamictons up to 2.5 m thick. The diamictons contain striated pelite clasts and show a strong down-slope clast fabric. They are interpreted as soliflucted tills.

The significance of the Crossbrae Farm site is the presence of a peat deposit that apparently represents an Early Devensian interstadial. The most recently obtained radiocarbon dates indicate that the peat is older than the Middle Devensian and therefore cannot constrain significantly the age of the last glaciation of this part of Buchan. The presence of coarse gravel of possible glaciofluvial origin is important. This gravel unit is most likely to be of Late Devensian age and as such provides no support for the view that part of Buchan escaped glaciation in the Late Devensian. However, it is possible that the gravel is older, as glacial deposits at the Howe of Byth site in Buchan (Hall et al., 1995b) and in the vicinity of Teindland in lower Strathspey (Hall et al., 1995a) have been ascribed to cold stages in Oxygen Isotope Stages 4 and 3.

Site 6 Howe of Byth Quarry

Howe of Byth gravel quarry [NJ 839 753] lies at the confluence of a broad north-east trending valley and a meltwater channel orientated north-west to south-east and which now drains the Moss of Fishrie (Map 4). The site, on Sheet 97, is thought to provide evidence of the glaciation of north-east Scotland in both the Late and Middle Devensian (Hall et al., 1995b). The sequence at the site also includes peat of the Late-glacial (Windermere) Interstadial overlain by cryoturbated gravel laid down during the following Loch Lomond Stadial.

The lithostratigraphy of the site was first set up informally by Hall et al. (1995b), but many of the units have been renamed subsequently by Sutherland (1999) in order to adhere more closely to international lithostratigraphical guidelines. As outlined in Chapter 8, the new names have been incorporated into the regional lithostratigraphy only where appropriate (Table 7). The original and new names are given in (Table A1.5) for comparison.

The oldest deposit is the Howe of Byth Gravel Formation, up to 13 m thick (Figure A1.10), which rests on granite, psammite and weathered red arkosic sandstone bedrock. It consists mainly of quartzite cobbles, which show crude subhorizontal bedding and are derived from the local Devonian conglomerates, together with sparse lenses of imbricate pebble gravel and cross-bedded sand. Palaeocurrents were directed towards the south. Brown gravel-rich diamictons, with a silty sand matrix, occur in the upper part of the H owe of Byth Gravel For mation and thicken to the north. In 1997, new sections in the northern part of the quarry showed the unit to comprise up to 8 m of gravel and gravelly diamicton. The beds and lenses of quartzite gravel were discontinuous and 20 to 80 cm thick, dipping gently southwards. The diamicton units were reddish brown in colour and comprised sandy matrix-supported quartzite gravel. Erect pebbles were noted at two locations within these diamictons indicating cryoturbation. The pit had been extended northwards by 1999 and erect pebbles were also observed towards the top of the unit. The gravels could now be seen to occupy a broad depression in the bedrock surface and the lowest beds contained much soft red sandstone similar to the underlying bedrock. The gravels were heavily iron stained with much black ‘iron-pan’ in the upper 4 m or so. The diamictons were seen to be greenish grey in colour and some of them could be traced for over 200 m southwards.

The Howe of Byth Gravel Formation is interpreted as forming an ice-proximal fan, deposited by meltwater draining southwards, with associated debris flow activity. In the distal parts of the fan, lenses of imbricate gravel and cross-bedded sand suggest formation of ephemeral bars. Although the high-energy, ice-proximal nature of the formation implies that ice was present along the northern coast of Buchan, the only material of nondurable Moray Firth origin observed within it was a single clast of grey mudstone.

Luminescence dating has given ages of 45 ± 4 and 37 ± 4 ka BP for sands within the Howe of Byth Gravel Formation (Hall et al., 1995b). If correct, these imply that the formation is Middle Devensian in age. This is contemporary with the Skjonghelleren Glaciation of western Norway between 42 and 36 ka BP (Larsen et al., 1987) (Figure 43). However, the degree of weathering observed in 1999 may cast doubt on the dating results. Similar developments of iron-pan were present in the gravels at Tillybrex, near Ellon, which are probably pre-Devensian in age (Chapter 8), and have been seen in the oldest gravel unit at Leys, likely to be of OIS 8 age (see below). A problem is that the gravels contain much Old Red Sandstone-derived iron oxide as coatings to clasts. This material is easily mobilised and can give the impression of prolonged weathering. It is noteworthy that the most advanced development of staining and pan is in the proximal part of the fan where Old Red Sandstone material is most abundant in the gravel.

Overlying the Howe of Byth Gravel is the Byth Till Formation. This freshlooking, reddish brown, massive, matrix-supported, silty sandy diamicton is up to 3 m thick, but averages 0.8 to 1 m. Clast lithology is dominated by quartzite, psammite and Devonian sandstone. The Byth Till has a strong west–east clast fabric. The distinctive lithology and fabric suggest that it was deposited by a different ice stream to that from which the Howe of Byth Gravel was derived. The Byth Till is placed in the East Grampian Drift Group and is probably of Late Devensian age.

In the northern part of the quarry, the Byth Till Formation is overlain by up to 3 m of quartzite cobble gravel, the Auchmedden Gravel Formation. In lithology, this unit resembles the Byth Gravel except that it is less weathered. Clast imbrication suggests deposition by meltwater flowing to the south from an ice front to the north; not obviously that which laid down the underlying Byth Till. The Auchmedden Gravel dates from the deglaciation phase of the last ice sheet, while ice remained to the north of the site.

To the south, the Auchmedden Gravel thins and the Howe of Byth Gravel extends close to the surface. The Thinfolds Peat Bed, up to 1 m thick, lies within a shallow basin on the surface of the till. Pollen analysis shows typical Late-glacial (Windermere) Interstadial pollen assemblages. This dating is supported by a radiocarbon age determination of about 11.3 ka BP (SRR–4830) (Table 8) and by luminescence ages of 13 ± 1.4 and 14 ± 4 ka BP. Overlying the peat is a unit of cryoturbated gravel about 1 m thick, the Todholes Gravel Bed, and 1.3 m of Holocene peat (Figure A1.10).

Further notes taken on visits to the quarry in 1999 to 2000 are given in Hall and Connell (2000).

Site 7 Kirkhill and Leys quarries

The Middle and Late Pleistocene deposits preserved in and around Kirkhill [NK 011 528] and Leys [NK 005 525] quarries (Map 6), are unique and represent the most complete sequence of this age known on land in Scotland. The deposits ((Figure A1.11); (Table A1.6)) apparently record evidence of events spanning three 100 ka interglacial/glacial climatic cycles. A wide variety of sediments deposited under glacial, glaciofluvial, periglacial and fluvial regimes, are present, together with two buried soils of interglacial and interstadial aspect (Connell et al., 1982; Hall, 1984a; Connell and Hall, 1987; Hall and Connell, 1991; Hall and Jarvis, 1993a).

Kirkhill Quarry (Plate 22), alas now completely filled by landfill waste, exploited an east–west-orientated felsite dyke that cores a drumlinised ridge of similar orientation. Leys Quarry is situated in a felsitic sand and gravel deposit that lies within an ice-scoured depression to the west of the ridge, carved out of weathered basic igneous rock.

Lithostratigraphy

The lithostratigraphy described below generally follows that set up in the papers cited above. In order to adhere more closely to internationally agreed guidelines on lithostratigraphical nomenclature, new names have been proposed by Sutherland (1999). The latter have been adopted where appropriate into the present scheme presented in (Table 7) and (Table A1.6).

Unit 1 - Leys Till Formation

Pits opened in the floor of Leys Quarry revealed up to 2.5 m of till resting on weathered bedrock. The till is a matrix-supported silty clayey sandy diamicton with pebbles of mainly quartzite and basic igneous rocks. Matrix colour is dominantly olive brown (2.5Y 4/4), but in its upper part the diamicton shows extensive yellowish brown (10YR 5/6) to strong brown (7.5YR 4/6) mottling and contains many grussified basic igneous clasts. It passes down into grussified basic igneous and kaolinised psammite bedrock and contains reworked masses of this material, demonstrating that weathering of the bedrock predates deposition of the till (Hall et al., 1989).

The Leys Till contains many locally derived clasts, but the dominance of quartzitic material indicates derivation from the west. It is the oldest known till in Buchan and has no known correlative deposits.

Unit 2 - Denend Gravel Formation (Leys Lower Gravels)

The overlying Denend Gravel Formation comprises 3 to 5 m of coarse felsite-dominated pebble to boulder gravel, with interstratified sand units. Clast shapes are dominantly subrounded to rounded. Clast lithologies comprise over 90 per cent felsite, mainly of very local origin, together with intermediate igneous rocks, gneiss, vein quartz and schist. The main primary structure in the gravels is crude parallel stratification. In contrast, large scale deltaic planar cross-stratified gravels with thick fining-upwards foresets dipping at 20° to 25° towards the west-south-west occur in the north-western corner of Leys Quarry. Massive, matrix-supported units of coarse gravel are also locally developed. Orientation of both planar cross-stratified gravels and trough cross-stratified sands indicates deposition by water moving towards the west and south-west.

Postdepositional disturbance of the gravels is widespread. Bedded sands and gravels show low amplitude flexures and higher amplitude folds, and minor faults are widespread within sand units. The folded gravel beds are further disrupted by large wedge structures with upper widths that may exceed 10 m. The wedges are infilled with sand and gravel of varied character ranging from downwarped beds of stratified gravel to unstratified, yet crudely sorted, granule to boulder gravel (Hall and Connell, 1986, fig.4). Truncation of anticlinal fold crests and the thickness of gravel beds preserved by downwarping and collapse in the wedge structure indicate that at least several metres of gravel were eroded prior to deposition of flat-bedded sands and gravels, probably part of the Pitscow Sand and Gravel Formation (see below). Locally, these gravels have more conspicuous angular clasts of felsite and appear less disturbed.

Postdepositional folding, faulting and development of wedge structures are interpreted as the result of melt-out of ice blocks originally buried within the gravels. The coarse, but variable calibre of the gravels indicates high, fluctuating palaeoflow regimes. Sedimentary structures indicate a braided channel environment, with meltwaters draining to the north-west, with subsidiary deposition within a proglacial lake. The gravels probably formed as part of a valley sandur extending from an ice margin lying relatively close to, and east of, Leys. Importantly, the easterly source of meltwater suggests that the ice mass was not that which deposited the underlying westerly derived Leys Till Formation.

Unit 3 - Kirkton Gelifluctate Bed (Kirkhill Gelifluctate 1)

In the south-east face at Kirkhill, Unit 4 is underlain by and interbedded with angular felsite rubble up to 2 m thick derived from adjacent intensively frost-shattered bedrock in channel walls. Clasts in the rubble show a well developed downslope fabric, indicating transport by avalanche or gelifluction. Silt cappings indicate incipient pedogenesis (Connell and Romans, 1984, p.76).

Unit 4 - Pitscow Sand And Gravel Formation (Kirkhill Lower Sands and Gravels)

This unit comprises up to 4 m of light olive brown sand with thin gravel beds. Most of the sands are horizontally stratified, but trough and planar cross-stratification occurs locally. In places, the sands seem to have been deposited in a series of stacked, wide and shallow channels, with thin gravel lags evident at their bases. Cross-stratification suggests transport from the east.

This unit is interpreted as a periglacial fluvial or proglacial glaciofluvial deposit. Deposition of the basal beds took place under periglacial conditions judging from the occurrence of syndepositional ice-wedge casts. The declining numbers of angular clasts in the upper beds at Kirkhill, together with the absence of ice-wedge casts and other features of cryoturbation at these levels may indicate amelioration of climatic conditions.

Unit 5 - Kirkhill Palaeosol Bed (Kirkhill Lower Buried Soil)

The Kirkhill Palaeosol Bed is developed on the eroded upper surface of the sands and gravels of Unit 4 (Connell and Romans, 1984, p.70). The upper 19 cm of the profile generally consists of light grey (10YR 7/1) sand and sparse gravel. Felsite clasts are bleached and softened and the sand fraction is depleted in less resistant heavy minerals (Connell et al., 1982). This conspicuous bleached horizon occurred throughout Kirkhill Quarry, and in the north-east face of Leys Quarry. The base of the grey horizon is marked by a sharp, undulating junction with a lower, greyish brown (10YR 5/2) mottled horizon up to 7 cm thick. This horizon is locally enriched in organic carbon and free iron as a result of translocation of these materials in the soil. Analysis of this ‘Bs’ horizon has indicated the presence of protoimogolite/allophane. This mineral is also found in podzols formed during the Holocene in north-east Scotland (information from M J Wilson, Macaulay Institute of Soil Research, Aberdeen, 1994) and suggests that the palaeosol was also formed under humid temperate climatic conditions. An iron pan is found towards the base of the profile (Connell et al., 1982).

At both Kirkhill and Leys, bleaching and iron/manganese staining associated with localised cementation occur extensively in the gravels and sands of Units 2 and 4. The Kirkhill Palaeosol seems therefore to be part of a weathering profile that in places extends throughout the entire thickness (about 4 m) of gravels.

Micromorphological analysis of the palaeosol has also revealed structures typical of soils of arctic environments (Connell and Romans, 1984, p.70). Superimposition of periglacial soil characteristics on an earlier interglacial soil horizon seems to have produced a composite soil horizon. The stratigraphical position of the palaeosol indicates that it developed during a pre-Ipswichian interglacial, but not necessarily the Hoxnian (OIS 11) as originally suggested (Connell et al., 1982; Hall and Connell, 1991).

Unit 6 - Swineden Sand Bed (Kirkhill Organic Muds and Sands)

At Kirkhill, the truncated Kirkhill Palaeosol Bed is overlain by 0.1 to 0.7 m of poorly stratified and weakly organic sands. These sands probably represent slope-wash deposits and are penetrated by a series of frost-cracks (Connell 1984b, p.63). Beneath the sands lies a 1 to 5 cm band of black to brown weakly laminated organic mud that drapes the truncated palaeosol in the south-west face.

The organic mud contains pollen of Poaceae, together with a marked arboreal component of mainly Pinus and Alnus, together with charcoal. The overlying sands show a reduction in arboreal pollen and an increase in grasses and Calluna, possibly reflecting the establishment of an open, treeless environment (Connell et al., 1982). Sampling of an equivalent organic sequence in the west face at Kirkhill suggests that two components exist in the pollen spectra. Open, grassland types represent pollen contemporaneous with sedimentation of the sands. Recycled pollen, with an important arboreal component, is perhaps derived from older soil horizons at the site (information from J J Lowe, Royal Holloway University of London, 1990).

Initial radiocarbon dating gave finite ages for three samples, but contamination was suspected (Connell et al., 1982). Later dating of a much larger sample gave an age of over 47 360 radiocarbon years BP (SRR–2416) and confirmed that the sediments are beyond the range of radiocarbon dating (Hall, 1984).

Unit 7 - Camphill Gelifluctate Bed (Kirkhill Gelifluctate 2/3)

At Kirkhill Quarry, up to 2.5 m of periglacial mass movement deposits are developed above the organic Swineden Sand Bed. In Leys Quarry, a series of ice-wedge casts, spaced 6 to 9 m apart, penetrate downwards from the upper surface of this unit. The structures are filled with sand from the same unit, and not gravel from Unit 8 above. Examination of the upper surface exposed during quarrying showed that the casts form part of a fossil polygonal network. The deposit is cryoturbated with erect pebbles in its upper 0.5 to 0.9 m. The basal part of the gelifluctate is undisturbed and shows subhorizontal bedding. The evidence indicates multiple phases of periglacial climate. A luminescence date of 142 ± 19 ka BP (Duller et al., 1995) has been obtained from an exposure of this unit in the north-east face of Leys Quarry. This date, if correct, falls within OIS 6 and constrains the ages of the overlying deposits.

Unit 8 - West Leys Sand And Gravel Formation (Leys Upper Sands and Gravels)

The West Leys Sand and Gravel Formation rests in shallow channels (less than 1 m deep) scoured into Unit 7 and are only seen in the north-east corner of Leys Quarry. The sandy pebble gravel is very pale brown to pale brown (10YR 7/3-6/3) in colour and is dominated by quartzite and psammite. Crude horizontal bedding of clast-supported gravels is common, with local development of planar cross-beds indicating water flow from the north-west. These thin, probably glaciofluvial gravels and sands are separated by a sharp subhorizontal contact from the overlying unit. They represent outwash or subglacial meltwater deposits related to the emplacement of the overlying till.

Unit 9 - Rottenhill Till Formation (Kirkhill Lower Till)

This is a yellowish brown (10YR 5/4) matrix-supported, silty sandy diamicton. Clasts include mainly pebbles and cobbles of quartzitic metasedimentary rock, pelitic schist and felsite. Clasts of red granite are common at Kirkhill and are probably derived from the White Cow Wood area to the west. Present also are red-stained quartzite cobbles, probably derived from Devonian conglomerates, and grey granite from the Strichen intrusion to the north-west. Clay mineralogy is dominated by kaolinite and illite. Normally, the till is massive, with poorly defined subhorizontal jointing, and a weak north-west–south-east fabric (Connell, 1984b, p.64). Both clast lithology and fabric indicate that the till was deposited by ice moving from the north-west or west.

Unit 10 - Fernieslack Palaeosol Bed (Kirkhill Upper Buried Soil)

Evidence of soil development on the surface of the underlying Rottenhill Till Formation includes:

The palaeosol is truncated, but resembles the B and C-horizons of gleyed brown earths developed on similar parent materials in eastern Scotland during the Holocene (Connell and Romans, 1984, p.76). Support for this interpretation is provided by the presence of Alnus, with Betula and coryloid grains from a single sample in the profile (Connell et al., 1982). The palaeosol appears to be quite widely preserved south and west of Kirkhill as a mottled zone in the upper part of the till. On the simplest interpretation and in the light of the luminescence date from Unit 7, it was probably formed during the Ipswichian Interglacial (OIS 5e).

Unit 11 - Corsend Gelifluctate Bed (Kirkhill Gelifluctate 4)

A range of periglacial deposits and features formed in the interval prior to deposition of Unit 12 and after formation of the weathering profile within Unit 10. In stratigraphical order from oldest to youngest these are:

1. Truncation and cryoturbation of the Fernieslack Palaeosol Bed, with erection of pebbles in the uppermost 0.6 m.
2. Deposition of a thin (0.6–0.8 m) brownish yellow (10YR 5/3) silty gelifluctate unit with angular felsite clasts (Corsend Gelifluctate Bed).
3. Incipient pedogenesis under periglacial conditions, with development of silt droplet fabrics in the gelifluctate.
4. Ice-wedge growth. The presence of an ice-wedge cast extending downwards from the surface of the gelifluctate and filled with matrix material from Unit 13 indicates that glaciation of the site began while ground ice remained in the wedge (Connell, 1984c, p.68).

Unit 12 - Corse Diamicton Formation

This unit has been observed to overlie both sediments of Unit 11 and Unit 9. In turn, it is known at both Kirkhill and Leys Quarries to be either overlain by the Hythie Till Formation (Unit 13) or to be reworked into the base of that till. The Corse Diamicton was formerly regarded as a component of the East Leys Till, but these units, while of broadly similar lithology, appear to lie below and above the Hythie Till respectively.

The Corse Diamicton Formation was first exposed at Kirkhill as a large mass of dark grey clay consisting largely of reworked Mesozoic mudstone. This black (10YR 3/1) to dark grey (5Y 4/1) silty clay contained dinoflagellate cyst assemblages of Late Jurassic/Early Cretaceous age (Connell, 1984c, p.69) and had a smectite-dominated clay mineralogy. The unit contained sparse small, rounded pebbles of quartzite or psammite. At Leys, the unit included deformed bodies of pale pinkish yellow fine-grained sand, very similar to the glacial rafts at Oldmill Quarry (see below).

Unit 13 - Hythie Till Formation (Kirkhill Upper Till)

This unit is a mainly brown (10YR 4/3) massive, sandy silty matrix-supported diamicton, up to 3 m thick. Clast lithology is dominated by psammite and pelitic schist, with basic igneous rocks and felsite. A single clast of rhomb porphyry typical of those found in Norway was recovered from this unit. A characteristic feature of the till at Kirkhill is the presence of boulders of basic igneous rock that were probably originally formed as corestones. Clay mineralogy there is dominated by an interstratified smectite/chlorite mineral. At Leys, the till contains rare clasts and thin lenses of red clay (Hall and Connell, 1986, p.23).

The dominance of quartzitic and basic igneous rocks in the Hythie Till Formation indicates a western provenance. The unit can be correlated on the basis of stratigraphy and lithology with tills along the valleys of the North and South Ugie Water, which also have strong west–east fabrics (Hall and Connell, 1991).

Unit 14 - East Leys Till Formation

This till is a very dark grey (2.5YR N/3) massive clayey silty diamicton with scattered matrix-supported pebbles and cobbles. Clast lithologies are dominated by dark grey-stained gneissose quartzite and psammite, with chalk, pelites and reddish brown (possibly Devonian) sandstone. The till is weathered in its upper 2.5 m, with grussified basic igneous, softened chalk clasts, the absence of shell fragments and development of reddish brown (2.5YR 4/4) to dark greyish brown (10YR 4/2) hues. The clay mineralogy of fresh diamicton comprises more or less equal proportions of interstratified mica smectite, illite and kaolinite.

Striated shell fragments appear below a depth of 2.5 m and include specimens of Macoma sp. and Astarte sp. The matrix contains reworked Late Jurassic/Early Cretaceous dinoflagellate cysts, Mesozoic and Tertiary pollen, and Quaternary foraminifera (information from R Harland, BGS, 1980 and L A Riley, Bovingdon, Herts, 1997). Together these organic remains indicate derivation of the matrix from Mesozoic mudstone, early Tertiary sediments and Quaternary glaciomarine and marine muds in the Moray Firth basin. Southward transport by ice from the Moray Firth is supported by the presence of chalk clasts and by the transport of dark gneissose rocks comparable to the Inzie Head Gneiss of the Fraserburgh area.

Unit 15 - Kirkhill Church Sand Formation (Kirkhill Upper Sands)

These sparsely developed sands and subordinate gravels overlay the Corse Diamicton Formation in the north face of Kirkhill Quarry, where they appear to fill a shallow channel. Cross-bedding indicates transport towards the west. The heavy mineralogy of the sand fraction resembles that of the Corse Diamicton in that it contains significant amounts of garnet together with apatite.

As no diamicton similar to the Hythie Till Formation was identified in exposures along the north face of Kirkhill Quarry, it is possible that a hiatus occurs at this point in the sequence. Alternatively, the channel within which the sands sit may have been locally eroded into the Hythie Till as well as the Corse Diamicton (as accepted here). If this was the case, the sands relate to the subsequent advance of ice derived from the Moray Firth, which deposited the East Leys Till. It is likely that outwash delivered through the channel contributed the sediment that was formerly visible at Sandhole [NJ 998 521], which formed as a delta built in Lake Ugie during the Main Late Devensian Glaciation (Connell, 1984d)(see below).

Unit 16 - Manse Gelifluctate Bed (Kirkhill Gelifluctate 5)

This unit overlies the Hythie Till and includes felsite rubble, crudely bedded and gently dipping sand, and pebbly diamicton. At Leys, the unit displays involutions with stone pillars, near-surface concentrations of frost-heaved clasts, erected and frost-cracked pebbles and associated development of ice-wedge casts (Hall and Connell, 1986, p.24). Ice-wedge casts have also been noted within sands of the fan delta at Sandhole (Connell, 1984d, p.85).

Discussion

The Pleistocene sequence in the Kirkhill area comprises a record of three complex stadial–interstadial–interglacial cycles. In the oldest cycle, the till of Unit 1 is succeeded by the outwash gravels of Unit 2, then by the periglacial deposits of Unit 3 and at the base of Unit 4 and, lastly, by a period of interglacial conditions during which the Kirkhill Palaeosol developed. A second cycle begins with the overlying organic sands and muds of Unit 6, which record a climatic deterioration, which then deepens to provide the periglacial slope deposits of Unit 7 and culminates in glaciation of the site, with deposition of Unit 9, the Rottenhill Till. Subsequently, the Fernieslack Palaeosol Bed developed on the till and its pedogenic characteristics imply a return to interglacial conditions.

The youngest cycle commences with the varied periglacial deposits of Unit 11. This phase of climatic decline terminates with the deposition of the Hythie Till Formation on to permafrost. The flanks of the drumlinised ridge at Kirkhill are underlain by this till (Unit 13), suggesting that it was laid down during the final shaping of the feature. An important distinction is made here between the Corse Diamicton and the East Leys Till, both dark grey, mud-rich diamictons derived from the Moray Firth, but separated by the Hythie Till. This situation resembles that at Oldmill (see below) where a raft of glaciomarine mud derived from the Moray Firth is overlain by a thin till of inland derivation (Hall, 1984a). Subsequent deglaciation involved minor glaciofluvial deposition and was accompanied, or succeeded by, a phase, or phases of periglacial activity until soil formation began at the start of the present interglacial.

Chronostratigraphical information is provided by the luminescence date of 142 ± 19 ka BP (Duller et al., 1995) from sandy gelifluctate in Unit 7, which together with the overlying Rottenhill Till date to OIS 6. If this is correct, then the Fernieslack Palaeosol Bed is likely to be of OIS 5e age. The Corse Diamicton probably correlates with the Whitehills Glacigenic Formation of late Middle or, more likely, early Late Devensian age. The age of the Kirkhill Palaeosol Bed and underlying glacial and periglacial deposits is uncertain. The simplest model would have them of OIS 7 and OIS 8 ages respectively, but they may be older.

Site 8 Oldmill Quarry

This disused gravel pit [NK 024 439] on Sheet 87(E) Peterhead (Map 7) is excavated into the western flank of the Hill of Ludquharn, a triangular-shaped, till-covered, rock-cored hill that appears to have been partially streamlined by ice moving towards the south-east. Deltaic sands and gravels at least 10.4 m thick (McMillan and Aitken, 1981) are overlain by over 6 m of till. The deltaic deposit consists of coarse-grained quartzofeldspathic sand with beds and lenses of cobble gravel composed mainly of pink granite with some weathered grey granite and gabbro. Foresets exposed in the eastern face of the pit indicate deposition by meltwater flowing southwards into a lake ponded up within the valley of the Burn of Ludquharn.

In 1984, the sand and gravel unit was overlain unconformably by about 1.6 m of brown till containing conspicuous basic igneous clasts and other igneous and metamorphic rock-types of western provenance and possessing a weak west to east fabric (Hall, 1984a). The till incorporated at its base several large (many metres across) erratic masses of Jurassic shale and glaciomarine mud probably derived from the Moray Firth ((Plate 24)c). In 1998, the pit had been extended eastwards into the hill where the till was thicker and now comprised two units separated by a gradational contact (Plate 8c). Other photographs are given in Merritt and Connell (2000).

The lower unit, up to 5 m thick, is a dark grey to dark bluish grey, very stiff, massive, matrix-supported, muddy (silty clay to clayey silt) diamicton with moderately dispersed angular to well-rounded clasts up to 40 cm across. These consist of red sandstone, red granite, basic igneous rocks, psammite, slaty semipelite, well-rounded liver-stained quartzite (from ORS conglomerate) and sparse flint and white sandstone. Several discontinuities and laminae of iron-stained sand occur towards the base of the lower till unit, lying parallel to its sharp, planar to gently undulating, unconformable contact with the underlying sand and gravel. The sand laminae locally pass laterally into lenses up to 1 m thick.

The number of sand lenses within the lower till unit increases towards the south-eastern face of the pit, where they show evidence of complex shearing and interfolding with the bluish grey mud and pebbly mud diamicton. Some of the sand bodies reach 2 m in thickness. The orientation of many overfolds, sheath folds and reversed faults indicate that ice flowed towards the east-south-east. The shearing and attenuation of bedding that has occurred towards the base of the unit is typical of subglacially formed ‘glacitectonites’, similar to those described at the Boyne Limestone Quarry (Site 2) and Gardenstown (Site 3).

The upper till unit is mottled yellowish brown and grey, up to about 2 m thick, and is cut by closely spaced vertical fissures. It is similar in composition and structure to the lower unit. The boundary between the units seen in 1998 is a gradational colour change that occurs over a vertical thickness of 10 cm.

Discussion and correlations

The dark grey clayey till with lenses (rafts) of sand, glaciomarine mud and Jurassic shale is typical of diamictons of the Whitehills Glacigenic Formation (Chapter 8) that were emplaced by ice flowing towards the south-east from the Moray Firth. The rafts, in particular, are diagnostic of this unit at the type section of the Whitehills Glacigenic Formation at the Boyne Limestone Quarry (Site 2). Here, as at Boyne, they are most common towards the base of the unit and become increasingly sheared and attenuated upwards, merging into homogeneous pebbly diamicton (deformation till). Both units of till at Oldmill contain a suite of rock types that generally match with outcrops lying to the north-west of the site, supporting the east-south-east direction of emplacement deduced from the orientation of some glacitecton ic structures in the rafts. However, the significant number of clasts of red granite in the till poses a problem, because although the Forest of Deer Granite is likely to underlie the site (British Geological Survey, 1992), it is grey and the only known outcrops of red granite occur 5 km to the east of the site. The simplest explanation is that they have been derived from the unit of gravel underlying the till, which contains much red granite (see below).

Most researchers would agree that the upper till unit seen in 1998 represents the weathered top of the lower till. However, it is not clear whether the brown till with the weak west–east fabric, basic igneous clasts and western erratic provenance originally recorded by Hall (1984a) is contiguous with this unit of weathered till. It could represent a separate unit (and glacial event) and Connell and Hall (1987) correlated it with the Kirkhill Upper Till (now Hythie Till Formation). If the brown diamicton is simply the weathered top of the grey till, it indicates that during the time taken to deposit the till sequence at Oldmill, the direction of ice movement swung from south-east to east. This is consistent with evidence from the Whitehills Glacigenic Formation at the Boyne Limestone Quarry and Gardenstown, indicating that East Grampian ice gradually became dominant over ice that had earlier pushed inland from the Moray Firth (Peacock and Merritt, 1997, 2000a). The weathering profile is probably postglacial in origin, like that developed on Late Devensian till in Holderness, east Yorkshire (Madgett and Catt, 1978). It is also like that seen affecting blue-grey tills of the Banffshire Coast Drift Group exposed in pipeline trenches north of Peterhead (Chapter 8 Essie Till Formation).

The origin of the basal sand and gravel unit is more puzzling. While clearly formed as a southward prograding glaciofluvial delta, the predominant clasts of red granite in the unit are most likely derived from the Peterhead Granite, some 5 km to the east. As the site lies within a valley draining towards the north-east, the ponding is most likely to have been caused by ice encroaching from the north or ponding of the Ugie valley system to the east. The simplest explanation is that the lake was formed by the ice that later overrode the site and deposited the two units of till. If so, ice must also have occupied the coast to the east, blocking drainage in that direction and causing meltwaters to be routed towards the site carrying the clasts of red granite. A similar situation seems to have occurred in the vicinity of the Kirkhill site, 8.5 km to the north-north-west of Oldmill, where the felsitic Denend Gravel Formation was also deposited by westerly directed meltwaters. Those gravels, however, are considered to be considerably older (see Site 7 Kirkhill and Leys).

The solution to this problem of correlation could be that the sand and gravel at Oldmill is much older than the till units overlying it. For example, Connell and Hall (1987) cited evidence (since corroborated by them verbally) from Smith et al. (1977) that ice-wedge casts in the sand and gravel are truncated by an overlying till of western provenance. Furthermore, the ice wedges are associated locally with a thin gelifluctate unit and many of the granite cobbles and boulders are bleached implying that a hiatus of unknown duration occurred between the deposition of the sand and gravel sequence and the tills.

Site 9 Philorth valley, Fraserburgh

The concealed sediments within the Philorth valley, on Sheet 97, include a sequence of estuarine deposits and interbedded peat. This sequence provides important evidence of the pattern of relative sea level change in north-east Scotland during the Holocene. Because of the marginal setting of the Philorth valley relative to the centre of postglacial isostatic rebound in the western Highlands, these deposits preserve a more detailed record of sea-level movements than is recognised from most other coastal localities (Smith et al., 1982; Smith, 1993).

The Water of Philorth is a small stream that drains north-eastwards to reach the coast 3 km south-east of Fraserburgh (Map 4). The estuarine deposits there are concealed beneath brown silty clay and blown sand at the northern end of the valley [NK 011 635], west of Milltown. The buried sequence records changes in relative sea level during the middle and late Holocene, including a transgressive episode not recognised elsewhere in Scotland. Mapping, levelling and coring by Smith et al. (1982), proved a succession of sands and gravels overlain by peat, with interbeds of grey sand and micaceous sandy silt, beneath the brown silty clay (Figure A1.12).

The most detailed study of the organic sediments was done on cores from Milltown and Philorth Home Farm. Smith et al. (1982) undertook pollen analysis and 14C dating of the deposits at the latter site and nine radiocarbon dates were obtained from the study area (including two from a core at Milltown and five from a core at Mains of Philorth). These analyses showed that the basal peat, grey sand, and the peat below the micaceous sandy silt are both of early Holocene age. Pollen from the sequence indicates scattered stands of birch and pine, with hazel and willow in the general area. The valley floor was subjected to a fluctuating water table, which resulted in the development of a variety of communities typified by sedges, grasses and aquatic plants.

Pollen from the top of the peat above the grey sand layer, in the overlying micaceous sandy silt and in much of the peat above it, indicates an increase in the occurrence of oak and alder; the silt is associated with high values of pine and oak. The top of the peat and the overlying brown silty clay yielded pollen indicating birch-oak woodland, with alder and local freshwater aquatic communities.

The grey sand layer recorded in cores from the lower end of the Philorth valley was deposited by one of a series of tsunami waves that struck the east coast of Scotland some 7000 years ago. The tsunami resulted from the second of three huge submarine landslides, known as the Storegga slides, which occurred on the continental margin at the southern end of the Norwegian Sea (Long et al., 1989; Smith and Dawson, 1990).

The age of the micaceous sandy silt is constrained by both the pollen and the radiocarbon dates, which indicate deposition during the middle Holocene (Table A1.7). The association of Plantago maritima pollen, together with the increase in representation of pine and oak, is characteristic of a littoral depositional environment (Traverse and Ginsberg, 1966) and indicate that the silt is of marine-estuarine origin. Smith et al. (1982) suggest that this deposit represents local expression of the Main Postglacial Transgression (when sea level was relatively higher than at present). They interpret the radiocarbon evidence as showing that the transgression was underway in the area by 6300 ± 60 14C years BP and that it culminated after 6096 ± 75 14C years BP. Surface peat growth had recommenced by 5700 ± 90 14C years BP. These dates are somewhat later than those proposed for this event from sites farther to the south (Smith, 1993) and may be taken as evidence that the transgression was diachronous and resulted in the formation of a time-transgressive shoreline. The brown silty clay present at the surface began to accumulate at about 4760 ± 60 14C years BP as the result of a second rise in relative sea level, again to above present OD.

Site 10 Ugie valley

Deposits of interbedded red, reddish brown, grey and green silt, clay and fine-grained sand occur widely in the valley of the River Ugie and its tributaries on Sheet 87E (Map 7). It is generally believed that the deposits were laid down in a lake (Glacial Lake Ugie), or lakes, that were dammed by ice to the east in the coastal area during the last glaciation ((Figure 42); McMillan and Aitken, 1981; Hall 1984). The complex relationship between periglacial phenomena, deglacial landforms and lacustrine sediments suggests that a significant re-advance of coastal ice occurred, providing important evidence supporting, but not proving, a two-phase Main Late Devensian glaciation.

Lacustrine deposits

The glaciolacustrine deposits (Ugie Clay Formation, Chapter 8) are mostly preserved beneath glaciofluvial terrace gravels at elevations up to 50 m OD. Locally, they reach a thickness of over 17 m and seem to have been derived partly from sediment sources to the west and north, and partly from coastal ice (Peacock et al., 1977; McMillan and Aitken, 1981; British Geological Survey, 1992). At Baluss Bridge [NK 002 470] near Mintlaw (Figure A1.13), a bed of organic mud 0.17 m thick interbedded with the laminated silts contains Palaeozoic and Mesozoic palynomorphs, presumably derived from offshore (Connell et al., 1985; Hall and Connell, 1991). The bed lies below flat-bedded terrace sand and gravel some 4 m thick, and is underlain by till and gravel. It can be traced north for some 500 m, where it reaches the terrace surface at about 35 m OD.

Though no lake shorelines have been mapped, measurements of the upper levels of the Ugie lake are provided by deltas at elevations of about 80 m and 50 m above OD (Figure A1.13). South of Morriston, in the upper valley of the North Ugie Water, eastward-dipping deltaic foresets some 2 to 3 m high were formerly seen in a gravel pit [NJ 915 578] cut into a flat-topped terrace with a surface level slightly over 80 m OD. Near New Leeds, in another gravel pit [NK 002 500], south-westward dipping deltaic foresets, of which 2 to 2.5 m were exposed, underlay thin topset beds that were locally cryoturbated. Spot heights on the delta top are 78 m, 80 m and 84 m OD. The New Leeds delta was fed by glacial meltwater from a shallow channel to the west, and/or from an ice-front that lay a short distance to the east (see below). A near-surface, thin layer of red clay, possibly glaciolacustrine, was observed in 1990 at about 62 m OD in the south-east corner of Leys Quarry.

The presence of a lake with a water level a little above 50 m OD is supported by the occurrence of a glaciolacustrine delta with a surface level at 52.5 m OD at Denhead ((Figure A1.13); Peacock et al., 1977; Connell, 1984d), where the sediments, disrupted by ice-wedge casts, were carried from glacier ice lying to the north. This delta is perched above a younger meltwater channel, indicating that meltwater from the north continued to enter the lake as its level fell from 80 m to below 50 m. The gravels in the valley of the North Ugie Water at Strichen, and at Stuartfield on a tributary of the South Ugie Water, may also be deltaic deposits related to the 50 m lake, but no supporting evidence is available from the sedimentology. The Stuartfield deposit is large in relationship to the small catchment, suggesting that it was derived from a nearby ice-front.

No exits have been identified for the 80 m and 50 m glacial lakes in the lower Ugie catchment. However, since all the ground towards both the north and east coasts lies below 50 m OD, lakes at this level must have been dammed by coastal ice belonging to the Moray Firth ice stream as well as ice from the south-east associated with the Logie-Buchan Drift Group. Thus both ice-streams were present at the time of formation of the glacial lakes. However, the 80 m lake is likely to have been restricted to the valley of the North Ugie since no associated deposits or landforms occur elsewhere.

A narrow, east-south-east-trending ridge (normally less than 80 m wide), which locally reaches a little over the 55 m contour, extends for nearly 2 km from west of Blackhill [NK 017 568] to near Strathstodley [NK 017 568], straddling sheets 97 and 87E (Figure A1.13). Mapping suggests that it is formed of sand, gravel and morainic material (Glentworth and Muir, 1963). Its morphology and composition invite comparison with recessional moraines that were formed in lakes in Glen Spean at the close of the Loch Lomond Stadial (Sissons, 1979). This ‘Strathstodley Ridge’ is interpreted as such a moraine, which was possibly laid down in a small lake co-existing with the 80 m Ugie Lake discussed above. It is likely to have been formed close to the maximum westward limit of the coastal ice at this time.

Terrace gravels and their relationship to the lake sediments

The low glaciofluvial terraces that extend along the valley of the North Ugie Water north of Longside ((Figure A1.13); British Geological Survey, 1992) slope gently from about 45 m OD to just below 25 m OD over a distance of about 4 km. Much or all of the terrace deposit, which varies from sand to boulder gravel (Peacock et al., 1977; McMillan and Aitken, 1981), was derived from the north-west as indicated by cross-bedding and the azimuth of palaeochannels seen in former quarries. The contact of the terrace gravels against the till of the surrounding slopes is rarely represented by a marked back feature because of extensive gelifluction of till over the gravel. The gravels overlie red till at the former Ardlow Quarry [NK 073 505], and red silt at Auchlee ((Figure A1.14); see next). At Dumpston Quarry [NK 029 500], the terrace deposits include numerous boulders of grey granite, possibly Strichen granite, up to 0.5 m across.

The surface of the terrace gravels is commonly in the form of low mounds and hollows, with a relief of a few metres, similar to those seen formerly in cross-sections in Auchlee quarry. Air photographs taken in 1968 show that the fossil tundra polygons seen at this locality (Clapperton and Sugden, 1977) are mostly restricted to a terrace fragment (A on (Figure A1.14)) immediately to the south of the North Ugie Water, and to remnants of a slightly lower terrace at (B) and (C). Terrace fragment (A) is truncated by a low scarp on its south-western side. The morphology of the elongated depression (D) shows clearly that it is not a river channel. It is believed to be a subsidence feature formed by the melting of a mass of buried ice during climatic amelioration. As such, any tundra polygons would have been destroyed. The three shallow kettleholes (E, F and G), which are partly floored by peat, are interpreted as having formed during the continued melting of the ice that formerly underlay depression (D). Kettlehole (E) is the site of the former Loch of Auchlee, from which the neighbouring farm takes its name, and the other kettleholes in the area, including (H) and (I) were probably also occupied by lochans prior to agricultural drainage.

Exposures seen in the former Auchlee quarry in 1974 showed 1 to 2 m of fine-grained to cobbly gravel overlying a minimum of 3 m of fine- to coarse-grained sand with interbeds (0.1 to 1 m thick) of red, grey and brown silt and clay. The sand, which is known to pass downwards into greyish brown to greenish grey silt (McMillan and Aitken, 1981), is interpreted as a prodeltaic lacustrine deposit that passes upwards into the horizontally bedded braided terrace gravel. Its reddish colour associates it with the deposits of the Logie-Buchan Drift Group lying to the east.

It is envisaged that the glaciolacustrine deposits were overlain by glaciofluvial gravels derived from upstream as the lake level fell. The gravels subsequently became permanently frozen with the formation of tundra polygons, probably prior to the Windermere Interstadial (13 to 11 ka BP) (Ballantyne and Harris, 1994), though Gemmell and Ralston (1984) attributed the formation of the polygons to the Loch Lomond Stadial (11 to 10 ka BP). The thawing of the permafrost could have been associated with the rise in water level to at least 30 m OD following the deposition of the terrace gravels at Faichfield (see below), and would certainly have been completed during the relatively warm interstadial when the ground ice (or relict glacier ice) also disappeared.

In contrast to those at the confluence of the North Ugie Water and the South Ugie Water, the terrace gravels at Faichfield (Figure A1.13) were transported by streams flowing from the Dens channels to the south (Hall, 1984), as shown by the orientation of cross-bedding foresets and palaeochannels. In a former quarry [NK 063 467] where the surface level is about 30 m OD, the gravels were seen to be capped for a distance of some tens of metres by reddish brown laminated silt and clays up to 1.5 m thick (Peacock et al., 1977). The silts and clays are indistinguishable from those below the gravels elsewhere and, unless they are overbank deposits, suggest that there was a rise in water level following deposition of the fluviatile or glaciofluvial gravel. Such a rise could have been caused either by the re-establishment of an ice-dammed lake with a water level close to 30 m OD, or by a marine incursion following a relative rise in sea-level. The latter is thought improbable as the highest raised beaches in the vicinity stand at 15 to 16 m OD in the vicinity of St Fergus (see below). The reestablishment of the lake is more likely to be the result of a significant glacial re-advance of one or both of the coastal ice streams. Supporting evidence includes the record at the former Ednie clay pit [NK 085 502], where red deposits enclosed irregular masses (possibly rafts) of dark grey till, dark blue clay and sand, the latter two items possibly derived from overridden lacustrine deposits (Wilson, 1886; BGS records). The timing of such a re-advance is unknown, but it is tempting to suggest that it followed the cold interstadial between about 22 and 18 ka BP, recognised in the northern North Sea basin (Sejrup et al., 1994, 2000) and correlates with the subsequent Dimlington Advance (Eyles et al., 1994) after 18 ka BP (Figure 42), (Figure 44).

Site 11 St Fergus

Interest in the Quaternary deposits around St Fergus, 6 km north-west of Peterhead (Map 7), began in the mid-19th century following the discovery of an apparently in situ arctic marine fauna (Jamieson, 1858, 1906). It occurred in a stoneless grey clay exposed in a former brick pit close to present sea level near Annachie [NK 105 532] (Figure A1.15). Similar deposits were subsequently encountered in BGS boreholes drilled to the east of the village (McMillan and Aitken, 1981) and in site investigations and construction works at the gas terminal from 1980 onwards. The dominantly massive, very dark grey to dark greenish grey, calcareous silts containing shelly horizons were described and named as the St Fergus Silts by Hall and Jarvis (1989). The unit is named formally here as the St Fergus Silt Formation of the Banffshire Coast Drift Group (Chapter 8; (Table 7)). Early geomorphological work in the area by Walton (1956) and Donner (1963) indicates the presence of raised beach deposits in the area to heights of up to about 30 m above OD, but the most prominent feature was reported at about 15 m OD.

St Fergus Silt Formation occurs on sheets 87E Peterhead and 97 Fraserburgh where it is overlain mainly by alluvial deposits, including peat and lacustrine silt, which mark the site of a former freshwater loch (Anderson in Scott, 1890), and by blown sand (Glentworth and Muir, 1963). The formation probably extends to the north of the Loch of Strathbeg (Figure A1.13) where grey plastic clays have been recorded (Wilson, 1886). Dark grey sandy silts and clays are also widespread beneath terraced glaciofluvial sand and gravel deposits in the vicinity of Savoch [NK 047 588] and dark grey silts occur at 9 m above OD at Coralhill [NK 061 607]. The southern boundary of the formation is obscured by thick surface peat and/or organic mud, but probably lies immediately south-east of St Fergus village. To the west, the St Fergus Silt Formation is bounded by a till slope and a Late-glacial raised beach at about 16 m OD, whereas to the east, the deposits locally form the backslope of the abandoned Postglacial (Flandrian) cliff (Figure A1.15).

The dark grey silts, clays and fine-grained sands of the St Fergus Silt Formation form a generally coarsening upwards succession (Peacock, 1999) with a maximum reported thickness of about 16 m above (McMillan and Aitken, 1981). The surface level is generally 7 to 10 m, but locally up to 16 m above OD adjacent to a low ridge that borders the outcrop to the north of Annachie Bridge [NK 105 531] (Figure A1.15) and extends towards Rattray Head. Exposures of faulted, tilted and deformed masses of bedded silt and sand in the ridge led Hall and Jarvis (1989) to interpret the feature as a (push) moraine formed by a re-advance of ice from the east or north-east. This conclusion is supported by borehole evidence that indicates that the St Fergus Silt Formation passes under, and is therefore mainly older than the moraine feature. If correct, the surface level of the formation in the vicinity of the ridge is likely to be anomalous due to proglacial glacitectonic disturbance. A cutting through the cliffline backing the Flandrian raised beach at Pittenheath [NK 104 554] also revealed faulted, tilted and sheared masses of bedded sand and black clay.

Another low north-trending ridge is cut by the cliffline just to the north of St Fergus village [NK 097 522]. In 1990, the ridge was observed to consist of planar and ripple bedded sand with thin beds of gravel and red diamicton and local ‘streaks’ of grey clay and clayey sand. The sequence is folded into a gentle anticline trending parallel to the ridge and is down-faulted to the east locally. The ridge is draped and flanked by reddish brown diamicton. The folding may either have been associated with the advance that formed the other ridge at St Fergus, or an earlier incursion of the ‘Logie-Buchan’ ice stream.

Shell fragments have been reported from many boreholes, and the presence of dropstones (gneiss, chalk and red sandstone) suggests either a proximal glaciomarine environment or a situation in which stones were ice-rafted by sea ice from a nearby shore (compare with Peacock et al., 1978).

Ice flowing out of the Moray Firth probably eroded the chalk fragments from Upper Cretaceous strata that crop out on the sea floor to the north of Fraserburgh. In excavations west of Annachie Bridge, the deposits consist of both massive and laminated calcareous silts over 5.8 m thick with dropstones up to cobble size (Hall and Jarvis, 1989). Here the St Fergus Silt Formation is underlain by dark grey diamicton also with erratics of chalk and granite.

In excavations at a locality [NK 1023 5285] west of Annachie Bridge, Hall and Jarvis (1989) reported the occurrence of a shell bed at 4.6 m depth. It was formed entirely of valves, mostly paired and unabraded, of Hiatella arctica together with low concentrations of the cold water foraminferid Elphidium excavatum forma clavata. Previously, a poorly preserved boreo-arctic assemblage including the bivalves Hiatella arctica and Yoldiella sp. and the ostracods Roundstonia globulifera and Cytheropteron montrosiense had been recovered from the former clay pit at Annachie (Jamieson, 1858, 1865; Brady et al., 1874). However, it seems likely that temperate species in the fauna, such as the bivalve Thyasira ferruginea, have either been misidentified, or have been reworked from older sediments during the formation of the moraine (Peacock, 1999). The faunal distribution and the general coarsening-upwards sequence perhaps indicate shallowing and a reduction in the rate of sedimentation through time. The Hiatella bed has yielded adjusted radiocarbon ages of 14 915 ± 205 BP (Lu–3028) (Hall and Jarvis, 1989) and 14 345 ± 65 BP (Beta-101953), though these figures are regarded as approximate (Peacock, 1999). Additionally, D/L amino-acid ratios of 0.071 (Lond–584) (mean, SD 0.005, n=9) have been obtained on Hiatella arctica from the same excavation (information from G Sykes). These ratios are commensurate with a Middle or Late Devensian age.

The position of the St Fergus Silt Formation within the local stratigraphy is difficult to ascertain because it rests on a variety of older deposits. The records of commercial boreholes associated with the St Fergus Gas Terminal suggest that the formation overlies an undulating surface of greyish brown or green clayey diamicton. Although the clasts in these diamictons are mainly of local igneous and metamorphic lithologies, in at least one borehole [NK 095 545] the weakly calcareous, very dark greyish brown (2.5 Y 3/2) sandy silty clay matrix contains a rich dinoflagellate flora. Possible Late Jurassic to Early Cretaceous cysts were obtained together with some Quaternary forms (information from R Harland, 1980). A provenance either from the north-west in the Moray Firth, or from the north-east to east in the western North Sea is indicated. In a borehole near North Blackwater

[NK 100 537], the St Fergus Silt Formation is reported to rest on reddish brown diamicton (presumably Logie-Buchan Drift Group) that in turn overlies greyish green till. In another borehole immediately east of Annachie, red till is reported to overlie grey till, but to be capped by dark grey till (McMillan and Aitken, 1981). Both records suggest that Logie-Buchan Drift Group deposits could be locally present below the St Fergus Silt Formation. On the south-eastern side of the gas terminal, probably close to the former North Blackwater Farm [NK 100 537], the St Fergus Silt Formation is reported to rest on dark shelly till with chalk erratics (Hall and Jarvis 1989). Similar dark diamicton with chalk fragments and numerous erratics of granite, probably derived from the local, but unexposed, St Fergus Granite, has been seen in spoil heaps towards the sea. Wilson (1886) recorded a dark coloured till containing supposed Old Red Sandstone flagstone and marl erratics in the St Fergus area and an erratic of ooidal sandstone at Pittenheath [NK 100 550].

The back feature of the Late-glacial raised beach located on (Figure A1.13) stands at just under 15 m above OD [NK 093 529] and is fronted by a platform up to 120 m wide. The platform is underlain by sand and gravel that passes seawards into silt. The latter is possibly the uppermost part of the St Fergus Silt Formation, which implies that the two deposits are roughly contemporaneous. The sand and gravel is disposed as ‘herring bone’ sets of cross-beds, suggesting intertidal deposition.

Discussion

Most of the molluscan fauna in the St Fergus Silt Formation would not be out of place in a glaciomarine setting and the adjusted radiocarbon ages between 14 ka and 15 ka BP on shells of Hiatella arctica fall within a period when high-arctic faunas lived in the northern North Sea (Sejrup et al., 1994; Peacock, 1995). This poses a problem, however, because raised beaches and raised marine deposits generally have been assumed to be absent in peripheral areas of eastern Scotland on theoretical grounds that the contemporary sea level was below OD (Sissons, 1976; Lambeck, 1995). Much more ice must have been present in the vicinity, both onshore and offshore in the North Sea basin. The isolated occurrence of the St Fergus Silt Formation may be explained if, apart from the St Fergus area, much of the coastline of north-eastern Buchan east of Troup Head and north of Aberdeen was ice covered until relative sea level had fallen close to or below its present level.

The surface level of the St Fergus Silt Formation and the height of the supposed raised beach backing it, imply a marine transgression to a mean sea level of at least 12 m above OD and therefore considerable glacio-isostatic depression at about 15 ka BP (Peacock, 1997). This is taken as supporting evidence for total ice cover in Buchan during the Main Late Devensian glaciation (Hall and Bent, 1990) and for coalescence of the Scottish and Scandinavian ice sheets during at least an early phase of that glaciation (Sejrup et al., 2000) (Figure 41), (Figure 44).

Site 12 Sandford Bay

In a series of papers Thomas Jamieson described important coastal cliff sections together with exposures in clay pits in the area immediately south of Peterhead (Jamieson, 1858, 1860a, 1865, 1882b, 1906). The thick succession of Pleistocene glacial sediments he recorded proved seminal to his subsequent reconstruction of the relative timing and pattern of glaciation in eastern Buchan (Figure 36). Since Jamieson’s early work the majority of the sections have now become obscured and the clay pits are long disused. At present, the only exposures available for study are those cliff sections in Sandford Bay, close to Sandford Lodge [NK 125 434] and Burnhaven [NK 124 440] (Map 7). These sections have been described briefly by Connell and Hall (1984c) and Hall and Connell (1991). Generally exposure inland from the coast is very poor, but a number of recent road improvement schemes and borehole surveys have added important information on the lithology, sedimentology, thickness and distribution of the glacial deposits (Figure A1.16). The Sandford Bay sections are some of the few sites in Buchan where tills of the Hatton Till Formation of the Logie-Buchan Drift Group are accessible, together with older glacial deposits.

Glacial stratigraphy described by Jamieson

Immediately to the south of Peterhead, in Peterhead Bay, Jamieson (1882b, 1906) recorded drift deposits overlying outcrops of ‘somewhat soft and disintegrated’ red (Peterhead) granite. Overlying the bedrock was 3.05 to 3.66 m of hard ‘grey boulder clay full of granite and gneiss debris’. In places he described the grey boulder clay as thinning into a ‘mass of coarse granite rubble’. Above the grey boulder clay he described up to 3.05 m of ‘hard and firm’ unstratified red clay from which he reported clasts from Old Red Sandstone conglomerate.

Wilson (1886) also recorded sections in Peterhead Bay, together with others farther south in ‘Brickworks Bay’ probably at about [NK 123 451]. He noted m of ‘red stony boulder clay with blocks of Old Red Sandstone’ resting on ‘bands of yellow silt and sand’. He did not record any of Jamieson’s lower boulder clay. To the south of Brickworks Bay, Jamieson recorded outcrops of red clay at about [NK 126 447], near Peterhead Prison. Here he noted (1906) ‘an irregular, undulating band of blue clay’, a bed of gravel containing a ‘few broken shells’, many blocks of red sandstone, and ‘a bit of hard Chalk with a possible Belemnite fragment attached’. Amongst the shell fragments in the gravel he observed (nomenclature not revised) Cyprina islandica, Pecten islandicus, Astarte arctica, Panopoea, Mytilus, Cardium, Fusus, and Balanus.

Some of the most important sections recorded by Jamieson were those in the long-established Invernettie brick and tile works pit on the north side of Sandford Bay [probably close to NK 126 441]. In 1858, he published the following section (bottom up) from the pit.

Clasts of red and grey granite, schist, ‘greenstone and trap’, greyish sandstone and flint were noted in unit 4, together with broken shells in a coarse reddish sand. Later (1865) he recalled finding Astarte borealis, Cyprina islandica and Littorina squalida (nomenclature not revised). Jamieson also recorded (1858, 1882b) that an almost complete skeleton of a large bird had been found in the clay pit at a depth of either 7.6 m (1858) or 9.1 m (1882). If correct, it would place the find either within the base of unit 4 or close to the boundary between units 4 and 3. He also noted that a large vertebra (possibly fish) had been found earlier at a depth of 11.6 m, probably from unit 1. Seal bones have also been reported from Invernettie (D Page in Turner, 1870), but the horizon is uncertain.

In his 1906 paper, Jamieson noted that the upper reddish brown clays at Invernettie were somewhat different to the usually vivid red clays of his Red Series deposits. The upper 6 m or more were composed of ‘darkish mottled brown clay, varying much in colour, as if red and blue had been jumbled together’. Furthermore, in places, the ‘Red Clay’ was ‘curiously streaked with clay of a dark bluish colour, derived apparently from a different source’.

Jamieson recognised the following sequence of events.

1. Ice advance from the west to beyond the present coastline, depositing the lower grey boulder clay, with clasts predominantly of granite and metamorphic rocks derived from the underlying bedrock and outcrops, farther west.
2. Deposition of glaciolacustrine (or possibly glaciomarine) laminated silts, sands and clays of brick red, dark brown, yellow and brownish grey colour. Deposition occurred following retreat of the earlier ice mass and in front of ice advancing from the south-south-east. The varying colour of the deposits suggested that the sediment was derived from both ice masses (see Site 10 Ugie Valley).
3. Ice advance from the south or south-east deposited a thick sequence of the red and reddish brown shelly diamictons and clays with local ‘streaks of dark bluish clay’. The presence of conspicuous clasts of Old Red Sandstone rocks suggest derivation from offshore and from the south.

From his observations and his wide knowledge of the drift deposits in the region, Jamieson (1906) concluded that the ‘Red Clay’ was deposited ‘close upon the retreat of the ice which lodged the grey Boulder-Clay’. It is likely that the inclusions of dark bluish clay within the red deposits were either derived from his older ‘Indigo boulder clay’, known from the Ellon and Slains area to the south (see Site 15 Ellon), or from the interaction of ice from the Moray Firth and Strathmore in the Peterhead area. The significance of the bird, seal and fish bones at Invernettie is uncertain. It is possible that the vertebrae noted by Jamieson and the seal bones reported by Page (in Turner, 1870) are one and the same.

Glacial stratigraphy in recent exposures at Sandford Bay

Sections described at Sandford Bay over the last 23 years have confirmed much of Jamieson’s earlier interpretation. Additional information on the provenance and sedimentology of the deposits has, however, allowed amplification of many of his conclusions.

The stratigraphy described below is a composite of cliff exposures near Sandford Lodge [NK 125 434] and Burnhaven [NK 124 440] close to the site of the old Invernettie brick pit.

Interpretation of the sedimentology and stratigraphy at Sandford Bay

Unit 2

This diamicton is the oldest unit in the local succession. It is named here as the Sandford Bay Till Member of the Hythie Till Formation (East Grampian Drift Group) (Table 7) Locally, it includes a lower sub-unit interpreted as a deformation till derived from the immediately underlying grussified granite. The main unit contains more far-travelled westerly derived material, has a strong west–east clast fabric and is interpreted to be a lodgement till laid down by East Grampian ice as it advanced to a position beyond the present coastline. The west–east or west-south-west orientated striae recorded by Jamieson (1882b) on nearby Stirling Hill were probably formed during this advance. It is noteworthy that two well-rounded erratics of Norwegian rhomb porphyry (information from J A Dons, Mineralogisk-Geologisk Museum, Sars’ Gate, N-Oslo, Norway, 1979) have been found in this unit, similar to those found in the Hythie Till Formation at Kirkhill; they probably all have been derived from older deposits.

Unit 3

The colours of these deposits, which are assigned here to the Ugie Clay Formation of the Logie-Buchan Drift Group, suggest that sediment was supplied both from the retreating East Grampian ice sheet and from ‘Logie-Buchan’ ice advancing from the south. The deposits contain dropstones and thin beds of diamicton (debris flows) and are interpreted as glaciolacustrine. Peacock (1975) has also reported red ‘flow till’ units associated with red-coloured silts and clays in the area at Boddam.

Only a single unit of dominantly red-brown laminated sediment has been recorded in the cliff exposures. However, in the 28 boreholes drilled at the nearby Peterhead Power Station site (average drift thickness 13.7 m; only a few hundred metres to the south-west and centred on NK 125 433) multiple laminated units, up to 2.0 m in thickness, have been recorded interbedded with red-coloured diamictons. As no exposures are available to check the sedimentology or structural characteristics of these deposits it is unclear if these multiple units reflect a number of discrete ponding events followed by readvances, or a stack of glacitectonic thrust slices emplaced by a single readvance of ‘Logie-Buchan’ ice.

Unit 4

This thick unit of massive to locally sheared diamicton was emplaced by the north or north-westwardly advancing Logie-Buchan ice. It is assigned here to the Hatton Till Formation of the Logie-Buchan Drift Group. Both the matrix, and elements of the clast assemblage, are derived mainly from deposits that crop out on the sea bed to the east and south of Buchan. The distinctive reddish brown, calcareous matrix of the diamictons contains rich palynomorph assemblages (information from R Harland, BGS, 1980). These contain Permo–Triassic miospore species, Jurassic to Early Cretaceous dinoflagellate cysts, together with sparse Early Pleistocene cysts of Operculdinium israelianum and Tectatodinium pellitum. Additionally, striated hinge fragments of the bivalve Arctica islandica from the diamicton have yielded the following Total D/L amino-acid ratios (information from D Q Bowen, University of Wales, Cardiff, 1988):

These ratios indicate an Early Pleistocene age for the shells, which are clearly reworked. Like other units of the Logie-Buchan Drift Group (see Site 15 Ellon and Site 16 Kippet Hills) the evidence demonstrates that the matrix of the diamicton is derived mainly from fine-grained, red coloured, Permo–Triassic sedimentary rocks and Early Pleistocene marine beds (possibly the Aberdeen Ground Formation) present to the east and south of Buchan.

Jamieson (1906) was the first to note the locally darker brown matrix colour of diamicton in the Invernettie pit, together with inclusions of dark bluish clay. Similar, apparently sheared, laminae of black or dark grey clay a few centimetres thick have been recorded in the Burnhaven exposure. Furthermore, in 1980 extensive road works exposures for the Invernettie Diversion centred on [NK 120 447] revealed large rafts of very dark grey (5Y 3/1), massive, clayey, shelly, calcareous, clast-poor diamicton in units up to 2.0 m thick and over 50 m in length. These rafts were disrupted by south dipping, low-angle, thrust faults. The margins of the rafts merged over a few centimetres into the enclosing reddish brown diamictons. Darker colours predominated close to the rafts, passing outward into more vivid reddish brown colours suggesting that the rafts were being ‘digested’ into the diamicton matrix and altering its colour locally. Palynological analysis of the very dark grey diamicton forming the rafts (information from R Harland, BGS, 1980) showed it to be rich in possibly Late Jurassic to Early Cretaceous dinoflagellate cysts. Though being thrust from south to north when deposited, the grey diamicton apparently had an original provenance to the north or north-west of the site.

There are two explanations for the origin of the dark coloured rafts. They could have been reworked from outcrops of the similarly coloured Pitlurg Till Formation in the Slains area to the south (see Site 15 Ellon). Alternatively, they have been derived from a younger dark till deposit. BGS offshore Borehole 72/31 (Owens and Marshall, 1978; Evans et al., 1981), 13 km east off Peterhead, penetrated 23.65 m of grey-black ‘boulder clay’ beneath a veneer of pebbly, shelly sand. This diamicton sequence yielded Quaternary miospores and microplankton accompanied by a rich assemblage of reworked Late Jurassic and Early Cretaceous palynomorphs indicating a likely north-west provenance for the deposit. While the sequence is not directly dated, it is possible that part of the deposit could be correlated with the late-Late Devensian Essie Till Formation of the Banffshire Coast Drift Group (Table 7). This unit crops out north of Peterhead, but it is unclear how far south the ice responsible for its deposition penetrated to the east of Peterhead.

BGS boreholes to the west of Peterhead (McMillan and Aitken, 1981) also locally record thin bands of greenish grey to dark grey laminated silty clay within stiff red brown diamicton e.g. [NK 04 NE 10]. This suggests that glacitectonic incorporation of dark coloured clay into Hatton Till Formation diamicton units is quite widespread. This evidence of the incorporated rafts and ‘digestion’ of dark clays into the otherwise reddish brown matrix of the Hatton Till Formation around Peterhead indicates that it is essentially a thick glacitectonic–deformation till complex (see also Site 10 Ugie Valley and Site 17 Errollston).

Unit 5

The gravels and sands overlying the Hatton Till Formation are assigned here to the Kippet Hills Sand and Gravel Formation of the Logie-Buchan Drift Group although they are not as shelly as equivalent deposits farther south. They represent outwash from Logie-Buchan ice after its withdrawal to the east and south of the area.

Summary

Recent interpretation of the glacial deposits exposed in the Sandford Bay sections broadly follows that of Jamieson in the 19th and early 20th centuries. Early expansion of ice from inland to, and beyond, the present coast line was succeeded by the deposition of a glaciolacustrine (or possibly glaciomarine) sequence that probably received sediment both from the westward retreating East Grampian ice sheet and northward advancing Logie-Buchan ice. Subsequently the Logie-Buchan ice sheet advanced onshore and deposited a thick deformation till complex derived from Permo–Triassic red beds, Early Pleistocene marine sediments and pre-existing dark grey, shelly, clayey diamictons. The ice advanced over 5 km inland into the lower Ugie valley, west of Peterhead, damming meltwater to create an extensive proglacial lake (Lake Ugie). All of the deposits exposed in the Sandford Bay sections are assigned to the Main Late Devensian glaciation (OIS 2).

Site 13 Windy Hills

Quarries in the Windy Hills area (Map 5) provide outstanding exposures of deeply weathered pre-Quaternary gravels that give a unique insight to the geological and geomorphological evolution of Buchan during the Tertiary. Primary sedimentary structures within the gravels indicate their deposition by a river that flowed north-eastwards across an exposed Neogene land surface (Chapters 3 and 4). Quaternary modification of the deposits has been generally limited to minor glacial erosion, local deposition of a thin overlying till and to subsequent periglacial churning (Cryoturbation).

The white quartz and quartzite gravels that crop out in the vicinity of Windy Hills (Figure A1.17), about 12 km south-east of Turriff, are thought to be of early Pleistocene or Neogene age (Plate 4a). They have been assigned to the Windy Hills Gravel Member of the Buchan Gravels Formation (Chapter 4). Similar deposits occur at Dalgatty (Hospital) Wood [NJ 735 460] and Delgaty [NJ 744 508]. Type sections of the Windy Hills Gravel Member occur in a sand and gravel pit ((Figure A1.17)a, Pit 1) [NJ 800 400] (Clapperton, 1977; McMillan and Merritt, 1980, Auton, 2000) and in BGS Borehole NJ73NE/2 (Merritt, 1981), sited adjacent to the pit. The borehole proved 11.3 m of predominantly quartzite gravel, interbedded with pale yellow, clayey pebbly sand, overlying deeply weathered and kaolinized Dalradian schistose pelite. Reference sections have been recorded in a small gravel pit (Pit 2) [NJ 802 401] (Koppi and Fitzpatrick, 1980); at Pit 3 [NJ 793 393] (Kesel and Gemmell, 1981; Gemmell and Auton, 2000) and in BGS Borehole NJ73NE/1, about 400 m north-north-west of Mosslip (Merritt, 1981).

Spreads of quartzite and vein-quartz pebbles around Windy Hills had led previous workers to conclude that the gravel deposit caps a low, north-east-orientated ridge (about 2 km in length) overlooking the River Ythan. Recent trial pits and Ground Probing Radar (GPR) traverses in the area (Clapperton and Gemmell, 1998; Greenwood et al., 1995; Gemmell and Stove, 1999) indicate that in situ gravel is present only beneath the high ground at the south-western and north-eastern ends of the ridge. The quartzite and quartz pebbles on the lower ground constitute a veneer of soliflucted material (head) capping till (Figure A1.17). In particular, GPR traverse 1 (Greenwood and Raines, 1994; Greenwood et al., 1995), aligned across the axis of the ridge in the shallow col, between the two main gravel outcrops, showed no signatures characteristic of the Windy Hills Gravel Member throughout its 756 m length. The reflectors were generally more characteristic of till deposits containing scattered large boulders. Similar responses were evident at the southern end of traverse 5, but channelled gravels were imaged at the northern end of the traverse and in adjacent traverse 3.

Both the in situ and soliflucted spreads of the Windy Hills Gravel Member contain quartzite clasts that are generally comparatively fresh and bear ‘chatter’ or percussion marks. In contrast, scattered cobbles of granite and schistose metamorphic rocks within the gravel are normally decomposed to kaolinitic sand. Logs of BGS boreholes NJ73NE/1 and NJ73NE/2 indicate that decomposed fragments of underlying schistose strata are commonly incorporated into the basal parts of the unit, suggesting that most of the weathering of both the in situ gravel and underlying strata has occurred since the gravel was deposited. Flint clasts are also present in the deposit, although they generally comprise less than 1 per cent of the gravel fraction; some are weathered with a thick, dull grey or white patina, which is common in flints from the Chalk. Rare clasts of Lower Cretaceous chert have also been recorded (Flett and Read, 1921) and Jamieson (1865) reported the presence of ‘chalk-fossils’ within some of the flints; unfortunately he gave no specific identifications and his observations have not been confirmed by subsequent investigations.

The matrix of the gravel contains two types of quartz sand grains (Hall, 1983). Predominant are angular grains that show abundant evidence of breakage and conchoidal fracture and only limited edge rounding. Kesel and Gemmell (1981) suggested that these features were indicative of glacial transport. Less abundant are rounded grains with surface textures indicative of environments of high silica mobility, with deep etch pits and smooth precipitation surfaces. These latter surfaces are indented by crescentic chocks and coalescing impact pits, which Hall (1982) attributed to a later phase of highenergy subaqueous transport. A few of the rounded grains have been broken subsequently. The dominant heavy mineral in the sands is ilmenite, with subsidiary staurolite, andalusite, zircon and garnet (Hall, 1983). The gravel and sand is bound in places with white silt and clay, comprising mainly b-axis disordered kaolinite with minor illite. Discrete beds of silty sand also occur.

North-east outcrop

Primary sedimentary structures are variably developed in the in situ gravel in the Windy Hills type area. Horizontally interbedded gravels and sands are illustrated by Kesel and Gemmell (1981, fig.3 F and G) and Clapperton (1977, fig.15). Cross-bedding, dipping at about 20° towards the north-east was recorded by McMillan and Merritt (1980) from a gravel pit [NJ 800 400]. Imbrication of gravel clasts within the cross-sets was seen to dip at about 40° towards the south-west and cross-bedding dipping at about 20° towards the north-west was recorded in a small gravel working [NJ 8035 4015]. Interstratification of horizontal and cross-bedded units is clearly displayed in the GPR traverse 4 ((Figure A1.17)c), aligned west to east and adjacent to the type section north-north-east of Windyhills. The radar penetrated the gravel to a depth of about 14 m.

Clast imbrication recorded by Clapperton (1977) was taken to indicate that the gravels were laid down by water that flowed east-north-eastward (roughly parallel to the present line of the ridge), a view supported by McMillan and Merritt (1980).

South-west outcrop

A series of benches is developed towards the south-western end of the body of Windy Hills Gravel Formation that crops out north of Mosslip. The surfaces of the benches, which slope gently south-eastwards are dissected by a series of narrow glacial drainage channels, apparently up to about 5 m deep, that trend north-westwards. A trial pit ((Figure A1.17)a, Trial Pit C) on the floor of one of the channels showed a stratified sequence of disturbed quartz and quartzite gravel beneath thin peaty soil. The stratification within the disturbed gravel dipped at shallow angles towards the channel axis, indicating that the gravel in the floor of the channel probably accumulated by mass movement (slippage and solifluction) from the channel sides. It also suggests that each channel may be considerably deeper (more than 5 m) than its present subdued surface expression implies.

Large angular clasts of relatively fresh metamorphic rocks have been recovered from beneath kaolinised quartz and quartzite gravel exposed in trial pit A near to the western edge of the south-western outcrop of the Windy Hills Gravel Member ((Figure A1.17)a). The pit, which was sited on the lowest bench, penetrated apparently in situ gravel to a depth of more than 2 m, overlying reddened diamictic gravel. The gravel contained fresh, angular metamorphic clasts, some of which appeared to be striated. This suggests that the gravel at the western end of the outcrop may include some glacially reworked material.

Sunnybrae outcrop

A small outcrop of quartz–quartzite gravel has been mapped adjacent to the Sunnybrae Centre at the western end of the Windy Hills area. Whether this deposit is essentially in situ or soliflucted is unclear on present evidence, but it is of sufficient thickness to have been formerly worked in a small gravel pit.

Interpretation

The gravel exposed in a working at [NJ 793 393] appears to be in situ. Kesel and Gemmell (1981) interpreted ‘planar foreset’ bedding and imbrication from the gravel pit as indicating transport from the west. However, GPR traverse 2 ((Figure A1.17)b) aligned north-westwards and adjacent to the pit, which penetrated the gravel to a depth of about 10 m, shows the true nature of the cross-bedding at the site. The GPR profile indicates the presence of large-scale, shallow (more than 10 m wide and up to 2 m deep), trough cross-stratified units with the axes of the troughs apparently aligned south-west to north-east.

Preservation of well-developed stratification and imbrication, as well as large-scale cross-bedding seen on many of the GPR profiles, from both outcrops of the Windy Hills Gravel Member implies little disturbance of the bulk of the deposit since its deposition. Nevertheless, the soliflucted gravel infilling glacial drainage channels and the presence of striated clasts indicate possible glacial and glaciofluvial reworking of parts of at least the western end of the south-west outcrop. Reworking of the upper parts of the in situ sequence by periglacial activity is also widespread in both outcrop areas. This is indicated by a well-defined layer of cryoturbated material, typically between 0.6 and 1.0 m thick, that caps many of the exposures. This layer is best developed where the gravel crops out at the surface and has been described by FitzPatrick (1975b, c), Gemmell and Kesel (1979) and McMillan and Merritt (1980). Cryoturbation features include the widespread and pervasive development of an erect clast fabric within the top of the gravel. This fabric was seen to truncate an ice-wedge cast which extended for 1 m into the underlying gravel at the type section in the gravel pit [NJ 800 400] (McMillan and Merritt, 1980). Clasts commonly fine upwards within the cryoturbated layer of gravel and become more numerous towards the top of the unit (Fitzpatrick, 1975b, c). Many of the cobbles have a thin layer of silt capping their upper surface. These silt cappings have been interpreted as the infilled remnants of voids, left following melting of sheaths of ice that formed when the sediments were affected by permafrost activity (Fitzpatrick, 1987). At Woodhead [NJ 788 384], a peat bed lying beneath 1.5 m of soliflucted till has yielded a radiocarbon date of 10 780 ± 50 BP (SRR–1723) (Table 8), implying that significant periglacial activity took place in this area during the Loch Lomond Stadial (Connell and Hall, 1987).

Where the Windy Hills Gravel Member is overlain by thin spreads of ‘moderate brown’ till, as in the gravel pit type section (Bremner, 1916; Kesel and Gemmell, 1981; Clapperton and Gemmell, 1998) unweathered erratics have been incorporated locally into the upper parts of the gravel, probably by frost churning (Clapperton, 1977). The clasts within the till also display a pronounced erect fabric indicating that much periglacial activity postdates the last glacial event preserved at the site. The till includes sparse unweathered striated pebbles of metamorphic and igneous rocks, but most of the clasts are of quartz and quartzite derived from the underlying gravel. The presence of the till close to the highest point of the north-western outcrop of gravel indicates that ice probably overrode the whole of the outcrop of the Windy Hills Gravel Formation at least once during the Quaternary. The precise age of this glaciation is not known.

Although recent interpretations are agreed that the Windy Hills Gravel Member is primarily of fluvial origin, Jamieson (1858, 1865) originally suggested that the deposit was locally derived and of preglacial marine origin. However, he later proposed a glacial derivation from the floor of the Moray Firth (Jamieson, 1906). Wilson (1886) concluded that the gravel was a residual deposit from a denuded Chalk cover and that it had been glacially reworked. Flett and Read (1921) suggested that the gravel outcrops were remnants of formerly more extensive marine deposits resting on an ancient land surface. The more recent evidence, cited above, clearly suggests that most of the in situ Windy Hills Gravel Member was laid down by a pre-Quaternary river and that subsequent erosion has lowered the adjacent landscape, preserving the deposits in their present hill-top locations (Figure 20). The nature of the cross-stratification evident in the GPR profiles supports the interpretation that the bulk of the Windy Hills Gravel Member in its type area is a fluvial deposit. It was probably laid down by waters that flowed north-eastwards across an exposed Neogene land surface.

If the evidence of breakage of grains cited above is accepted as implying glaciofluvial transport, the degree of alteration is demonstrably less than seen in the Denend Gravel Formation at Leys Quarry (see Site 7 Kirkhill and Leys). The deposition would have had to have occurred in the early Pleistocene, perhaps the Baventian (Table 1), in order to have allowed sufficient time (over half a million years) to weather the gravels under temperate conditions.

The character and depth of the weathering that affects the deposit also indicates that during the Quaternary, modification of the bulk of the sequence has been limited to glacial erosion, deposition of overlying till and subsequent periglacial churning. Consequently the deeply weathered gravel at Windy Hills preserves unique evidence concerning the long-term evolution of the landscape in north-east Scotland, both before, during and after the Quaternary ice ages.

Site 14 Moss of Cruden (‘Buchan Ridge’)

The Moss of Cruden, 11 km south-west of Peterhead, is an area of unusual geological interest (Hall, 1993a) (Figure A1.18). Together with the Moss of Auquharney, the Corse of Balloch and the Hill of Aldie, it forms part of the so-called ‘Buchan Ridge’ (Flett and Read, 1921) ((Figure 22), (Map 7)), a gently rounded, featureless, east-north-east-trending ridge that attains an elevation of about 140 m above OD. It provides the type localities for the Smallburn Sandstone, a small outlier of Devonian sandstone, the Moreseat Sandstone, a concealed outlier of Early Cretaceous age, and the Buchan Ridge Gravel Member of the Buchan Gravels Formation, of Neogene age. It also includes the important Late Pleistocene interstadial site at Camp Fauld. These sites are described in turn.

Smallburn Sandstone Formation

The Smallburn Sandstone Formation occupies a small area on the northern flank of the Moss of Cruden [NK 023 403]. Exposure is poor, but shallow pitting indicates that the sandstone rests on diorite and is partially concealed beneath Lower Cretaceous Moreseat Sandstone (Figure A1.18). The Smallburn Sandstone is a red-brown arkose, whose grains comprise dominantly quartz, with sudsidiary feldspar and lithic fragments of diorite and granite. Quartz and hematite coatings occur on grain surfaces, but development of quartz overgrowths is limited. Preserved porosity varies from 13 to 19 per cent. Clay mineralogy is dominated by interstratified mica-smectite, with some illite (information from R A Downie, Downie Geoscience Ltd, 62 Durwood Avenue, Shawlands, Glasgow). No palynomorphs appear to be preserved in the formation (information from L A Riley, Bovingdon, Herts, 1997).

Moreseat Sandstone Formation

Weathering and glacial disturbance of the Moreseat Sandstone Formation is widespread and it is only in the ground north of Moreseat that relatively fresh material was recovered from the base of trial pits (Figure A1.18). The Moreseat Sandstone here is a very dark grey, porous silty fine-grained glauconitic quartz arenite. Bioturbation is common, and burrows and moulds of mollusca are present. The rock contains well-preserved palynomorphs indicating a Late Hauterivian to Early Barremian age (information from

W Braham). Thin sections reveal a spicular quartz sandstone with a mixed terrigenous/calcareous mud matrix. Grains are dominantly of quartz, but, prior to the development of voids (moulds) after the dissolution of mollusc fragments and foraminifera, the dominant grain type would have been sponge spicules. Feldspar grains and bioclastic debris have also undergone significant dissolution. Macroporosity is around 30 per cent (information from A R Duncan, Maud, Aberdeenshire, 1997).

The Lower Cretaceous rocks north of the Moss of Cruden were originally fine-grained sandy siltstones, but they are typically decomposed to clayey fine-grained sands and sandy silty clays of vivid orange, tangerine, yellow and white hues.

Buchan Ridge Gravel Member

The Moss of Cruden is the type area of the Buchan Ridge Gravel Member, part of the Buchan Gravels Formation (compare with Flett and Read, 1921; McMillan and Merritt, 1980) (Figure 22). Exposures also occur at Skelmuir Hill (Bridgland et al., 1997), Whitestones Hill (Kesel and Gemmell, 1981) and Den of Boddam (Bridgland et al., 1997), and other concealed, small gravel masses may well exist (Flett and Read, 1921). At Moss of Cruden, the gravel deposit forms a broad ridge, orientated north-east–south-west, some of the highest ground in this part of Buchan. The type section of the Buchan Ridge Gravel Member occurs in BGS Borehole NK04SW/3 on the highest point of the ridge (McMillan and Merritt, 1980) (Figure A1.18), which proved at least 25 m of coarse flint and quartzite gravel, interbedded with pebbly sand and commonly bound by white kaolinitic clay. Reference sections have been recorded in temporary pits at Moss of Cruden (BGS records), Skelmuir Hill [NJ 987 413] (Bridgland et al., 1997), Boddam [NK 113 415] (Bridgland et al., 1997) and in BGS Borehole NK04SE/6, near Berryley (McMillan and Merritt, 1980).

The Buchan Ridge Gravel Member is poorly exposed. Temporary excavations and boreholes show white, clay-bound coarse gravels with minor sand and silt units (Merritt, 1981). Gravel clasts are dominantly flint with subsidiary quartzite and vein-quartz. Clasts are generally well-rounded pebbles and cobbles and bear numerous chatter marks. The deposits originally contained less resistant clasts, probably mainly granite, but these have decomposed to ghosts of white sandy clayey silt. This decomposition extends throughout the known thickness of the deposit into the underlying granite and gneiss bedrock. Both sand and gravel units are bound and, in places, supported by, white, sandy clayey silt consisting of well-ordered kaolinite with minor illite (Hall, 1982).

The base of the Buchan Ridge Gravel Member in places contains large nodules of flint and boulders of quartzite (Plate 5a), (Plate 5b). Quartzite blocks up to 30 cm in diameter were recovered from a trial pit close to the western margin of the gravels at Moss of Auquharney [NK 019 400]. In pits at an elevation of approximately 130 m [NK 028 403] the base of the gravels revealed a lag of unworn, nodular flint with blocks up to 25 cm in diameter. In places the gravel is partially cemented by iron concretions containing goethite. Similar deeply weathered boulder gravel containing large nodular flints occurs on Skelmuir Hill at an elevation of about 148 m OD (Bridgland et al., 2000). The Buchan Ridge Gravel Member rests on kaolinised granite and metasedimentary rock (McMillan and Aitken, 1981; Hall et al., 1989) and on highly weathered Lower Cretaceous Moreseat Sandstone (Hall and Jarvis, 1994).

A ground probing radar survey (Greenwood et al., 1995) on the northern slope of the Moss of Cruden has confirmed that the feather edge of the Buchan Ridge Gravel Member rests on weathered Lower Cretaceous sandstone (see Appendix 3). Within undisturbed parts of the Buchan Ridge Gravel Member, low-angle cross-stratification dips towards the axis of the ridge ((Figure A3.3); line 1). Resistivity depth probes and ground probing radar indicate that around 15 m of Buchan Ridge Gravel Member lies beneath the crest of the ridge, resting on weathered granite. The gravels here fill a channel running north-east–south-west and parallel to the line of the ridge.

The Buchan Ridge Gravel Member is overlain by a variable thickness of till, and disturbance attributed to overriding ice is common. At Corse of Balloch, BGS Borehole NK04SW/4 (McMillan and Merritt, 1980) indicates that in situ Buchan Ridge Gravel Member is overlain by almost 10 m of till, derived in large part from the underlying gravel. Temporary excavations east of Camp Fauld revealed rafts of flint and quartzite gravel within till (Whittington et al., 1993). To the north of the ridge, a sheet of soliflucted flint gravel, up to 2 m thick, extends well downslope on to weathered granite (Hall, 1993a). This gravel forms part of the ‘Flint-quartzite Head’ deposits described in Chapter 6 that are widespread around the Buchan Ridge on sheets 87W Ellon and Peterhead 87E Peterhead (Maps 6 and 7).

Camp Fauld

This site [NK 049 410] lies on the southern slope of the Moss of Cruden, about 900 m north-west of Moreseat ((Figure A1.19), (Map 7)). It is important for the presence of a buried peat of probable Early Devensian age that separates two till units (Whittington et al., 1993). The lithostratigraphy described below generally follows that set up by Whittington et al. (1993), but in order to adhere more closely to internationally agreed guidelines on lithostratigraphical nomenclature, new names have been adopted here ((Table 7); A1.8). Those proposed recently by Sutherland (1999) are not appropriate.

The stratigraphy is known only from trial pits. The oldest deposit recognised is the Camp Fauld Till Formation. This is at least 2 m thick and is a grey (2.5Y N5) to very dark grey (7.5YR N3) massive to crudely horizontally bedded diamicton with sparse pebbles. Clast lithologies comprise mainly rounded quartzite and flint, with lesser amounts of subangular granite, basic igneous, red sandstone and soft grey siltstone clasts. The matrix is a micaceous sandy silt and clay mineralogy is largely kaolinite, illite-smectite and minor glauconite. The diamicton rests on, and locally passes down into, and incorporates rafts of Moreseat Sandstone. The direction of ice flow during deposition of the Camp Fauld Till is difficult to establish owing to the dominance of very locally derived material. The presence of clasts of grey granite and red sandstone suggest transport from local outcrops to the west of the site.

The Berryley Peat Bed is up to 0.5 m thick in pit A and rests on the mottled top of the Camp Fauld Till (Figure A1.19). The peat is truncated by, and both reworked and sheared into, the lower 30 cm of the overlying Aldie Till. In Pit B, the peat bed rests on bedrock and is overlain by sands and gravels containing stringers of peat. This unit is separated from the overlying Aldie Till by a gravel-rich diamicton with a strong downslope clast fabric (Figure A1.19). This intervening deposit, named here as the Hardslacks Gelifluctate Bed, appears to mark a phase, or phases, of periglacial conditions and slope instability.

The Aldie Till at Camp Fauld is a greenish grey (5GY 6/1) sandy diamicton that is massive or shows a crude horizontal bedding and is up to 2.5 m thick. Clast lithology is dominated by broken rounded pebbles and cobbles of flint and quartzite reworked from the Buchan Gravels Formation, with sparse mica-schist, granite, basic igneous and red sandstone clasts. The matrix is kaolinitic quartz sand. Away from the type-site, the character of the Aldie Till varies greatly in response to marked lateral changes in bedrock and even drift geology. On the Buchan Ridge Gravels at Hill of Aldie, the Aldie Till closely resembles that at Camp Fauld. However, in the area underlain by the Moreseat Sandstone Formation, it is indistinguishable from the underlying Camp Fauld Till. To the north and east of Camp Fauld, the Aldie Till partly incorporates large masses of weathered brown sand with sparse quartzite clasts of unknown derivation and origin. This sand, named here as the Moreseat Farm Sand Bed, was also probably formerly exposed at an infilled gravel pit [NK 0529 4108], from where a cobble of brown, porous arkose of Lower Cretaceous appearance was recovered in 1979. BGS Borehole NK04SE/2 shows that this unit of sand is up to 7 m thick, and rests on a dark grey till that may be equivalent to the Camp Fauld Till.

South and east of Moreseat, the Aldie Till passes beneath red clayey diamictons of the Logie-Buchan Drift Group. It thickens on the southern and eastern slopes of the Moss of Cruden in accord with the apparent eastward carry of flint and red granite from the Moss of Cruden, and its weak north-west–south-east clast fabric. These features indicate that the unusual till thickness is probably a product of only minor subsequent glacial erosion as it lies in the lee of the Buchan Ridge. The presence of red granite erratics on the Moss of Cruden has been used to infer ice movement from the north-east prior to the last glaciation (Bremner, 1934a, 1939; Synge, 1956). However, the situation is complicated by the presence of coarse-grained red granites, similar to the Peterhead Granite, in the ground 1 km north and west of Moreseat.

The Berryley Peat Bed has yielded radiocarbon ages (SRR–3724) of 39 590 ± 550 BP for the humic fraction and 34 600 ± 910 BP for the humin fraction (Whittington et al., 1993) and luminescence ages of 160 ± 16 and 251 ± 24 ka BP (Duller et al., 1995). The significance of these contrasting ages is uncertain. Pollen analysis of peat layers from the adjacent pits suggests that the peat formed during an interstadial in which birch–pine woodland first developed, probably followed by tundra vegetation in which Bruckenthalia spiculifolia was present. This heath is found today in the Balkan Mountains, but is extinct in Scotland (Whittington, 1994). The vegetation assemblage of the Berryley Peat, including the presence of Bruckenthalia, is akin in certain aspects to that found at other Early Devensian sites in Scotland at Allt Odhar and Sel Ayre. Hence, the Berryley Peat may correlate with Oxygen Isotope Stage 5c or 5a, when pine–birch woodlands developed widely in north-west Europe during the Brørup and Odderade Interstadials (Behre, 1989) (Figure 7).

Site 15 Ellon, Bellscamphie

In the Leask area, to the north-east of Ellon, there are three lithologically distinctive till units belonging to each of the main drift groups in the region (Maps 6 and 7). Two of these tills may predate the Late Devensian. Several sites, including ‘Bellscamphie’, provide evidence of the pattern and timing of movement of three major ice masses affecting eastern Buchan (Figure 50).

In 1906, T F Jamieson described several localities in the Ellon area at which a distinctive shell bearing ‘indigo boulder clay’ was found underlying younger glacial and glaciofluvial deposits (Figure A1.20). Jamieson regarded the ‘indigo boulder clay’ as the oldest glacial deposit in the Ellon area and suggested deposition by ice moving from the Moray Firth. All of Jamieson’s original sections have long been obscured (Hall and Jarvis, 1993b). Recent investigations, notably a programme of shallow pitting, have led to a revision of the glacial stratigraphy in this area (Hall and Jarvis, 1993b; Hall and Jarvis, 1995).

Lithostratigraphy at Bellscamphie and Pitlurg

Mechanical excavations into the sides and floor of the abandoned railway cuttings at Bellscamphie [NK 019 337] and Pitlurg [NK 021 346]

(Figure A1.20) have shown the existence of four lithostratigraphical units (Table A1.9). The nomenclature set up by Hall and Jarvis has been formalised here and extended. It does not follow that set up by Sutherland (1999), who replaced the Bellscamphie Till with the Elton Till (presumably Ellon Till).

Bellscamphie Till Formation

The oldest unit identified is the Bellscamphie Till, which rests on Ellon Gneiss. It is massive, overconsolidated and the clasts, which are mainly subangular, are dominated by psammite derived from local sources. Flint clasts are conspicuous and were very probably transported from outcrops of the Buchan Ridge Gravel Member to the north and north-west, especially as the clay mineralogy of the till matrix is dominated by kaolinite. Sparse palynomorphs of probable Mesozoic age also have been recovered from the matrix of this till.

Pitlurg Till Formation

This unit is equivalent to Jamieson’s ‘indigo boulder clay’ and overlies either the Bellscamphie Till or bedrock. The slightly calcareous clay-silt matrix of the diamicton is massive and overconsolidated with varied clay mineralogy. Palynological preparations of the matrix (information from W Braham, Hemel Hempstead, Herts, 1990) have yielded rich assemblages of late Oxfordian to Ryazanian dinoflagellate cysts indicating a provenance for the dark grey clay matrix in the Kimmeridge Clay Formation of the Moray Firth. Other components of Callovian and mid-Barremian age would also support a north-north-west to north-west provenance. Clast lithologies include large proportions of dark mafic igneous and metamorphic rocks. Small numbers of red-brown sandstones of Devonian or Permo–Triassic age, together with a distinctive white (possibly) Mesozoic sandstone, could have been derived either from the north, or from the bed of the North Sea to the east. At Pitlurg, the Till shows a weak north–south fabric (Hall and Jarvis, 1995).

Very sparse, small, abraded shell fragments are present in the till. Amino-acid D/L ratios of shell in the Pitlurg Till at Bellscamphie are 0.272 (LOND–590) and 0.212, 0.101 (LOND–589). The range of ratios indicates a mixed population of shells of different ages. Correlation of the Arctica sp. ratios with the D/L sea level events recognised by Bowen and Sykes (1988) indicates that the shells are of OIS 7 age or younger, with the single D/L ratio of 0.101 indicating an age equal to, or younger than, OIS 5e.

Kippet Hills Sand And Gravel Formation

The clast composition of the Kippet Hills Sand and Gravel Formation includes psammite similar to that found around Collieston. Distinctive components of

the assemblage are limestones/dolomites and mudstone clasts of Permian and Neogene age and possibly also Mesozoic provenance. These lithologies indicate transport from the bed of the North Sea to the south and east (see Site 16 Kippet Hills).

Hatton Till Formation

The distinctive red-coloured Hatton Till rests with a sharp planar contact on underlying units. Locally, thin beds of till are interstratified with the Kippet Hills Sand and Gravel Formation suggesting deposition by debris flow processes. However, at the south-west end of the Bellscamphie cutting, interbedded red, grey and brown diamictons occur indicating probable glacitectonic reworking of older deposits into the base of the Hatton Till. The clast lithologies in the diamicton are similar to those of the Kippet Hills Sand and Gravel Formation with which it shares a provenance.

Discussion

The Hatton Till and Kippet Hills Sand and Gravel Formation are typical representatives of the Late Devensian Logie-Buchan Drift Group. It is confirmed that Jamieson’s ‘indigo boulder clay’ (Pitlurg Till) forms an extensive stratigraphical unit concealed beneath deposits of the Logie-Buchan Drift Group in the Ellon area. Boreholes, trenches and shallow pits indicate that the Pitlurg Till extends northwards from Slains to Auchleuchries (Figure A1.20). The deposit may extend north-eastwards beyond Hatton as boreholes there show thick, stiff, dark grey clayey tills lying beneath deposits of the Logie-Buchan Drift Group (Merritt, 1981).

The Pitlurg Till lies immediately beneath deposits of the Logie-Buchan Drift Group in the Bellscamphie area, but the situation is more complicated in the surrounding area. For example, at Ellon, Jamieson (1906, p.22) recorded up to 6 m of grey till derived from the west lying between his ‘indigo boulder clay’ and ‘Red Clay’. Stream sections near Lintmill Bridge [NJ 988 308], east of Ellon, show Hatton Till overlying grey till with a sandy matrix and a strong west–east macrofabric. At Bearnie [NJ 967 339], some 4 km north of Ellon, 1 m of Hatton Till overlies a dark grey sandy till up to 2.7 m thick. This till, named here as the Bearnie Till Member of the Hythie Till Formation, has a multimodal clast fabric, but the clast composition suggests a local provenance. Furthermore, the till matrix has yielded a rich dinoflagellate cyst assemblage dominated again by late Oxfordian to Ryazanian elements derived from the Kimmeridge Clay Formation in the Moray Firth. These Jurassic palynomorphs are probably derived by local reworking of the nearby Pitlurg Till.

At two sites north-east of Pitlurg [NK 048 363; NK 037 356], road improvements have revealed up to 3.5 m of brown to dark brown till overlying metamorphic bedrock and with clasts predominantly derived from it. This till also contained common rounded flint clasts derived from outcrops of the Buchan Ridge Gravel Member to the north and has a strong north-north-west–south-south-east clast fabric. The lithology and unusual fabric characteristics of this till suggest that it is a correlative of the Bellscamphie Till. As the till is overlain directly by sands and gravels of the Kippet Hills Sand and Gravel Formation, other units in the local succession must have been eroded away or never deposited. Part of the succession is probably also absent in sections in the immediate vicinity of Bellscamphie where Hatton Till rests on Pitlurg Till with no intervening grey till of western provenance (Bearnie Till). This is likely because the Hatton Till at Bellscamphie apparently incorporates material from older diamictons.

The stratigraphy at the former gravel pit at Tillybrex [NK 002 348], 2 km north-west of Bellscamphie (Figure A1.20), is also of interest (Hall and Jarvis, 1995). Thin red diamicton and clay rests on brown diamicton of inland derivation, up to 11 m of gravel and a basal yellow-brown till of westerly derivation with psammite and flint clasts (Merritt, 1981). The gravel, named here as the Tillybrex Gravel Formation (Chapter 8, (Figure 50)) showed a range of sedimentary structures indicating flow towards the west, and included a number of syngenetic ice-wedge casts, indicating deposition under permafrost conditions. The gravel included many badly weathered clasts (Merritt, 1981) and was locally heavily cemented by iron and manganese oxides, suggesting a period of weathering after deposition (Jones and Milne, 1956). The stratigraphy at Tillybrex leaves little doubt that there are at least two tills of inland derivation in the Ellon district and that deposition of these tills was separated by a period of interstadial and perhaps interglacial conditions.

The provenance of the Pitlurg Till is uncertain. The presence of Jurassic palynomorphs suggests transport from the Moray Firth and this accords with the weak northerly fabric of the till at Pitlurg. Transport directly from the north is unlikely, however, as any ice movement from this direction would be across the flint gravels of the ‘Buchan Ridge’ and the till contains little flint. Transport via the Ythan valley, as suggested by Jamieson (1906) is unlikely, as no occurrences of dark grey clayey diamictons derived from the Moray Firth are known between Turriff and Ellon. A component of movement onshore from the bed of the North Sea is suggested by the small proportions of Devonian and Mesozoic clasts within the till.

The only information on the age of the Pitlurg Till is provided by amino-acid ratios of contained shell fragments. As argued above, these data suggest that the till is no older than Ipswichian in age and do not rule out formation during an early phase of the Late Devensian, in which case it is probably equivalent to the Whitehills Glacigenic Formation (Table 7). More dates are needed to resolve the age of this deposit. The ‘indigo boulder clay’ has traditionally been correlated with other dark grey clayey tills and erratics found beneath, or incorporated into, tills of inland derivation and probable Devensian age (Jamieson, 1906; Hall, 1984a). An OIS 6 age has been suggested for these deposits (Hall and Connell, 1991). One of them, the Benholm Clay Formation, at Burn of Benholm (Site 26), is almost certainly pre-Devensian in age. Thus it would seem that blue-grey tills were laid down both in the Late Devensian and also before the Ipswichian, probably in OIS 6.

The Bellscamphie Till rests on bedrock and is the oldest till in the immediate area. It was deposited by ice moving from the north-west or west. Possible correlative deposits in the Ellon area are the thin tills of inland derivation that rest on bedrock and underlie dark grey clayey till east of Hatton and underlie the Tillybrex Gravels at Tillybrex (Merritt, 1981) (Figure A1.20). Farther afield, it probably correlates with one of the two pre-Devensian tills of inland derivation at Kirkhill (Hall and Connell, 1991) and with the Camp Fauld Till on the Moss of Cruden (Whittington et al., 1993).

Site 16 Kippet Hills, Slains

The Late Devensian sands, gravels and tills at Kippet Hills provide important evidence concerning the direction of movement of ice that laid down the Logie-Buchan Drift Group and the pattern of deglaciation in the coastal lowlands north of Aberdeen. In particular, the presence of marine shells of Early Pleistocene age, together with clasts of sandstone, dolomite, limestone and shale in the Kippet Hills Esker (Plate 15), indicate that much of the glaciofluvial gravel was initially derived from offshore.

The Kippet Hills area, 7 km east of Ellon on Sheet 87E ((Map 7); (Figure 50)), includes ridges and hummocky terrain underlain by the Kippet Hills Sand and Gravel Formation of the Logie-Buchan Drift Group (Figure A1.21). These landforms and deposits have been the focus of research and controversy for more than 150 years (Gordon, 1993b). In particular, the presence of unusual clasts, such as pebbles of Cretaceous flint and Jurassic limestone, as well as shells of Early Pleistocene marine molluscs in the sands and gravels have resulted in a variety of interpretations of their origin. They have been explained as in situ Crag deposits of marine origin, formed during the late Tertiary (Jamieson, 1858), as marine deposits ploughed up by ice into their present topographic form (Jamieson, 1865) and as glacigenic deposits (Jamieson, 1882a). Wilson (1886) suggested that the sands and gravels were interglacial marine beds, equivalent to those at Clava and King Edward, but that red clays overlying the sands and gravels were of glacial origin. Bell (1895) concluded that both the shelly gravels and the clays were glacial deposits, but Hull (1895) disagreed with Bell’s interpretation and invoked marine submergence to explain both the clays and the sands and gravels.

It is now generally accepted that the Kippet Hills area contains landforms, including esker ridges, kames, terraces and kettleholes, that were formed by glaciofluvial deposition during the melting of the Late Devensian ice sheet (Gemmell, 1975; Merritt, 1981; Smith, 1984; Hall and Jarvis, 1995). The landforms and sediments provide important clues as to the directions of ice movement and the pattern of deglaciation in the coastal area to the north of Aberdeen.

The most striking landform in the area is the Kippet Hills Esker (Figure A1.21) which trends approximately south to north and is up to 15 m high. The esker ridge passes northwards into an undulating, triangular-shaped feature at Knapsleask composed of sand and gravel and interpreted by Smith (1984) as a kettled delta. Exposures in the presumed delta have revealed up to 12.2 m of shelly gravels and sands at Whitefields Sand Pit (NK03SW/12) (Merritt, 1981). These deposits display bedding dipping at low angles to the north and north-east; trough cross-beds and clast imbrication also indicate that palaeocurrents flowed in that direction. Exposures in the southern flank of the landform showed a red clayey diamicton draping a steep slope. It is likely that this is a former ice-contact slope draped by debris flow diamicton deposited from an adjacent ice margin. Discontinuous spreads of red till (Hatton Till Formation) were recorded as capping the sequence by Gemmell (1975) and Smith (1984). Hall and Jarvis (1995) have noted red till units interbedded with the sands and gravels, and clasts of red clayey diamicton are known from the gravels at Whitefields pit. Grain-size analysis (Merritt, 1981) of samples from the sand pit and from Borehole NK03SW/9, sited on top of the triangular feature, do not show a clear upward-coarsening sequence that would be expected if the deposit were a true delta. The sedimentological evidence suggests the feature is more likely to be an ice-contact outwash fan supplied with sediment transported through the esker tunnel from the south.

The Kippet Hills Esker was described in detail by Jamieson (1858, 1860a, 1865, 1882a, 1906), who noted its sinuous form, and the presence of clasts of igneous and metamorphic rocks, of local provenance, and clasts of a distinctive suite of more exotic provenance. The latter included significant numbers of red and grey sandstone, limestone and calcareous shale clasts, of possible Jurassic and Permian origin. Jamieson (1860a, 1882a) listed the molluscan faunas and noted similarities between the fossil assemblages in the Kippet Hills Sand and Gravel Formation and those from the Red Crag of Norfolk, although certain shells in the Scottish material were seen to be anomalous. Cambridge (1982) examined newly collected shells from the Kippet Hills area, but none of this material was confirmed as being attributable to an assemblage equivalent to that within the Red Crag. Recently, hard matrix material that formed the internal cast of an unidentified bivalve shell was recovered from gravels in the Whitefields pit and has yielded a palynological assemblage dominated by Spiniferites spp. together with the dinoflagellate cysts Achomosphaera andalousiensis and Amiculosphaera umbracula (information from L A Riley and K J Gueinn, Network Stratigraphic Consulting Ltd, Potters Bar, Herts, 1997) The latter has a range from late Miocene to Early Pleistocene (Powell, 1992; Harland, 1992) and confirms the presence of late Tertiary to Early Pleistocene reworked material within the Kippet Hills Sand and Gravel Formation.

Amino-acid (D-alloisoleucine: L-isoleucine) ratios on Arctica islandica reported by Smith (1984) and by Hall and Jarvis (1995) imply an Early Pleistocene or early Mid-Pleistocene age for these shells. Hall and Connell (1991) and Hall and Jarvis (1995) have suggested derivation of the shells from the Early to Middle Pleistocene Aberdeen Ground Formation, which occurs offshore (Stoker et al., 1985). The mixed nature of the fauna recovered from the Kippet Hills gravels is also demonstrated by the presence of the bivalve Macoma balthica, known from both modern finds and Jamieson’s published species list (Jamieson, 1882a). This mid-Early Pleistocene to present-day species is first noted in deposits of late Baventian age in East Anglia (Norton and Spaink, 1973) and suggests a younger reworked element in the fauna. Additionally, total D/L ratios measured on two Arctica sp. shells from the Whitefields pit have given values of 0.107 and 0.103 (AAL–1300; information from J T Andrews, Institute of Arctic and Alpine Research, University of Colorado, Boulder, USA, 1980). Owing to the sample preparation method employed by the laboratory at the time of the analyses, it appears likely these values are underestimates and the values should be close to 0.16 and 0.15, respectively (Miller et al., 1982). These higher D/L ratios would be typical of Arctica from the last interglacial (Ipswichian) This combined evidence indicates that the fauna recovered from the Kippet Hills deposits represents shells reworked from deposits spanning the age range late Neogene to the Ipswichian Interglacial.

Merritt (1981) showed that the stratigraphical sequence at Kippet Hills is similar to that recorded from the adjacent Ellon area (see Site 15 Ellon). BGS Borehole NK03SW/10, near the highest point on the esker ridge proved 3.6 m of Hatton Till overlying 21.0 m of well sorted sand and gravel of the Kippet Hills Sand and Gravel Formation. The borehole sequence correlates with the exposures recorded from the fan at Knapsleask described above. A stone count of gravel exposed in Whitefields pit showed that 41 per cent of the clasts were of dolomitic limestone and calcareous siltstone (possibly of Jurassic origin). High proportions of similar rock types have been recorded from the Hatton Till and Kippet Hills Sand and Gravel Formation at Bellscamphie railway cutting by Hall and Jarvis (1995) (see Site 15 Ellon). Significant numbers of light grey (10YR 7/2) dolomite clasts have recently been recorded from the gravels at Whitefields pit. This conspicuous lithology contains a rich miospore assemblage of Late Permian (Zechstein) age (information from A R Duncan, Maud, Aberdeenshire, 1996; L A Riley, Bovingdon, Herts, 1997). The most likely source for these clasts is the offshore outcrop of Permian rocks in the western North Sea basin. These finds confirm Jamieson’s original suggestion that the deposits contain clasts derived from Permian rocks.

The provenance of exotic clasts and shells from the Kippet Hills Sand and Gravel Formation, and the Hatton Till, indicates that the deposits have been transported and reworked by Late Devensian ice and meltwater that carried the material north-westwards (onshore) from the North Sea basin. The orientation of the esker ridge and its connexion northwards into an ice-contact outwash fan at Knapsleask, indicate a hydrological gradient within the ice from south to north. This gradient resulted in northward flow of meltwaters across the Kippet Hills area during the decay of ice that impinged on this part of the Buchan coastal fringe during the Late Devensian.

Site 17 Errollston Clay Pit, Cruden Bay

The Pleistocene succession in the Cruden Bay–Port Erroll area, some 15 km east-north-east of Ellon on Sheet 87E, consists mainly of stiff, reddish brown pebbly clayey diamicton with subordinate beds of sand and laminated mud. Interest has revolved around the significance of shell fragments and organic horizons preserved within the sequence, especially that exposed in the former brick and tile works at Errollston [NK 086 372] ((Figure 50), (Map 7)). The site must have been relatively poorly exposed during the late 19th and early 20th centuries as few substantial records were made of the deposits. This is in contrast to the Ellon–Slains area, immediately to the south-west, which figured prominently in the evolving knowledge of the complex Pleistocene stratigraphy of Buchan (see Site 15 Ellon).

Pleistocene succession

Jamieson (1858) recorded poorly exposed sequences of pale grey and reddish coloured sands and red clays intermittently visible along the Water of Cruden westwards to Ardiffery. He noted the presence of striated erratic clasts in the clays and fragments of shell. He considered that there might have been about 46 m of Pleistocene sediments present. In his 1906 paper he noted a railway cutting exposure at Port Erroll from which he recorded a bed of uniform, unstratified red clay, 9.1 to 12.2 m thick, with small clasts dispersed throughout (possibly a deformation till). Jamieson considered that the sandy and gravelly sediments were deposited in close proximity to a retreating ice margin, whereas the more clay-rich deposits were a product of more distal sedimentation. He concluded that the deposits were laid down by a glacier advancing north-eastwards along the coast, having derived the red colour from Old Red Sandstone strata in Strathmore.

Bremner (1916) reported the presence of red brick clays at Errollston up to an elevation of about 30 m OD, with red boulder clay occurring up to about 61 m. He considered the brick clays to have been deposited in a glacial lake impounded by an ice margin at Port Erroll and partly fed through a meltwater channel now occupied by the Water of Cruden. In 1943, he recorded ‘lumps and sheets’ of dark shelly boulder clay occurring as transported masses within his Strathmore (red) boulder clay in the brickworks and at sections near the mouth of the Water of Cruden. He considered these masses to be the remnants of a possible Scandinavian-derived boulder clay originally laid down in the North Sea basin.

Eyles et al. (1946) recorded the following section in the Cruden Bay Brick and Tile Company’s works in 1944.

4.6 to 6.1 m of stiff, reddish brown, unbedded boulder clay with a few stones and boulders overlying up to 1.8 m of bedded sand, sandy clay and silt (occasionally in pockets)

They also noted borings at the site proving between 12.2 and 24.4 m of clay.

FitzPatrick (1975a) described a section [NK 088 370], which revealed approximately 3.0 m of red clayey material with an absence of bedding and numerous erratics that he interpreted as a till rather than a lacustrine sediment. This unit overlay a 1 m thick unit of laminated material (containing a 1 cm-thick lamina of black organic mud) overlying sand. FitzPatrick interpreted these sediments as representing an initial ice advance from the east, followed by a minor still-stand or retreat during which proglacial sands and organic muds were deposited in a small pond. Subsequently, ice advanced again and deposited the overlying till which locally incorporated masses of the organic sediment. On final retreat of the ice, permafrost caused a subcuboidal structure to develop in the till. A sample of the black organic clay yielded a radiocarbon date of 36 720 + 1030/-910 BP (SRR–596) (Harkness and Wilson, 1981). This date was significant as it appeared to date the latest advance of ‘Strathmore’ ice into eastern Buchan.

FitzPatrick’s (1975a) interpretation of the genesis of the deposit overlying the organic mud was disputed by Clapperton and Sugden (1977), who suggested that the sequence of interbedded laminated clays, silts and sands with beds of flow till all exhibited slump and settlement structures associated with glaciolacustrine sedimentation. They questioned the validity of the radiocarbon date and whether the organic sediment had been reworked.

Peacock (1984) recorded another section in a pit [NK 086 372] during 1974.

The graded silt/clay laminae in unit 2 were interpreted as turbidites and the overlying diamictons as mudflows. The origin of unit 3 was more problematic, perhaps being deposited as a thick debris flow or as a subaqueous till.

A borehole drilled close to the works at [NK 0895 3719] (Merritt, 1981; NK03NE/13) confirmed earlier records that the sediments described in the brick pit were only the upper portion of a sequence over 20 m thick dominated by red-brown diamicton with subsidiary laminated silty clays. The borehole log records that the red-brown units are mixed with dark grey gravelly till between 13.6 and 14.6 m depth. It is tempting to suggest that this dark grey material is a glacial raft similar to that noted by Bremner (1943), especially as another mass of dark yellowish brown sandy stony diamicton (2.0 m long and 0.5 m thick) has been observed within the red-brown clayey diamicton. The latter mass is lithologically similar to tills of the East Grampians Drift Group known from a number of sites in the area, mostly below red-brown clayey diamicton or related sediments (see Site 15 Ellon).

Cabrera (1976) recorded a section in a pit at [NK 086 370] as part of a study of soils in the area. The deposits were noted as being over 7.0 m thick with an upper, rather massive unit of sparsely stony diamicton overlying beds of sand below about 4.0 m. Cabrera’s work included detailed particle size analyses, clay and fine sand mineralogy and soil micromorphology of the upper 5.0 m of sediment. This work provided strong support for the Glentworth et al. (1964) study of the provenance of the ‘Red Series’ sediments in Buchan. Cabrera confirmed that the red-brown sediments at Errollston were derived from Devonian and Triassic red sedimentary rocks beneath the western North Sea rather than from the Old Red Sandstone of Strathmore. Additional supporting evidence comes from palynological analysis of the matrix of the red-brown clayey diamicton at the site. information from R Harland (BGS, 1981) reported a mixed assemblage of Permo–Triassic miospores together with Jurassic, ‘Tertiary’ and Quaternary dinoflagellate cysts.

The Quaternary Research Association field excursion in 1984 allowed sampling of black clay laminae in the pit from a sequence equivalent to Peacock’s (1984) unit 2. Palynological analysis of this material (Connell et al., 1985) revealed a rich assemblage of miospores and dinoflagellate cysts of Late Jurassic to Early Cretaceous age, confirming the concern about the dating expressed by Clapperton and Sugden (1977). The work indicated that the black clays were derived largely from Mesozoic material and were therefore contaminated with ‘old’ carbon. Hence the 36 720 BP date on the organic clay has no significance for the Pleistocene chronology of the site. It is of interest that the black clay with concentrated, redeposited palynomorphs, both here and at Mintlaw (see Site 10 Ugie valley), is equivalent to unlithified cannel coal, but here reworked in a glaciolacustrine environment from sediments many millions of years old.

Reappraisal of Peacock’s (1984) description of the section suggests possible alternative explanations for the genesis of the formerly exposed diamicton beds overlying the laminated units in the pit. He noted locally developed folded and slickensided silts and sands in unit 4 together with local shearing and contortion within the diamictons of unit 3. It has also been recorded that the black clay laminae (Peacock’s unit 2) are deformed and undulatory with single layers spitting into two between inclusions of red clay (information from J E Gordon, Scottish Natural Heritage, 1999). While these structures possibly resulted from slump and gravitational settling of glaciolacustrine clays and diamictons, it is more likely that the dislocations and slickensiding results from subglacial tectonism as the sediment was overridden by glacier ice, probably from seaward. Thus, the thick succession of diamictons and clays proved in the area may represent a glacitectonite/deformation till complex.

Further sedimentological and textural work is required to resolve the genesis of the diamictons, but the works is no longer in operation and sections are severely degraded. Re-examination of Cabrera’s (1976) micromorphological thin sections may help resolve the issue.

Conclusions

Stratigraphical work in the area of the Errollston clay pit over the last 25 years has confirmed the presence of a thick sequence of red-brown, calcareous and shelly, diamicton interbedded with laminated clay. It forms part of the Logie-Buchan Drift Group and derives its colour from erosion of sedimentary rocks (Devonian and Triassic) that crop out beneath the western North Sea. These materials were brought onshore by a strong westerly or north-westerly ice movement during the Late Devensian. Associated black laminated clays and masses of black clay containing Mesozoic palynological assemblages are likely to have been derived from sediments resembling the Pitlurg Till Formation (see Site 15 Ellon) which are known to occur beneath Logie-Buchan Drift Group deposits to the south-west and which may have been more extensive formerly. The radiocarbon date from black laminated clay in the pit has no significance in terms of the Pleistocene history of the area.

Uncertainty still surrounds the mode of deposition of the diamicton formerly exposed in the pit. It has been interpreted either as till or as a deformed glaciolacustrine sequence. Reappraisal of Peacock’s (1984) description suggests that either the diamicton has been overridden by a late-stage readvance of ice or that it forms part of a thick glacitectonite/deformation till complex.

Site 18 Mill of Dyce Quarry

The sequence at Mill of Dyce, on Sheet 77 (Map 9), provides an insight into the sedimentary environments that were associated with ponding of glacial lakes in the major valleys during Late Devensian deglaciation. It indicates that retreat of a glacier in the Don valley was punctuated by a stillstand, during which an ice-contact delta accumulated in a proglacial lake downstream from the site. Radiocarbon dating of intercalated organic sediments towards the top of the sequence also controversially suggests that glacier ice remained buried within the delta sediments, until at least 11 550 14C years BP. It also implies that a later lake was impounded behind the delta, until at least the latter stages of the Windermere Interstadial.

Quarrying at the Mill of Dyce (formerly known as the St Fergus) sand and gravel pit has revealed an extensively deformed succession of diamicton, gravel and sand which pass laterally into less deformed sediments. The pit [NJ 871 152] is located in the valley of the River Don, 11 km north-west of Aberdeen. It was excavated into moundy topography principally underlain by sand and gravel. These glacial sediments infill the valley, although the river has cut a narrow channel through their northern margin

(Figure A1.22). Much of the moundy ground has been quarried away and the workings largely backfilled and landscaped, but in 1989 the unquarried portion reached an elevation of about 55 m above OD. The western margin of the moundy ground is an abrupt scarp (ice-contact) slope, whereas to the east the mounds merge into an undulating glaciofluvial terrace, which has been modified by excavations and landscaping associated with the expansion of industrial estates on the periphery of Dyce.

Exposures in the Mill of Dyce pit

Previous descriptions of the sediments at Mill of Dyce (Murdoch, 1975, 1977; Peacock in Harkness and Wilson, 1979; Auton and Crofts, 1986) have been reviewed by Aitken (1995). Murdoch (1977) recognised three sedimentary units at the pit. His basal unit, comprised up to 6 m of medium-grained sand typically showing very large-scale, planar cross-bedding, dipping at 28° towards the south-east. Murdoch’s middle unit, comprised 6 to 8 m of poorly sorted bouldery gravel that was highly contorted (dips of 32° were common, and locally bedding was near vertical). Characteristically, the amount of deformation declined upwards within the exposures. Murdoch’s top unit comprised bedded, fine-grained sand, which locally capped the succession and was best developed in swales between hummocks. He interpreted the basal unit as being of deltaic (possibly subglacial) origin, his middle unit as being a supraglacial deposit, deformed by the melting of underlying buried ice, and the uppermost unit as a later stage of subaerial fluvial deposition.

The coarsening-upward sequence exposed in the Mill of Dyce pit during 1984 was recorded as exceeding 16.2 m in thickness (Auton and Crofts, 1986). The succession included a lower unit of clayey sand, greater than 5.2 m thick, and cross-bedded in its upper 4 m. The sand was overlain by 10 m of poorly bedded gravel with boulders at the top. The gravel was overlain, in places, by up to 0.7 m of bouldery diamicton. Clasts within the gravel and the diamicton comprised a mixture of Dalradian metamorphic and Caledonian igneous rock types.

Exposures in the pit during 1988 and 1989 were logged in detail by Aitken (1995), who recognised five lithofacies associations (LFA). The main characteristics of each association are given in (Table A1.10); their relationships and interpreted palaeoenvironments are shown in (Figure A1.22).

LFA 1

LFA 1 occurred in the floor of the pit and was interpreted by Aitken (1995) as representing mainly deltaic bottomsets, with the cross-bedded upper parts of the association indicating a transition into foresets.

LFA 2

LFA 2 was exposed in the northern parts of the pit. It was interpreted as representing the cores of longitudinal or medial bars formed by proximal braided glaciofluvial drainage. The deformation structures were interpreted as indicating collapse following melting of buried glacial ice.

LFA 3

LFA 3 of Aitken, which probably equates with Murdoch’s middle unit, graded laterally into LFA 2. It was interpreted as an ice-marginal deposit derived, in part, from basal debris carried upwards through the ice along shear planes to reach the snout of a glacier, which occupied the valley of the River Don to the west (upstream) of the Mill of Dyce. The debris was subsequently reworked by cohesionless debris flows and supraglacial and englacial meltwaters, to produce the better sorted sands and imbricated gravels.

LFA 4

LFA 4 cropped out in the western face of quarry, adjacent to the ice-contact slope forming the western margin of the moundy deposits. It graded laterally into LFA 3. The sediments of LFA 4 were also interpreted as ice-contact deposits formed by in situ melt-out of debris-rich ice and resedimented by debris flows. Thus, much of the deformation was attributed to postdepositional loading or ice-push. Lenses of finer grained sediments, seen at the top of the association, were interpreted as kettlehole infills.

LFA 5

LFA 5 was restricted to swales in the upper surface of the moundy deposits and equates with Murdoch’s upper unit. Aitken interpreted these sediments as lacustrine overspill deposits, laid down by high-energy unidirectional waterflows.

In one exposure [NJ 8713 1496] at 40 to 45 m above OD, visible in 1979, but since quarried away, Murdoch’s middle and upper units (Aitken’s LFAs 3 and 5) were separated by an organic layer (J D Peacock, personal. communication, in Aitken, 1995). A unit of fine- to coarse-grained sand greater than 1 m thick, with beds and masses of coarse-grained gravel, locally dipping at 80° (equivalent to Murdoch’s middle unit), was overlain by an organic bed 3 m long and 32 cm thick. It consisted of grey, silty clay, thinly interbedded with organic sediment containing plant remains (including twigs) in beds 2 cm thick, dipping at 34°. The organic sediments were overlain by a 3 cm-thick yellow-brown clay that was, in turn, conformably overlain by 3 m of sand with interbedded cobble gravel. The gravel was disposed in a synclinal ice-collapse structure, truncated by 2 m of subhorizontally bedded coarse-grained sand with a thin gravel layer locally present at its base. The latter is taken to be equivalent to Murdoch’s upper unit, and dipped at 54° locally.

Radiocarbon dating of the upper and basal 5 cm of the organic deposit gave ages of 11 550 ± 80 BP (SRR–762) and 11 640 ± 70 BP (SRR–763), respectively (Harkness and Wilson, 1979; (Table 8)). Peacock (in Harkness and Wilson, 1979) suggested that the organic material was deposited in a fluvial backwater and that disturbance of the bed was caused by the melting of buried ice. This implies that, if the radiocarbon ages are correct, the final melting of buried glacier ice at Mill of Dyce postdated 11 550 BP, a view accepted by Auton and Crofts (1986), Munro (1986) and Aitken (1995).

Borehole and mapping evidence

Several boreholes were sunk by BGS in the vicinity of the Mill of Dyce pit during 1984. Borehole NJ81SE/17 to the east of the pit, sited on top of a dissected mound at 44 m OD, proved laminated, micaceous, pale grey-brown silt and clayey sand with scattered pebbles between 10.2 and 12.2 m depth overlying more than 7.9 m of sandy gravel. The laminated sediment was overlain by a 6.9 m-thick coarsening-upward sequence of sandy gravel that was overlain in turn by 1.3 m of clayey sand. Boreholes sited on the floodplain to the west of the site clearly show that this part of the Don valley occupies a glacially overdeepened basin, infilled by sandy silts, which in Borehole NJ81NE/1 exceed 16.2 m in thickness. These silty deposits (Glen Dye Silts Formation) extend upstream from Mill of Dyce for a distance of at least 7.3 km (Aitken, 1991) and crop out in the frontal bluffs of terraces, on the north side of the valley, up to about 50 m above OD.

In many boreholes and trial pits, sited on the floodplain of the River Don near Kintore, the laminated silts are overlain by sands and gravels containing interbedded organic sediments (Auton and Crofts, 1986; Aitken, 1991). Palaeontological examination and radiocarbon dating indicate, however, that these organic sediments are of late Holocene rather than Windermere Interstadial age (see Site 20 Nether Daugh).

Sedimentological interpretation of the Mill of Dyce sequence

All of the laminated silty deposits in the vicinity of the Mill of Dyce have been interpreted as glaciolacustrine deposits (Auton and Crofts, 1986; Aitken, 1995) and their presence has important implications for interpretation of the lithofacies that were seen in the pit. This interpretation agrees in most aspects with the model presented by Aitken (1995), which differs from that of Murdoch (1977), who suggested that much of the sequence in the Mill of Dyce pit was deposited as a subglacial delta. Subsequent investigations (Auton and Crofts, 1986; Aitken, 1995) indicate that much of the pit sequence represents an ice-marginal delta which prograded into a lake ponded by ice downstream (Figure A1.21). The lake level, which at its maximum probably stood at about 50 m above OD, varied as a result of periodic outflow, either subglacially or englacially, through the ice dam.

Deposition at Mill of Dyce commenced by the laying down of coarse ice-proximal gravels by high-discharge, sediment-laden streams and debris flows near the mouths of meltwater conduits at an ice margin. It occured during a stillstand in the active retreat of a glacier in the Don valley. The debris flow deposits, which have been extensively deformed, pass down dip into less deformed gravels that were deposited by sediment-laden meltwater streams. Deformation of the proximal gravels resulted from a combination of collapse, owing to the melting both of buried and buttressing ice, and ice push resulting from minor fluctuations of the ice margin. Both gravel deposits directly overlie sandy foresets in parts of the pit. The foresets represent progradation of the delta into the lake during periods of stable lake level. In places, foresets pass gradationally into bottomsets that were deposited by a combination of small-scale turbidity flows and sedimentation from suspension. The absence of dropstones in the bottomset beds may indicate that the ice margin had little direct contact with the ponded water body while the bottomsets were being deposited.

In the northern parts of the pit, the bottomset beds were overlain, with an erosive contact, by proximal braided outwash gravels. The latter were laid down during periods of delta incision and consequent erosion of any pre-existing foresets. Incision occurred as a result of the periodic falls in the level of the proglacial lake. Localised deformation of these gravels resulted from collapse owing to meltout of buried blocks of ice.

If Aitken’s interpretation of the sediments of LFA 1 at Mill of Dyce as lacustrine overspill deposits is correct, the youngest deposits were laid down after glacier ice had retreated from the Don valley. This retreat left a lake, upstream of Mill of Dyce, ponded by the debris flow and deltaic deposits, and the ice buried within them. In this model, lake overspill eventually incised the upper part of the deltaic sequence and deposited relatively thin, coarse-grained sands on top of the Late-glacial organic sediments at low points on the abandoned delta surface. If the radiocarbon dates on the organic material and the palaoenvironmental interpretation of LFA 1 are accepted, they imply that a deep lake survived, upstream of Mill of Dyce, through at least most of the relatively warm Windermere Interstadial; final drainage occurred some time after 11 550 BP.

Pattern of deglaciation in the adjacent area

The sequence recorded from the Mill of Dyce forms part of a complex suite of Late Devensian glaciofluvial deposits, laid down in the Don valley by meltwaters derived from the East Grampian ice sheet as it retreated westwards. Much of the sediment that was exposed at Mill of Dyce accumulated at a relatively late stage in this deglaciation. At an earlier stage, East Grampian ice and its associated meltwater drainage flowed directly eastwards, across the northern interfluve of the present river, towards the coastal ice stream responsible for laying down deposits of the Logie-Buchan Drift Group. The glaciofluvial sands and gravels deposited by the inland meltwaters, form flat-topped mounds and hummocky ridges, orientated east to west, extending between Mill of Dyce and Corby Loch [NJ 924 145] (Auton and Crofts, 1986). The sediments are well exposed in Lochhills pit [NJ 912 147], where the sequence is similar in many respects to that formerly exposed at Mill of Dyce.

Foreset bedding at Lochhills pit, and in several former workings around Corby Loch, indicates that many of the sequences were laid down as ice-contact fans and deltas (Aitken, 1998). The tops of the deltas indicate that water levels reached heights of up to 91 m above OD, probably ponded between coastal ice and retreating East Grampian ice. Further evidence of the confluence of the ice masses comes from the complex assemblage of northerly and easterly trending moundy outwash deposits occurring between Corby Loch and Potterton (Murdoch, 1975; Munro, 1986). The orientations of these features indicate the contrasting patterns of meltwater drainage that were associated with retreat of the East Grampian and coastal ice sheets.

It is likely that the ice emanating from the Dee valley caused the initial ponding in the Don valley immediately downstream of Mill of Dyce, diverting meltwaters towards Corby Loch. It is also likely that any sediments blocking the valley hereabouts would not have been dissected by drainage until the regional water table had lowered following the decay of the coastal ice stream and/or of sea level, possibly from a height of about 30 m OD.

Site 19 Strabathie

The Late Devensian glaciolacustrine sediments that were formerly exposed at Strabathie sand and gravel quarry, north of Aberdeen, provided rare evidence of the type of sedimentation that took place during the decay of the coastal Logie-Buchan ice stream and retreat of the East Grampian ice sheet. In particular, it indicates that a glacial lake formed, to a height of at least 30 m above OD, ponded between the ice masses as they retreated.

The coastal area north of Aberdeen comprises a belt of ridged hummocky terrain, underlain by a complicated drift succession derived from both the East Grampian ice sheet and Logie-Buchan ice stream (Auton and Crofts, 1986; Aitken, 1991) ((Figure 50), (Map 9)). Simpson (1955) noted a marked lateral and vertical lithological variations in the sequence and interpreted the hummocky features as kames. Synge (1956) interpreted the succession at Strabathie as a series of overlapping fan deltas and noted that deposits forming the ridges locally overlay red laminated clay. The detailed sedimentology at Strabathie was described by Thomas (1984) and Thomas and Connell (1985), who identified four main facies (Table A1.11). Aitken (1991) interpreted the morphology of the ridges, and sedimentary structures within the deposits, as indicating deposition of coarse-grained sediments as a series of subglacial esker, ice-marginal kame and fan delta deposits.

The abandoned sand and gravel pit at Strabathie [NJ 958 135] is located 1 km from the coast and 8 km to the north of Aberdeen. The eastern end of the pit, now a refuse tip, was excavated within a broad easterly trending ridge, standing up to 15 m above the surrounding terrain. The central and western parts of the pit were excavated into an area of low ridges and scattered hummocks. The pit is flanked on its northern margin by a deep meltwater channel; its southern margin is a steep south-facing slope.

Thomas (1984) and Aitken (1991, 1993) both noted that each of the lacustrine facies contained numerous dropstones. They also recorded the vivid red-coloured silt and clay lamination within the fine-grained sediments, and that the diamicton facies was also laminated.

Thomas (1984) proposed that the esker delta facies sediments were deposited by a subglacial meltwater stream that flowed from a tunnel at the base of an overhanging ice margin (Figure A1.23). It debouched directly into an ice marginal lake to the east of an ice front. Coarse-grained sediment was deposited at the tunnel exit and, as the glacier retreated, the sediments built up a ridge of off-lapping delta-fans. Pulsating turbidity currents deposited finer grained sediment away from the ridge, while coarse-grained material was deposited into the more distal parts of the lake from floating ice. The overhanging ice margin eventually collapsed, when the lake water drained away, and the decaying ice released diamicton to cap the proximal lacustrine facies. Aitken (1991) presented a similar interpretation of the sedimentary sequence, but found no evidence for an overhanging ice margin, which he considered would have been inherently unstable. He interpreted the diamicton facies as subaqueous, cohesive debris flows from the ice margin.

The lacustrine facies (2 and 3) at Strabathie contained two types of enigmatic structures (Figure A1.24).

The setting of the glaciolacustrine deposits at Strabathie and their sedimentary structures, suggest deposition in an ice-dammed lake that was ponded between the previously confluent East Grampian ice sheet and Logie-Buchan ice stream offshore The presence of the latter is indicated by the red laminae and red-brown diamictons. The lake at Strabathie was one of a series that formed as the East Grampian ice ‘un-zipped’ southwards from the coastal ice (Figure 42). The water level in the Strabathie lake would have stood at least as high as 30 m above OD, the height of the mounds at Strabathie pit. If the lake was connected to the sea, as seems probable, and assuming that falling sea levels accompanied deglaciation, the lake must have formed at an early stage in the deglaciation. By analogy with sea level data from the St Fergus site, about 40 km to the north, the ponding at Strabathie probably occurred before 15 300 BP. This is the approximate age of the St Fergus Silt Formation, near Peterhead (Hall and Jarvis, 1989), which were deposited when sea level stood at about 12 m above OD (see Site 11 St Fergus).

Site 20 Nether Daugh, Kintore

Fine-grained sediments, containing the remains of plants and insects, recovered from pits excavated in the floodplain of the River Don, provide an insight into the type of environmental change that affected river catchments in north-east Scotland during the late Holocene. In particular, they suggest that aggradation of fine-grained alluvial sediments, which infill abandoned river channels, may have resulted, at least in part, from soil erosion initiated by deforestation and the development of early agriculture.

The Nether Daugh site [NJ 800 160] lies at the foot of a terrace, adjacent to an abandoned meander cut-off, in the Don valley east of Kintore, on Sheet 76E ((Figure A1.25), (Map 8)). This part of the Don valley is infilled by at least 16.2 m of sandy silt, with up to 5 m of laminated micaceous brown and grey sandy silt, passing up into olive-green and olive-brown unlaminated silt. These silts underlie up to 4 m of sand and gravel, which is capped by up to 2 m of fine-grained alluvium and soil (Auton and Crofts, 1986; Aitken, 1991). The silts are interpreted as sediments that were laid down in a lake, ponded at the Mill of Dyce (see Site 18 Mill of Dyce) and which drained away after 11 550 14C years BP (Auton and Crofts, 1986; Aitken, 1991, 1995). The uppermost parts of the silt sequence may contain organic matter, but its presence has not been confirmed.

BGS Borehole NJ 81 NW3 (Auton and Crofts, 1986) drilled at Nether Daugh proved 0.2 m of ‘peat’ approximately 2 m below the base of the fine-grained floodplain alluvium. Two further boreholes ((Figure A1.25) boreholes 1 and 2) [[NJ 8003 1600] and [NJ 8002 1599], respectively], were drilled at the site in December 1986 (information from J Maizels and R Gunson, University of Aberdeen, 1986). Borehole 1 confirmed the presence of an organic interval, comprising 1.1 m of grey, organic mud, overlain by 0.5 m of grey sand with branches and twigs of Betula, and overlain in turn by 1.6 m of grey, organic sand. Subsequently, two pits were dug by mechanical excavator, in February 1988, in order to obtain samples of the organic material for palaeoecological analyses.

The stratigraphy of the two pits and the three boreholes from the site are similar (Aitken, 1991) and is summarised in (Figure A1.25). Both pits contained an organic unit at least 10 to 20 cm thick, overlying fine-grained silty sand and overlain by 4.0 to 4.5 m of clay and grey pebbly sands. The organic remains in Pit 1 [NJ 8002 1597] comprised twigs, branches, bark, leaves and seeds in a matrix of grey, silty clay. The upper part of the organic unit in Pit 2 [NJ 8004 1603] contained a higher proportion of organic matter. It comprised woody peat, with thin sand laminations that contained plant macrofossils and insect fragments, again within a grey, silty clay matrix. These organic sediments are interpreted as part of a late Holocene channel fill succession (Aitken, 1991).

Although only bulk samples of the organic remains were obtained, pollen analysis, and 14C dating of the organic unit was undertaken. Its pollen content was similar in both pits (Table A1.12) and indicated sparse vegetation. Non-arboreal pollen is dominant, particularly Gramineae and Cyperaceae, and to a lesser extent Ericales. However, Betula, Alnus and Corylus/Myrica were present in significant proportions. The environment was, therefore, dominated by grass with some heathland, perhaps on the valley sides. Cyperaceae, and possibly Myrica, probably grew in the silted, abandoned channel, while some stands of birch, alder and possibly hazel woodland were present locally (Aitken, 1991).

A single sample from Pit 2 was separated into plant macrofossil and organic fractions for radiocarbon dating. The respective fractions yielded dates of 3855 ± 50 14C years BP (SRR–3718 i) and 4120 ± 50 14C years BP (SRR–3718 ii) (Table 8). The slight discrepancy between the two ages is attributed to the presence of reworked older organic residues in the silt.

Palaeoentomology was carried out on a single 3.5 kg sample from Pit 2. The full list of Coleoptera is given in (Table A1.13). The limited beetle fauna describes a landscape consistent with that suggested by the pollen evidence. Both suggest that the organic unit developed in an abandoned former channel of the river. The plant debris created a wet, boggy surface with open pools of stagnant or only slowly moving water bordered by reed swamp and a ground layer of hygrophilous herbs and scrubby brush or woodland. There is some indication that this was located within a moorland landscape (Dinnin in Aitken, 1991). Some of the beetle species shed light on the interpretation of the channel fill and subsequent floodplain sedimentation. The leaf beetle Chrysolina fastuosa feeds on hemp and deadnettle, which are arable and cultivated land species. Furthermore, Selatosomus incanus (click beetle) and Trechus quadristriatus (ground beetle) are species of open ground, and are commonly found on arable land (Harde, 1981; Lindroth, 1985). These three species indicate the possibility of arable agriculture (Dinnin in Aitken, 1991).

The presence of only a single Elmidae, Esolus parallelopipedus, one of the commonest of British Elmids, is also significant. Analysis of sub-fossil assemblages from Warwickshire (Osborne, 1988) demonstrates that the distribution of even the rarest species of this genus are relicts of once widespread ranges. These beetles live on clasts in well-oxygenated water with stony or gravel beds. Osborne (1988) suggested that the local extinction of such species was caused by the influx of particulate sediment burying the coarse-grained river bed. He proposed that such valley fill was the result of soil erosion initiated by either deforestation or changes in land use management. The absence of Elmidae from the Nether Daugh site may have resulted from similar processes within the Don catchment.

Channel abandonment and aggradation of the floodplain probably represents the only significant geomorphological change of the River Don and its valley since early Holocene valley floor stabilisation. This alluvial aggradation probably occurred during the mid- to late Holocene, in a largely cleared landscape. There is limited evidence of arable activity, mainly indicated by the occurrence of certain Coleoptera.

The age of the organic sediments, at about 4000 14C years BP, corresponds with that of the earliest cereal grain found in north-east Scotland, from Balbridie, near Banchory (information from Aberdeen Art Gallery and Museums Environmental Archaeology Unit, 1989). Furthermore, Edwards (1978) and Edwards and Rowntree (1980) have shown that the late Neolithic and early Bronze Age was a period of human impact on vegetation in north-east Scotland, with large-scale forest clearance resulting in increased sedimentation into lochs on Deeside. There is no evidence for significant climatic changes that may have initiated alluviation (Lamb, 1977). Indeed it appears that Holocene climatic changes were subtle and may have had little effect on erosion except where vegetation cover was removed (Brown, 1987). Hence, the infilling of the channel at Nether Daugh, and the subsequent floodplain aggradation, may be partially a consequence of anthropogenic activity in the Don catchment.

Site 21 Rothens, Monymusk

The site at Rothens [NJ 6878 1710] has yielded a pollen spectrum that represents a ‘tripartite’ Late-glacial sequence, encompassing much of the Late-glacial (Windermere) Interstadial, the Loch Lomond Stadial and the early Holocene. It is unusual in that it appears to contain evidence of a climatic fluctuation within the Late-glacial Interstadial that may correlate with the ‘Older Dryas’ episode of Scandinavia.

The site is situated in hummocky terrain adjacent to the now abandoned Rothens gravel pit, 2 km north-west of Monymusk, on Sheet 76E (Map 8). Exposures in the gravel pit revealed a complex succession interpreted as proximal and distal outwash deposits and associated overbank facies (Aitken, 1991, 1998). The site lies at the western end of a small valley that extends from the Don valley at Monymusk to Kemnay. It has been interpreted as a major route of meltwater drainage which occurred while the main valley still contained ice (Aitken, 1991). The site is a roughly circular kettlehole with a diameter of 300 m and lies at an altitude of about 88 m OD. It is flanked on its eastern and southern margins by an arcuate ridge.

The kettlehole contains at least 5 m of fill. The lower parts of the sequence have been sampled for pollen analysis and the stratigraphy of the sampled is as follows:

Depth

Three distinct pollen assemblage zones occur and are numbered R-1 to R-3 (Figure A1.26). The sedimentary boundaries are slightly offset from the pollen zone boundaries, with changes in pollen stratigraphy preceding changes in sediment type. This may either represent a lag in the response of soil development and degradation, to changing environmental conditions, or indicate that the pollen zone boundaries are gradational. Details of individual zones are as follows:

R-1 Gramineae–Rumex–Cyperaceae Assemblage Zone (4.45–4.93 m)

This basal assemblage zone is dominated by non-arboreal pollen. It is characterised by high values of Gramineae (grasses), Rumex (docks) and Cyperaceae (sedges). Relative values of Ericales rise rapidly in the middle of the zone to attain a peak of 33 per cent TDLP (Aitken, 1991). However, absolute values for Ericales (heathers), remain relatively constant throughout the zone and the relative rise in Ericales may be, at least, in part, a statistical artefact (Figure A1.26). The zone is noteworthy for the high values of Myriophyllum cf. alterniflorum (water-milfoil). Pollen of woody plants is scarce, but Juniperus (juniper), Betula (birch), Salix (willow) and Corylus/Myrica (hazel/bog myrtle) taxa all occur. The upper boundary of the zone is placed at the rise in values for Artemisia (mugworts), the absolute rise in values for Ericales and the falls in Juniperus and Betula pollen. By comparison with other northern Scottish pollen spectra, zone R-1 is thought to be of Late-glacial Interstadial age (Aitken, 1991).

R-2 Ericales–Artemisia Assemblage Zone (4.25–4.45 m)

This zone is also dominated by herb pollen. It is characterised by constantly high values for Ericales (greater than 25 per cent) and increasingly significant values for Artemisia, which attains a maximum of 25 per cent near the top of the zone. Salix maintains its importance in the zone, but values for all other woody taxa are low. Aquatic plant pollen is absent. The upper boundary of the zone is placed at the marked decline in pollen of open habitat taxa, notably Artemisia, Compositae and Caryophyllaceae, and the beginning of the rise in values for woody plants, particularly Betula and Juniperus. By comparison with other northern Scottish pollen spectra, zone R-2 is thought to be of Loch Lomond Stadial age (Aitken, 1991).

R-3 Betula–Juniperus Assemblage Zone (3.69–4.25 m)

This zone is characterised by successive peaks in Gramineae, Ericales, Juniperus and Betula. Percentages of non-arboreal pollen decline throughout the zone, while the aquatic Potamogeton (pondweed) rises to a peak of 41.9 per cent at a depth of 3.70 m. Towards the top of the zone there is a rise in Corylus/Myrica pollen and a corresponding decline in Betula and Juniperus. By comparison with other northern Scottish pollen spectra, zone R-3 is thought to be of early Holocene age (Aitken, 1991).

Four radiocarbon ages have been obtained from the site (Aitken, 1991) (Table 8):

When compared with radiocarbon dated pollen sequences from other Scottish sites, each of the ages from the Rothens deposits appears to be less than expected (Aitken, 1991). The ages of the bottom three samples are at least 200 14C years less than expected, and the date of uppermost sample is perhaps up to 1000 14C years younger than expected. The reasons for these discrepancies are unclear, but contamination by humus-rich ground water may have been an important factor. Furthermore, as a result of the paucity of organic material, the segments selected for radiocarbon assay were relatively thick (up to 5 cm) such that each date represents a mean value for the quoted level. The dates are therefore considered to be unsuitable for reliable detailed chronostratigraphical correlation with other sites. Nevertheless, the radiocarbon ages broadly confirm both the interpretations of the local stratigraphy and the palynology. The organic deposit between 475 and

441 cm formed during the Late-glacial Interstadial, while the overlying grey, pebbly sand was deposited during the succeeding Loch Lomond Stadial. The two lowermost dates, on either side of a less organic silty clay unit, approximately bracket the supposed ‘Older Dryas Stadial’ although this is by no means conclusive.

Site 22 Nigg Bay, Aberdeen

The superficial deposits exposed in the Nigg Bay section (Map 9), to the south of Aberdeen city, represent a classic sequence that has been used extensively to establish the sequence of Pleistocene glaciations in the Aberdeen district, and more widely in north-east Scotland. Jamieson (1882b) provided the first detailed description of the succession. Others, principally Bremner (1928), Simpson (1948, 1955) and Synge (1963) subsequently added detail and debated the interpretation of the sequence. The reader is referred to Gordon (1993c), who reviews the long history of research on the site. This account focuses on describing and interpreting the deposits, and establishing the local lithostratigraphy (Table A1.14).

Although the deposits exposed at Nigg Bay provide evidence of a complex sequence of glacial and deglacial events affecting the Aberdeen area, they do not record all of them. Evidence from the lower Dee valley, and from recent gas pipeline trenches and road excavations to the west and north-west of Aberdeen city are also described here as they provide an important record of earlier events not represented in the section at Nigg Bay.

Nigg Bay Section

The Nigg Bay section, which occurs on the south side of the bay [NJ 965 045], is over 200 m in length and up to 20 m high (Figure A1.27). Since the mid-1970s material dumped at the base of the cliff has obscured most of the lower unit in the stratigraphy and allowed slumping and vegetation growth on the upper part. However, at present all of the main units are visible.

The exposed deposits only represent the uppermost 20 m of a 60 m-thick succession that occupies a broad buried valley lying between Torry and Tullos Hill (Map 9). The valley might have been a former course of the River Dee (Simpson, 1948), but evidence from borehole logs (Monro, 1986) and geophysics (Law, 1962) indicates that, within the broad feature, a narrow gorge-like channel cut into bedrock descends to 40.5 m below OD beneath the bay (Figure 46). This channel has an apparently irregular ‘up-and-down’ long profile suggesting that it was cut by subglacial meltwaters. Its excavation predates the drift succession at the site.

Borehole logs show that the channel is filled with about 30 m of drift, mainly till with subsidiary gravel and sand bodies (Munro, 1986). The till is described as grey or grey-brown in colour, suggesting that it was deposited by ice that advanced from the west or north-west. It is possible that the uppermost till units of the buried sequence correlate with the grey till seen in the cliff section, but they may have been deposited during an earlier event (see below).

The following account is based on the descriptions of Jamieson (1882b), Bremner (1928), Simpson (1948) and Synge (1963), augmented by observations made by E R Connell since the mid 1970s. The formal lithostratigraphy is established here. Four main units can be recognised in the cliff section above bedrock, which comprises granite-gneiss and is locally weathered.

Unit 1 Ness Sand and Gravel Member

This unit, up to 9 m thick, comprises coarse (locally bouldery) gravels and sands passing upwards into sands with minor gravel beds. Clast types are dominated by local metamorphic and igneous rocks, but they also include very sparse Norwegian erratics (rhomb-porphyry and larvikite) and ultrabasic rocks from the Belhelvie mass north of Aberdeen. Flint has been found recently in the gravel, as have a number of clasts of greyish brown diamicton (some of them ‘armoured’). A single unarmoured clast of red diamicton has also been found in the gravel close to the junction with the overlying till.

Interpretation Recent sedimentological and palaeocurrent analyses indicate that the sands and gravels were deposited by meltwater flowing north-eastwards.

They have been assigned to the Lochton Sand and Gravel Formation of the East Grampian Drift Group on account of their dominantly local clast composition (Synge, 1963). However, the presence of ultrabasic clasts derived from the north (Bremner, 1928), sparse Norwegian erratics (Read et al., 1923; Bremner, 1928) and sparse flints makes them sufficiently distinctive to be named here as the Ness Sand and Gravel Member. The clasts of greyish brown diamicton are of interest in that they resemble the overlying grey till in both texture and mineralogy. It seems likely that some clasts were introduced into the meltwater deposit directly by the ice mass that eventually deposited the grey till, whereas others may be derived from older till deposits in the area. The clast of red diamicton possibly originated from an older red, Old Red Sandstone-derived till once present in the area. The ultrabasic clasts, Norwegian erratics and flints were all probably reworked from older deposits when ice advanced into the area from the north-west, north or north-east.

Unit 2 Nigg Till Member

This unit, up to 10 m thick, typically comprises a very dark greyish brown to dark greyish brown (2.5Y 3/2–4/2), hard, sandy, silty diamicton containing predominantly local metamorphic and igneous clasts.

Interpretation This thick deposit of lodgement till rests unconformably on the underlying sands and gravels. Its base dips gently northward and it possibly correlates, in part, with the uppermost till units preserved in the buried channel. The unit is assigned to the Banchory Till Formation of the East Grampian Drift Group on account of its clast composition, which is dominated by local metamorphic and igneous types, and its strong north-easterly clast fabric (compare with Synge, 1963). It was deposited by ice that flowed down the Dee valley from accumulation centres in the Grampian Mountains.

Unit 3 Mill Of Forest Till Formation

Though it is absent locally, this unit ranges up to 5 m thick. It is typically a reddish brown (5YR 4/4), hard, sandy, diamicton with mostly Old Red Sandstone-derived clasts.

Interpretation The gradational transition between this unit and the underlying grey till suggests that both were deposited during the same glacial stage. The content of Old Red Sandstone-derived clasts and north-north-easterly clast fabric (Jamieson, 1882b, 1906; Synge, 1963) of the red diamicton indicate it was deposited by ice flowing northwards from Strathmore along the coast, at a time when the Dee valley ice had receded westward. The absence of Permo–Triassic material and of Early to Mid Pleistocene shells indicates that the unit should be correlated with the Mill of Forest Till Formation of the Mearns Drift Group rather than the Hatton Till Formation of the Logie-Buchan Drift Group, which occurs immediately to the north of Aberdeen. While generally less than 2 m thick in the section, the till is known to be up to 5 m thick in nearby boreholes and excavations.

Unit 4 Drumlithie Sand And Gravel Formation

Coarse gravels (locally bouldery) and sands pass northward into interbedded sandy gravels and sands with laminae and beds of red silt and clay. The gravels contain numerous Old Red Sandstone-derived clasts. The upper part of the unit displays a range of periglacial features, including cryoturbation structures, frost cracked clasts and possible ice-wedge casts. The unit is between 2.5 and 6 m thick.

Interpretation These glaciofluvial outwash deposits contain a similar erratic suite to the underlying red till and were almost certainly deposited from the ice that laid down that till. Simpson (1948, 1955) originally linked this unit with the moundy ice-contact gravels containing a similar erratic assemblage and thin beds of red diamicton at, for example, Balnagask (immediately north of Nigg Bay) and Loirston (some 3 km to the south-south-west). All were deposited during the southward retreat of the same ice mass. However, Jamieson (1865, 1874, 1882b, 1906), Bremner (1928, 1934a, 1938, 1943) and later Synge (1956, 1963) all concluded that the sediments formed part of a moraine that marked the limit of what Synge termed the ‘Aberdeen Readvance’. Synge (1963) sited evidence for a gravelly till and ‘faulting and overthrust structures’ at the northern end of the Nigg section, which he considered were formed at the exact margin of the re-advance.

Recent work has not been able to confirm Synge’s description of till and glacitectonic structures, but suggests that any deformation and faulting is more likely to have resulted from melt-out of buried glacier ice. Simpson (1948) also recorded a range of load, deformation and fault structures in the fine-grained sediments to the west of Nigg. Some may have formed by slumping of sediment from the adjacent valley sides whereas others may have resulted from having been deposited on top of melting buried ice.

Mapping of glaciofluvial sediments and landforms in the Aberdeen district, and farther afield in north-east Scotland, led Clapperton and Sugden (1975, 1977) and Murdoch (1975, 1977) to strongly support Simpson’s (1948, 1955) interpretations. The mapping identified continuity of glaciofluvial landform elements across the boundaries of the supposed ‘Aberdeen Readvance’. As a result, the sediments and landforms of the coastal zone around Aberdeen are now interpreted as having been deposited during the westerly retreat of the East Grampian ice sheet and the southward retreat of the coastal ice that deposited the Mill of Forest Till.

Unit 5 Tullos Clay Member

Recent exposures indicate that the sands and gravels of unit 4 tend to fine northward as originally suggested by Simpson (1948). He also recorded extensive deposits of red and khaki-coloured clays, silts and sands in excavations and boreholes up to 1 km west of the Nigg section. No fossils were recovered from the deposits, but Simpson suggested that they could be glaciomarine in part. Nearby at the former Torry claypit (to the north of Torry and approximately 1.5 km west-north-west of the Nigg section), Jamieson (1858) described a similar sequence of fine-grained sediments and recorded bird and fish remains (Peacock, 1999). These appear to confirm a glaciomarine influence and the deposits are named here as the Tullos Clay Member of the Ugie Clay Formation.

Summary

All the stratigraphical units exposed at Nigg Bay are believed to relate to the Main Late Devensian glaciation of the area (Clapperton and Sugden, 1977; Murdoch, 1977; McLean, 1977) (Table 7). The stratigraphy of the exposed sequence at Nigg Bay indicates a strong flow of ice north-eastwards from the Grampian Mountains along the Dee valley. This advance led to deposition of the outwash gravels and sands of the Ness Sand and Gravel Member and the thick unit of till, the Nigg Till Member. It is unclear if the thick till sequence preserved in the buried channel at Nigg Bay relates in part to this advance of Grampian ice, or an earlier advance of unknown age, or to both. The ‘exotic’ clasts in the Ness Sand and Gravel Member gravels are probably reworked from older deposits in the area. As the East Grampian ice receded, southerly sourced ice transported Old Red Sandstone material from Strathmore (and immediately offshore) into the Aberdeen area. The same ice stream also deposited the Logie-Buchan Drift Group sediments farther to the north. Retreat of this ice mass resulted in the deposition of both moundy ice-contact deposits and sheet-like spreads of the Drumlithie Sand and Gravel Formation throughout the district.

Sections west and north-west of Aberdeen

Important sections in the lower Dee valley are included here as they record events that were earlier than those interpreted from the Nigg Bay section.

Bremner (1934a) has recorded important sections in excavations for the Anderson Drive segment of the Aberdeen ring-road, near Bridge of Dee [NJ 927 037] (Map 9). At the base of the sequence he descibed a ‘dark shelly boulder clay’ containing arctic shells, many of which were striated. Unfortunately he provided no further information on this deposit and more recent investigations have failed to rediscover it (McLean, 1977). Although no Norwegian erratics were found in the deposit by Bremner (1934a), he speculated that it might have been deposited directly by Scandinavian ice, or more probably, by Scottish ice forced back onto the coast by Scandinavian ice in the North Sea (Figure 37). Hart (1941) reported limited textural and mineralogical data from the Anderson Drive deposit that suggest that it is of distinctly different provenance to the Benholm Clay Formation at Gourdon (see Site 26 Burn of Benholm). However, Hart’s data indicates some textural affinity with the Jurassic to Lower Cretaceous derived diamictons within the Banffshire Coastal Drift Group. Consequently, the deposit is here formally named the Anderson Drive Diamicton Formation of that group.

In the ring-road excavations recorded by Bremner (1934a), the dark shelly boulder clay was apparently overlain by ‘grey boulder clay with characteristic erratics from the north-west’ (from the Belhelvie, Barra Hill and Huntly basic igneous masses). Unfortunately, he did not describe the exact relationship between the tills. It is possible that erosion of a unit similar to this one provided the ultrabasic erratics recorded in the lower gravels of the Nigg Bay exposure, and that such a unit could be preserved within the thick sequence infilling the Torry–Tullos buried channel.

During the 1970s, extensive gas pipeline trenches to the west and north-west of Aberdeen provided important information on the local Quaternary sequence. Murdoch (1975, 1977) and McLean (1977) described two distinct and superimposed till units in this area. The lower till was dark brown (10YR 3/4) or locally blue grey (5PB 3/2) where gleyed (Munro, 1986). It was fine grained and possessed a strong south-easterly clast fabric. Erratics derived from outcrops to the north-west were also commonly recorded within this unit, which is patchily distributed and generally preserved only in bedrock depressions.

The lower unit is overlain by a much more laterally extensive unit of grey brown (10YR 5/5–7.5YR 4/5) relatively sandy till (Munro, 1986). Clast fabrics in this upper unit are less strongly clustered than in the underlying till, but display easterly or north-easterly orientations. The till contains abundant clasts of red granite to the east of the Hill of Fare granite. In 1998, an excellent exposure in a road excavation [NJ 879 062] near Kingswells, to the west of Aberdeen, revealed both tills in superposition. The lower of these two tills is here named the Den Burn Till Member and the upper the Kingswells Till Member, both of the Banchory Till Formation. It is probable that the Den Burn Till Member correlates with the ‘grey boulder clay’ reported by Bremner (1934a) overlying the Anderson Drive Diamicton in the ring-road excavations. The Kingswells Till Member is probably the lateral equivalent of the Nigg Till Member in the Nigg section. Both tills seen in the Kingswells exposures would now be described as lodgement/deforming-bed tills.

Murdoch (1977), McLean (1977), Munro (1986) and Auton and Crofts (1986) found evidence that the upper grey-brown till was interbedded with the coastal red till units (Mill of Forest and Hatton till formations) at a number of sites suggesting that deposition of both units occurred close together in time. Both Murdoch (1977) and McLean (1977) were of the opinion that the upper till unit was a supraglacial facies of the lower till. However, based on the evidence of distinct clast assemblages and fabrics, Sutherland (1984a) and Auton and Crofts (1986) concluded that both were basal tills and related to different glacial events.

Summary

The exposures in the lower Dee Valley and to the west and north-west of Aberdeen record important evidence of the pattern and relative timing of glacial advances. Evidence is present for ice advance from the north (Anderson Drive Diamicton Formation) and north-west (Den Burn Till Member). These sediments probably predate the deposits exposed in the Nigg Bay section and may, in part, provide sources for the ‘exotic’ clasts recovered from the Ness Till Member gravels. Later a strong ice flow from the west and south-west deposited the extensive Nigg and Kingswell till members. These tills grade up into, or are interbedded with, Old Red Sandstone-derived tills in the coastal zone and indicate some contemporaneous deposition. As no weathering horizons, soils or peat beds have been recorded interstratifed with the tills and gravels of the Aberdeen area, all units are considered to date from phases of the main Late Devensian glaciation (Clapperton and Sugden, 1977; Murdoch, 1977; McLean, 1977; Munro, 1986). However, some till units present in the buried channel at Nigg may date from earlier events. Radiocarbon dates from peat and lacustrine sequences younger than the glacigenic sediments in the area (see Site 21 Rothens, Site 18 Mill of Dyce and Site 23 Loch of Park) indicate that the region was finally deglaciated prior to 11 640 or 11 900 BP.

Site 23 Loch of Park

The organic sediments preserved within the large ice-scoured rock basin at Loch of Park, north-east of Banchory (Map 10), provide evidence of typical climate and vegetation changes that have occurred within the district since deglaciation. Pollen data and radiocarbon dates from the site indicate that a ‘tripartite’ Late-glacial sequence is preserved that passes upwards into Holocene deposits.

The Loch of Park ((Figure A1.28)a) is the site of a freshwater lake that was drained during the 19th century and is now a local nature reserve. It occurs at the south-eastern end of the basin, which extends for about 4 km along the northern margin of Sheet 66E Banchory and across the southern margin of Sheet 76E Inverurie. The basin was formed by ice that flowed eastwards from the Grampian Highlands, scouring the underlying decomposed Crathes Granodiorite. The ice deposited a thin spread of till on the basin floor. It also laid down morainic mounds at the eastern margin of the basin during its westward retreat. These moraines blocked the subsequent drainage of the lake eastwards and an interbedded sequence of organic muds, peats and fine-grained lacustrine sediments were laid down on top of the basal till. Mounds and narrow terraced spreads of sand and gravel were also laid down around the southern margin of the lake.

The organic sediments in the basin were first described by Vasari and Vasari (1968), who analysed the pollen content from a narrow diameter piston-cored borehole. The borehole, at about 70 m above OD, near the eastern end of the former lake ((Figure A1.28)a), penetrated interbedded peats, clays and organic muds to a depth of 5.6 m. The lowermost 1.70 m of the core was recorded as bluish clay, almost entirely devoid of pollen. Above this unit occurred 0.5 m of organic mud, 0.35 m of nutrient-rich silty peat (gyttja) and 0.53 m of clay. The upper clay was overlain by peat, which became increasingly sandy upwards. The three units above the basal blue clay (and their pollen assemblages) were interpreted as being of Late-glacial age and the upper peat as being Holocene.

The pollen spectrum ((Figure A1.28)b), between 2.10 and 3.85 m depth was assigned to the classical Late-glacial pollen zones I (cold), II (warm), III (cold), a combined zone III–IV, coinciding with the transition from cold to warm conditions, and the Postglacial (Holocene) IV (warm). The Late-glacial sediments, which were subdivided in to three subzones, Ia–Ic, are characterised by a dominance of non-arboreal pollen and the Holocene sediments by a marked increase in the proportion of Betula (birch), Pinus (pine) and Juniperus (juniper) pollen. A corresponding decrease was recorded in the proportions of grass and herb pollen in the Holocene deposits. Subzone Ia (mainly almost barren blue clay) and subzone Ib, (the organic mud) were thought to reflect progressive climatic amelioration. This was followed by a colder episode (subzone Ic).

The sequence was subsequently reinvestigated in 1972 (Vasari, 1977), when samples of organic sediment were taken for radiocarbon dating from a second boring. The borehole was sited immediately adjacent to the first and a similar stratigraphical sequence was proved. The lower sample, HEL–417, taken at the boundary between the top of subzone Ic and the base of zone II (Empetrum {crowberry} rise/Rumex {docks} fall), yielded a radiocarbon age of 11 900 ± 260 BP (Table 8). This was said by Vasari (1977) to correspond closely with the accepted radiocarbon age for the I/II boundary for much of north-western Europe. The upper sample, HEL–416 (at the combined zone III–IV boundary; between the Rumex and Empetrum maxima), yielded a radiocarbon age of 10 280 ± 220 BP. This was seen by Varasri (1977) as compatible with the generally accepted radiocarbon age of the Late-glacial/Holocene boundary in Britain (about 10 250 BP).

Site 24 Balnakettle

An unusual succession of Quaternary sediments occurs in a section at Balnakettle (Plate 26), on Sheet 66E, and in a nearby small gravel pit at Winney Hillocks (Map 10). It provides a unique insight into the sequence of events that took place during the glaciation and deglaciation of the upland flanking the western side of Strathmore. The stratigraphical and structural relationships between deformed and undeformed glacigenic sediments indicate that a late, local advance of East Grampian ice occurred after retreat of the Strathmore ice stream. The latter deposited till and outwash on the flanks of the upland.

Stratigraphy and structure

A complex, deformed sequence of glacigenic deposits of the Mearns Drift Group is exposed in the back scar of a landslip [NO 618 757] on the southern side of a western headwater tributary of the Burn of Caldcotts (the Burn of Balnakettle), north-east of Fettercairn on Sheet 66E Banchory. The deformed glacial sediments are moderate brown to red-brown in colour. They are interstratified with clast-supported gravel, composed predominantly of angular clasts of dark grey phyllite and greenish grey psammite, derived from underlying Dalradian bedrock. The interstratified sequence is overlain by a thin, undeformed, moderate to dark yellowish brown sandy till, containing a preponderance of Dalradian schistose psammite and semi-pelite clasts.

The sequence at Balnakettle (Figure A.1.29a) is exposed to a depth of about 8 m, over a distance of about 28 m, in a valleyside cliff at about 225 m above OD. The section, which is aligned east-north-east, forms part of the undisturbed back scar of a landslip in which material has slipped northwards towards the valley floor. The lower part of the succession consists of a tectonised sequence of tills, silts, clays, sands and gravels, in which bedding has been reorientated by folding and shearing so that many originally flat-lying contacts are now steeply dipping (some are almost vertical and a few are slightly overturned) ((Figure A1.29)b). The upper part consists of an undeformed, till unit that truncates the tectonised sediments.

Although original stratigraphical relationships between the units in the lower sequence are highly deformed, it is apparent that the grey, matrix-rich, phyllite-psammite gravel is the oldest deposit exposed in the section. The gravel is extremely angular and ranges in grain-size from boulder to pebble. The clasts exhibit a strongly developed, tightly clustered bedding-parallel fabric, which has an upright orientation in the lowest part of the sequence ((Figure A1.29)a, 6–10 m along the section). Fining-upward units that occur within the gravel higher in the section, indicate younging towards the west-south-west and small irregular lenses of the gravel are incorporated within a raft of red-brown till caught up in the overlying brown sandy till.

The till within the tectonised sequence is of two types. The predominant sediment is a moderate brown, stiff, silty and clayey matrix-supported diamicton. It contains subangular pebbles and cobbles of Dalradian metamorphic rocks (principally psammite, semipelite, and phyllite) granitic rocks and some Old Red Sandstone clasts (mainly quartzite and sandstone). The clasts display a tightly clustered fabric, and stoss and lee pebbles (compare with Benn and Evans, 1998) are present, some of which bear striations; all of these attributes suggest subglacial deposition.

The clast fabric and composition of the till at the east-north-eastern end of the section would appear to suggest deposition from ice moving north-eastwards (from Strathmore), but nearer to the middle of the section, where bedding has clearly been rotated into an upright position, the fabric is less tightly clustered. It has a south-directed vector mean, similar to that from the matrix-rich gravel, described above.

Reddish brown sandy silty diamicton, containing small pebbles of Dalradian and Old Red Sandstone rock types forms the second type of till in the tectonised sequence. It occurs as deformed lenses, caught up along upright shear planes bounding the displaced beds of silty angular gravel. It also forms a raft, containing lenses of angular phyllite/psammite gravel, within the upper yellowish brown till and pods of diamicton within the red-brown sand and gravel towards the west-south-western end of the section ((Figure A1.29)a, 20–26 m along the section).

Beds and lenses of red-brown laminated silt and clay occur within the tectonised unit of moderate brown till and an overlying unit of red-brown glaciofluvial sand and gravel. At the base of the sequence, and towards the west-south-western end of the section, the glaciofluvial sand and gravel fines upwards, from a gravel lag (containing angular to rounded boulders up to 0.5 m in diameter), into clast-supported coarse sandy gravel and finally into pebbly sand. The pebbly sands and gravels all contain a high proportion of Old Red Sandstone clasts (including purple and maroon sandstone and andesite) many of which are highly decomposed. In the middle of the section ((Figure A1.29)a, 14–22 m), the upward-fining gravel unit has been reorientated into an upright position, by folding and movement along steeply inclined curved shear planes, which are lined by red-brown silt and silty clay. The shear planes are inclined towards the north-west and east and total displacements, possibly of several metres, must have taken place along them during deformation.

Beds of red-brown silty sand are interstratified with gravel and pebbly sand higher in the glaciofluvial sequence. Towards the top of the section, the bedding is contorted into small-scale, tight flat-lying folds and is also displaced along low-angle thrust planes. The thrusts dip between 27° and 30° towards the south-east and north-west; the maximum measured displacement was about 0.3 m.

The upper till, with its slope-parallel base, cuts across and incorporates rafts of the underlying sediments (including pods of the underlying red-brown gravel). It has a sandy matrix and is moderate to dark yellowish brown in colour. The matrix exhibits pronounced subhorizontal fissile layering, suggesting deposition beneath the sole of a glacier (Benn and Evans, 1998). The clast fabric of the unit is weakly clustered compared with that of the underlying moderate brown till. The clasts within the sandy till are predominantly of Dalradian metamorphic and Caledonian igneous rocks, though a few rounded quartzite pebbles (probably initially derived from Old Red Sandstone conglomerate) are also present. The composition of the till, its clast fabric, and the included rafts of red-brown gravel all indicate deposition by East Grampian ice.

Moderate brown interbedded sands, silts and clays, similar to those at the west-south-western end of the main Balnakettle section, rest with a sharp, undulating contact on gravel, composed mainly of angular clasts of grey phyllite and psammite, in exposures to the north-east at [NO 619 758]. This gravel resembles the grey matrix-rich gravel at the base of the tectonised sequence. The exposures occur at about 200 m above OD, towards the base of large slipped blocks at the toe of the landslip. The exact relationship of the interbedded deposits and the underlying gravel, with the sequence exposed in the landslip scar some 25 m above is unclear.

Interpretation

The sorting, grain size and angularity of the grey, matrix-rich, gravel and its restricted composition (which clearly reflects the Dalradian bedrock that crops out in the flanks of the Burn of Balnakettle adjacent to the section), all suggest accumulation as a ‘head’ deposit formed by frost shattering and mass movement of debris in periglacial conditions. The fact that the gravel occurs both as discrete beds at the base of the tectonised sequence, and within rafts in the upper till (rather than being incorporated and homogenised), indicate that it probably accumulated prior to the deposition of all of the glacigenic sediments. Hence, it is the oldest deposit exposed in the landslip scar section and is probably of pre-Late Devensian age.

The oldest glacigenic sediments are the moderate-brown and red-brown diamictons that contain significant numbers of clasts derived from the Old Red Sandstone bedrock of Strathmore. The colour of most of the diamictons is not typical of tills of the Mearns Drift Group (reddish brown), but their compostion indicates derivation from Strathmore ice that had passed over the Dalradian outcrop; their tightly clustered clast fabrics are typical of subglacially deposited tills. The glacitectonic disturbance of the sequence, however, precludes precise interpretation of the direction of former flow of the Strathmore ice from the orientation of the till fabrics alone.

The preservation of head deposits within the moderate brown till suggest that little mixing of pre-existing sediments occurred during the deposition of the Mearns Drift tills. The head gravel was probably frozen onto the base of the Strathmore ice, close to the front of the ice as it impinged onto the upland. The absence of any fragments of an early till of clear East Grampian provenance within the Mearns Drift deposits (while the masses of head gravel are preserved), suggests that it was not present at the site prior to the advance of Strathmore ice.

The north-eastward movement of Strathmore ice across the Balnakettle site initially resulted in stacking of slices of the moderate brown till and head gravel by thrusting. This was followed by the deposition of the overlying glaciofluvial sand and gravel exposed at the west-south-western end of the section. The colour and composition of the sand and gravel are typical of outwash associated with the Strathmore ice. These glaciofluvial deposits were probably laid down as an ice-marginal fan or fan-delta directly on the moderate-brown and red-brown tills, and the intercalated head deposits, before East Grampian ice reached the site. Continued north-eastward movement of the Strathmore ice caused initial shearing within the sand and gravel, followed by tilting and localised folding of both the sand and gravel and the earlier developed thrust stack. This rotated both sequences into their upright positions, and was followed by the development of the small-scale thrusts and flat-lying folds towards the top of the sand and gravel unit.

Once the Strathmore ice had retreated on to the lower lying ground to the south, the East Grampian ice advanced into the area from the north-west and overrode the Balnakettle site. It cut across the deformed sequence and laid down the upper yellowish brown sandy till and incorporated the rafts of moderate-brown till and red-brown gravel. Mapping shows that the East Grampian ice advanced south-eastwards for a distance of at least 1.3 km beyond the Balnakettle site. It subsequently retreated north-westwards to a still-stand position about 500 m south-east of the site, where moundy deposits of subangular boulder, cobble and pebble gravel were laid down at the ice margin. These ice-contact gravels are interbedded with thin sandy laminated diamictons, which probably formed as debris flows. The gravels and diamictons contain the wide variety of igneous and metamorphic clasts, typical of deposits of the East Grampian Drift Group. They are well exposed in a small gravel pit at Winney Hillocks [NO 6204 7524]. Ice-wedge casts up to 1.8 m in length, infilled with fine-grained, subangular sandy gravel, are present in the middle part of the gravel sequence exposed in the pit. They occur beneath matrix-rich gravel, up to 2.1 m in thickness. This upper gravel shows evidence of cryoturbation, in the form of involutions and the local development of erect pebble fabrics. These features are attributed to ground ice development within the gravel under periglacial conditions that either occurred towards the end of the Dimlington Stadial, or, less probably, during the Loch Lomond Stadial.

Site 25 Knockhill Wood, Glenbervie

A Late-glacial (Windermere) Interstadial peat, intercalated between red-brown diamictons at Knockhill Wood, Glenbervie (Map 10), provides important evidence of the age of slope modification in Strathmore by landslipping. The 14C age, pollen spectra and plant macrofossils obtained from the peat, and its stratigraphical contacts suggest that the last movement occurred either during the Loch Lomond Stadial or when the climate ameliorated, during the earliest Holocene.

A composite organic deposit was discovered beneath reddish brown diamicton of the Mearns Drift Group during the resurvey of Sheet 66E Banchory in 1989. The deposits were located in a drainage ditch [NO 7667 8012] on the north-eastern side of a forestry track through Knockhill Wood, on the southern side of the valley of the Bervie Water, upstream of Glenbervie (Figure A1.30). The organic sediments consisted largely of peat, with subordinate layers of fine peaty sand and laminated clayey silt (Table A1.15). They rested upon an unstratified bouldery till, also reddish brown in colour.

The organic deposits are exposed along about 7 m of the ditch, and appear to form a wedge in the side of Knock Hill (Figure A1.31). They overlie bouldery till (unit J of (Table A1.15)) and are interstratified with clayey silts and sands. At the north-eastern end of the section, the upper part of the organic sequence (units B–F) is truncated by the overlying clayey diamicton of unit A. This diamicton is hard, but plastic, and apart from flecks of more vivid red clay, it shows little sign of internal stratification. Its base is slightly uneven and gradational over about 1 cm. Other sections lower down the hillside show that the basal till (unit J) reaches up to 8 m in thickness and varies laterally from being stiff and clayey, to friable and sandy. It rests on decomposed feldspathic andesite of the Montrose Volcanic Formation in a river cliff of the Bervie Water [NO 7642 8024].

The base of unit A incorporates wisps and fragments of peat where it overlies the truncated stretch of the organic sequence. The organic sequence itself is penetrated to a depth of about 0.8 m by a vertical, downward-tapering crack filled with red-brown clayey diamicton (Figure A1.31). The crack is either an ice-wedge pseudomorph, or a vertical fracture that has opened up owing to down-slope gravitational movement. The latter explanation is perhaps more likely as there is no evidence of any reorientation of clasts within the fissure infill, as would be expected if it were of periglacial origin.

Six samples taken from a monolith (Figure A1.32) for pollen analysis by M J C Walker (University of Wales, Lampeter, 1989), yielded relatively sparse assemblages. Only one hundred pollen grains were counted from each sample (Table A1.16). The pollen assemblages suggest an open, essentially treeless landscape, dominated by grass and sedge, with a number of herbaceous taxa present (e.g. Compositae, Cruciferae, Rumex and Caryophyllaceae). Tree pollen was sparse, although willow pollen was notable towards the bottom of the profile. The treeless nature of the landscape suggested by these assemblages is clearly indicative of interstadial or tundra climate, rather than warmer interglacial conditions.

The climatic inferences from the pollen analyses were supported by the results of plant macrofossil analysis by M Field (Keele University, 1991) (see (Figure A1.32) for sampling intervals). Sample B9 yielded 34 compressed Carex (sedge) fruits. Sample B10 yielded a well-preserved assemblage of macrofossils, but revealed the presence of only three taxa. A single seed of Viola sp. (violet), 81 compressed Carex fruits and hundreds of seeds of Monitia fontana subspecies fontana, a herbaceous annual to perennial of the ‘Blinks’ family (Stace, 1991) were recorded. The latter form is widely distributed at present in northern Britain and reaches north-westwards into northern Scandinavia. It occurs in damp habitats and its occurrence with Carex suggests a damp open landscape.

An initial radiocarbon age (GX–14723) of 12 460 ± 130 14C years BP was obtained for an acid-washed bulk sample of the silty peat. A sample taken subsequently was pre-treated to separate the alkali soluble (humic) and alkali insoluble (humin) components for independent age measurement. This was undertaken in order to identify any younger contaminants not completely removed by the pre-treatment. An age of 12 305 ± 50 14C years BP was obtained for the humic component (SRR–3687a) and 12 340 ± 50 14C years BP for the humin component (SRR–3687b) (Table 8). All of the radiocarbon dates indicate a Late-glacial Interstadial age for the lower part of the sequence. It is possible that the upper parts of the organic sequence may extend into the succeeding Loch Lomond Stadial, but at present, the evidence from the flora is inconclusive.

The stratigraphical, sedimentological and palaeontological studies of the sequence at Knockhill Wood and the radiocarbon dates suggest that the organic sediments were laid down in a cool damp environment, during the Late-glacial (Windermere) Interstadial. The organic sequence overlies a basal bouldery lodgement till of probable main Late Devensian age, laid down on top of weathered andesite bedrock. The basal till is assigned to the Mill of Forest Till Formation (Chapter 8). The origin of the overlying matrix-supported clayey diamicton is more problematic. Its compact structure and absence of well-developed internal stratification are typical attributes of lodgement or deforming-bed tills, rather than flow tills. No tills of Loch Lomond Stadial age are known from the district. Indeed, dating and palynological evidence from sites such as the Loch of Park indicate that the low ground was deglaciated throughout the Loch Lomond Stadial.

The upper diamicton has a gradational, but apparently conformable contact with the underlying organic sediments in the south-western part of the Knockhill Wood section, but it sharply truncates the upper part of the same sequence at the north-eastern (downslope) end of the exposure. This cross-cutting relationship, together with incorporation of fragments of peat in the till near the truncation surface, suggest that the diamicton has been emplaced as part of a landslip. The slip occurred by downslope movement along a gently curved, low-angle plane at the top of the laminated clayey silts (Figure A1.33). It is likely that slipping also occurred at the till/bedrock contact. If the tapering feature cutting the organic sequence is an ice-wedge cast, its presence would suggest that the movement probably occurred immediately following the Loch Lomond Stadial when rapid climatic amelioration took place.

Site 26 Burn of Benholm

The shelly pebbly clays and peat lenses exposed beneath the Mill of Forest Till at the Burn of Benholm, provide new insights into the glacial history of eastern Scotland prior to the Late Devensian. The peat lenses are remnants of an Early Devensian interstadial deposit of OIS 5c or 5a age. Amino-acid ratios indicate that shells in the underlying clays are of OIS 9 age or older, and the fabric and composition of the shelly sediments are consistent with their emplacement as glacially deformed rafts of marine sediment. The rafts were transported onshore when Scandinavian ice occupied the North Sea basin, during a pre-Late Devensian glacial episode.

Exposures in the banks of the Burn of Benholm, 14 km north of Montrose (Map 11), have for many years revealed a Quaternary sequence that includes a dark grey clay containing fragmentary marine shells, overlain by reddish brown till. In one exposure lenses of peat and organic silt intervened between the units (Plate 27); most of the lenses were incorporated within the red till, near to its base (Campbell, 1934; Donner, 1960, 1979).

Particular controversy has focused on the age and origin of the shelly clay and on the age and biostratigraphical significance of the peat lenses (Gordon, 1993d). The shelly clay has been interpreted as a till deposited by ice moving onshore, at a time when Scottish and Scandinavian ice sheets coalesced in the North Sea basin (Campbell, 1934; Bremner, 1938; Donner, 1960). More recently, Sutherland (1981) proposed that the shelly sediment is an in situ cold water marine deposit formed as a result of significant isostatic loading during an Early Devensian glaciation.

A pollen assemblage from peat collected by Campbell was subsequently described by Bremner (1943) as containing thermophilous tree pollen, and interpreted by him as indicating an interglacial environment. Later pollen analyses of the peat lenses, however, showed spectra dominated by non-arboreal types representing herb communities (Donner, 1960, 1979). This suggested an interstadial, rather than interglacial origin for the peat, which Donner (1960) initially interpreted as being of Late-glacial (Windermere) Interstadial age. He also proposed that the overlying red till was emplaced by landslipping, or gelifluction during the Loch Lomond Stadial a comparable origin to that proposed here for the Knockhill Wood succession (see above). A sample of the Burn of Benholm peat (Hel–1098) subsequently yielded a radiocarbon age of greater than 42 000 BP, which led Donner (1979) to reinterpret its age as being Early or Middle Devensian and the red till as being of the Main Late Devensian glaciation.

Because of the conflicting interpretations of the origin of the shelly clay and of the palaeoenvironmental conditions represented by the pollen in the peat deposits, a reinvestigation of the deposits at Burn of Benholm has been undertaken recently (Auton et al., 2000). The work included new excavations (Figure A1.34) adjacent to the stream sections first described by Campbell. One of these, trial pit BBP 4, revealed a new exposure of peat lenses beneath the base of the red till.

The Burn of Benholm site [NO 795 691] comprises a series of stream sections and trial pits at about 45 m OD along the Burn of Benholm. The lithostratigraphy is summarised in (Table A1.17). The sections and trial pits clearly show the Mill of Forest Till overlying Benholm Clay Formation in the manner described by Campbell (1934). In some of the exposures a sharp, undulating contact, varying from subhorizontal to steeply dipping, occurs between the two formations. In other sections, such as BBP 2, a zone of shearing and churning occurs with red-brown till mixed with a more gravelly unit at the top of the Benholm Clay. In trial pit BBP 4 sheared and folded lenses of peat and sandy silt occur within this gravelly unit. In section RC 1 deformed units of Mill of Forest Till and the underlying Birnie Gravel abut a knoll of andesite bedrock.

Clast orientations measured from the Mill of Forest Till all display tightly clustered fabrics suggesting subglacial deposition and indicate a north-easterly to east-north-easterly direction of ice movement. In contrast, eastwardly dipping stratification and north-westwardly dipping sheared contacts within the Benholm Clay, together with clasts of Mesozoic and Palaeozoic rocks are compatible with glacitectonic transport of the unit from offshore.

A sparse assemblage of fragmentary bivalve molluscs (some of which are ‘polished’ and striated) has been recovered from the Benholm Clay Formation. The assemblage included Arctica islandica, Macoma balthica, Cersatoderma edule and Mya sp. and is of boreal to low arctic aspect, but suggests no particular age other than, broadly ‘Late Pleistocene’. Samples from the Benholm Clay have also yielded rich palynomorph assemblages, which were completely dominated by Early Carboniferous, Early and Late Cretaceous and Palaeogene forms (Harland, 1993). This reworked material was probably derived from the North Sea basin. Small numbers of Quaternary dinoflagellate cysts were also recorded. Two main taxa were recognised: Protoperidinium sp. and Bitectatodinium tepikiense. Both are indicative of north-temperate to arctic, cold-water marine environments, possibly in close proximity to ice (Harland, 1983; Dale, 1985). A sparse assemblage of agglutinating foraminifera was found in two samples from the clay (Wilkinson, 1993). All of the foraminifera are considered to be reworked and, although their age is unknown, it is unlikely that they are younger than Palaeogene. The paucity of indigenous microfossils and the abundance of reworked material are consistent with interpretation of the clay as a former offshore deposit, reworked and transported onshore by ice.

Pollen analyses were carried out by M J C Walker on four samples of organic material from the Burn of Benholm Peat Bed exposed in Trial Pit BBP 4 and are reported fully in Auton et al. (2000). The pollen spectra in all four samples are broadly similar and are comparable to those reported by Donner (1960, 1979), although a wider range of taxa has been recorded. Collectively, the pollen evidence points to an open tundra vegetation of damp heath and grassland, with patches of low scrub, and isolated stands of tree birch probably in more sheltered localities. This confirms the view that the site probably does not contain an interglacial record; rather, it represents an interstadial episode with summer temperatures slightly below those of the present day.

A distinctive feature of the pollen spectra is the occurrence of a number of grains resembling Bruckenthalia spiculifolia. Bruckenthalia pollen has been found at a number of interglacial sites in north-west Europe (e.g. Turner, 1970; Menke and Behre, 1973; Phillips, 1976; Hall, 1980), although the species appears to be more typically associated with Early Weichselian to Early Devensian Interstadial deposits (Whittington, 1994). Indeed, Bruckenthalia spiculifolia has been cited as a biostratigraphical marker for Early Weichselian Interstadials (Pons et al., 1992). Bruckenthalia has also been found at the Crossbrae Farm and Camp Fauld (Moss of Cruden) sites described above, which may also be of Early Devensian age (Whittington et al., 1993, 1998; Whittington, 1994).

Three sets of ‘total’ amino-acid D/L (D-alloisoleucene: L-isoleucene) ratios have been determined on fragmentary shells of Arctica islandica collected from the Benholm Clay Formation in section RC 1. They are 0.37 (AAL 2950), 0.37 and 0.38 (AAL 2952) and 0.34 ± 0.024 (LOND 400). The third analysis (Lond 400) is regarded as the most accurate.

Conventional radiocarbon dating was carried out on a single bulk sample taken from a lens of the Burn of Benholm Peat in the upper part of the Benholm Clay Formation, between 2.0 and 2.3 m depth in trial pit BBP 4. The acid-insoluble residue was subdivided into its alkali soluble (humic) and alkali insoluble (humin) components for independent age measurement. The humic carbon sample, SRR–5201a, gave a radiocarbon age of greater than 50 850 14C years BP; the humin sample, SRR–5201b, one of greater than 50 600 14C years BP. Both dates are older than the upper limit of the radiocarbon chronology and provide a minimum (Mid-Devensian) age for the Benholm Peat.

The interstadial affinity of the pollen, the presence of Bruckenthalia spiculifolia and infinite radiocarbon dates from the Burn of Benholm Peat Bed, all suggest that the peat lenses are remnants of an Early Devensian interstadial deposit of OIS 5c or 5a age or older (Table 7).

Amino-acid ratios from the fragments of Arctica islandica suggest that the shells within the Benholm Clay Formation are of OIS 9 age or older. The fabric and composition of these shelly sediments are consistent with their emplacement as immature deformation till, formed by partial homogenisation of glacially transported rafts of marine sediment. The amino-acid ratios suggest that this till was deposited by ice during a post-OIS 9/11 glaciation. It may be derived from offshore sediments within the Coal Pit Formation, which is of probable Saalian to Weichselian age (Stoker et al., 1985), or from the Ling Bank Formation (which contains an interval of Holsteinian age), or correlate with a till within the Fisher Formation, of supposed Saalian age (Gatliff et al., 1994) (Chapter 5, (Figure 33)). Folded and sheared contacts between the shelly deposits, the peat lenses and the Mill of Forest Till indicate that the fossiliferous sediments were glacitectonised during the Main Late Devensian glaciation, when ice from within Strathmore overrode the site from the south-west.

Appendix 2 Bulk mineral resources

Description of sand and gravel resources by sheet area

The general extent of resources of sand and gravel in north-east Scotland is outlined in Chapter 2. The more detailed evaluation of individual accumulations on each sheet given below allows a greater appreciation of the variability, economic importance and abundance of the deposits. A wealth of mineral assessment data has been obtained since the 1940s, but the coverage is patchy.

A set of generalised maps (Maps 1–11) show the distribution of a selection of deposits, landforms and localities mentioned in the text. The form and extent of the most notable sand and gravel deposits are also described in Chapter 6.

History of resource appraisal

The first systematic descriptions of sand and gravel resources in north-east Scotland were published by the Geological Survey of Great Britain in Wartime Pamphlet 30 (Parts I and II). Part I dealt with deposits in the vicinity of Elgin, Banff and Aberdeen (Anderson, 1943); deposits around Stonehaven were considered in Part II (Anderson, 1945). A later descriptive account of the resources within the whole onshore area was published by the Institute of Geological Sciences (Peacock et al., 1977).

The first detailed assessment of sand and gravel resources, involving the logging and sampling of boreholes and trial pits sunk specifically for assessment purposes, was undertaken around Garmouth (Spey Bay). It is described in Mineral Assessment Report (MAR) 41 (Aitken et al., 1979) (Figure A2.1). This was followed by Mineral Assessment Reports 58 and 76 detailing sand and gravel resources around Peterhead (McMillan and Aitken, 1981) and Ellon (Merritt, 1981), respectively. Preliminary studies were subsequently made of the sand and gravel deposits around Aberdeen (Merritt and Peacock, 1983a), Inverness, Nairn, Forres and Elgin (Merritt and Peacock, 1983b), and Strathmore (Aitken, 1983). Two later reports (MARs 146 and 148) detail resources around Aberdeen (Auton and Crofts, 1986) and around Inverurie, Dunecht, Banchory and Stonehaven (Auton et al., 1988).

The assessment data from MARs 58, 76, 46 and 148 were subsequently re-evaluated and condensed in a ‘summary assessment report’ covering the eastern part of the district (Merritt et al., 1988). This contains derivative maps, including ‘target’ resource maps at 1:50 000 scale, which show the thickest and more laterally extensive of the dry deposits of gravel and sand, together with the more extensive water-saturated gravels that could be exploited by dredging. It also presents analyses of the deposits for end-use suitability and illustrates the results as spider diagrams (compare with Laxton, 1992). Eighty four target resources were recognised and 20 ‘prime targets’ identified.

It is emphasised that the suitability criteria presented in the report, and summarised at the end of this account, are generalised and based on computer manipulation of particle size analyses (gradings) mainly from borehole samples. The analyses also including some ‘as dug’ material from trial pits, workings and exposures. Sampling was undertaken at points within large spreads of inherently heterogeneous material. The results do not imply that aggregate from specific workings, or from within individual resources, could not meet some specific end-use requirements, following suitable onsite screening, washing, crushing and grading.

A further detailed assessment (MAR 149) was completed following the summary assessment report. It describes sand and gravel resources around Strachan, Auchenblae and Catterline (Auton et al., 1990).

The assessment surveys carried out between 1977 and 1990 were commissioned by the Scottish Development Department of the Scottish Office. Those since 1981 were funded by a consortium, which also included the Regional and District councils and representatives of the local aggregates industry. The number of boreholes, trial pits, workings and natural sections examined, and of resistivity soundings taken, during the course of each assessment is shown in (Table A2.1).

Sheet 95 Elgin

The spreads of sand and gravel on Sheet 95 (Map 1) constitute some of the most extensive resources in north-east Scotland. They have been assessed in detail only in the vicinity of Garmouth (Resource Sheet NJ 63), but are described more generally by Peacock et al. (1968) and as part of the resources of the Moray district (Peacock et al., 1977).

Much of the glaciofluvial sand and gravel between Elgin and the Spey valley occurs in coarsening-upward sequences that were laid down as fans at the mouths of drainage channels or in temporary ice-dammed lakes. Overall, these deposits fine north-eastwards and much of the sequence comprises silty and clayey glaciolacustrine sediment with a capping of sand and gravel. Other kettled terraced spreads were laid down on bodies of stagnant ice during deglaciation. Moundy sandy deposits predominate around Lhanbryde [NJ 274 612], west of Elgin, and towards the coast around Binn Hill [NJ 305 656]. The topography of the ground to the north and north east of Elgin is generally more subdued and is underlain by sand, silt and clay with minor amounts of gravel. These fine-grained sediments are commonly capped by or merge with gravelly Late-glacial raised beach deposits. The sandy glaciofluvial deposits around Elgin are worked in a pit at Woodside [NJ 237 632] and up to 15 m of fine-grained, bedded sand with minor amounts of pebble and cobble gravel was formerly worked in a pit at Duffushillock [NJ 216 596], in the moundy deposits south of Elgin.

The flights of terraces flanking the lower reaches of the Spey and Lossie comprise thick, easily worked resources of sand and gravel, lying above the water table. The kettled upper terraces flanking the Spey are composed of glaciofluvial gravel, which merges into moundy ice-contact deposits around Garmouth. Some of the lower gravel terraces are flat-topped and probably of Late-glacial age. An extensive gravel terrace on the western flank of the River Lossie is worked at Lochinvar Pit [NJ 183 614] ( Plate 10a). The gravelly alluvial deposits of the river valleys are less attractive, as most of the resource lies below the water table and could only be worked by dredging. The postglacial raised beach gravels have been worked extensively for shingle, but contain little sand.

Apart from the sand and gravel which flanks the coast between the mouth of Tynet Burn and Buckie, the only other resources of sand and pebbly sand occur as small moundy spreads on the interfluve to the east of the Spey. Exposures in the coastal area suggest that the former deposits are composed mainly of medium- to fine-grained sand. About 8 m of cross-bedded sand was recorded from a pit immediately west-north-west of Mains of Tannochy [NJ 386 637] and a similar thickness of more pebbly sand was formerly exposed in a pit on the eastern bank of the Gollachie Burn at [NJ 406 647].

The average grading of the near-surface gravelly deposits flanking the River Spey between Fochabers and the coast, is illustrated in Aitken et al. (1979; fig.2). This shows that most of the resources consist of gravel and sandy gravel. The uniform composition of the clasts from these deposits is illustrated in (Figure A2.2). Tough clasts of psammite and quartzite predominate and the proportion of non-durable rock types is minimal. The grading and composition of aggregates from the Garmouth area are probably representative of most of the sands and gravels in the lower reaches of the Spey valley.

Sheet 96W Portsoy

Deposits of sand and gravel are less widespread in the Portsoy district (Map 2). Notable resources are present within terraced glaciofluvial deposits in the valley of the Burn of Boyne and which extend across the interfluve to reach the Burn of Durn south of Portsoy. Less extensive terraced deposits occur in the valley of the Burn of Durn, upstream of Damheads [NJ 580 635] where they have been worked in a pit near Butterytack [NJ 568 613].

Much of the town of Cullen is built on sandy glaciofluvial deposits, which were formerly worked at Gallows Hill [NJ 513 664]. Potentially workable deposits of sand may occur beneath silts and clays in the valley of the Burn of Cullen, but they would lie close to, or below, the water table. Other resources include the well-rounded cobble gravels forming moundy topography along the coast, inland of Whyntie Head [NJ 630 660], and sands that underlie rounded hillocks in the vicinity of Home Farm [NJ 500 661], on the western side of the valley of the Burn of Cullen. Extensive, high quality resources may occur within the glaciofluvial sheet and ice-contact deposits, in the vicinity of Headtown [NJ 612 583]. These ice-contact deposits extend beyond the southern margin of Sheet 96W, where they are seen to be predominantly composed of fine- to medium-grained sand.

Sheet 96E Banff

The most notable sand and gravel resources of Sheet 96E (Map 3) lie west of the valley of the River Deveron, between Banff and Boyndie [NJ 640 638]. Glaciofluvial deposits assigned to the Blackhills Sand and Gravel Formation (see Chapter 8) form the Hills of Boyndie, stretching south-westwards from Banff to Ladysbridge [NJ 651 636]. Sand predominates, but gravel is more abundant towards the margins of the spreads. Clasts of quartzite, feldspathic red sandstone, granite and basic igneous rocks predominate within the gravel.

A significant resource of glaciofluvial sand and gravel is present on the western side of the valley of the River Deveron, but building developments in Banff have extended across the northern portion. Less attractive resources of water-saturated sand and gravel occur beneath the floodplain of the Deveron, underlying 2 to 3 m of alluvial silt. Farther east, in the vicinity of Paddocklaw Farm [NJ 662 617], Peacock et al. (1977) record sand and gravel containing clasts of metagreywacke, gneissose psammite, semipelite and slatey pelite. Notable resources may also be present within the glaciofluvial deposits flanking the Burn of Fishrie near its confluence with the River Deveron.

Sheet 97 Fraserburgh

The principal resources of sand and gravel on Sheet 97 (Map 4) are of four types.

Moundy glaciofluvial ice-contact deposits, which occur between New Aberdour and Loch Strathbeg.

Peacock (1983) has described these deposits in some detail. The glaciofluvial deposits are assigned to the Blackhills Sand and Gravel Formation (Chapter 8). Although predominantly sandy, the kettled, ice-contact deposits are typically very variable in grain size over short distances. Many of the deposits occur as coarsening-upward deltaic sequences (Peacock and Merritt, 2000d). As well as the ubiquitous clasts of hard metamorphic and igneous rocks, some of the more gravelly units contain sparse shells and clasts of soft sedimentary rocks (of Triassic, Jurassic and Early Cretaceous ages). Typical sections occur in pits near Blackhills at [NJ 926 609 and [NJ 926 615] (Plate 10b).

The terraced glaciofluvial deposits are composed predominantly of gravel derived from the local bedrock, but some of the lower lying spreads pass laterally into silt and clay. Considerable amounts of blown sand have accumulated adjacent to Loch Strathbeg. Along the coast, the blown sand is mainly medium grained and composed of quartz with minor amounts of feldspar; shell fragments and small pebbles are present locally.

Several deeply weathered outcrops of Devonian conglomerate constitute potential sources of coarse gravel in the area to the west of New Aberdour. This material, which is decomposed to a depth of several metres locally, also extends on to the adjoining Banff sheet.

Sheet 86E Turriff

The principal resources of sand and gravel on this sheet (Map 5) underlie high-level (glaciofluvial) and lower lying (alluvial) terraces within the valleys of the River Deveron and its main tributaries, the Idoch Water and the Burn of Fishrie. Similar, but less extensive deposits flank the ‘misfit’ valley (Towie Spillway) between Turriff and Fyvie. Significant resources of quartz-rich gravel are also present within outliers of the Neogene Windy Hills Gravels Member of the Buchan Gravels Formation at Windy Hills (Plate 4a), (Plate 4b), north-east of Fyvie (see Chapter 4 and Appendix 1); other smaller outliers occur at Dalgatty Wood and Delgaty near Turriff.

Minor scattered resources of gravel and sand form eskers and small kames in the Alvah–Rosyburn area, and moundy topography along the valley of the Idoch Water, in the vicinity of Balquhindachy Farm [NJ 763 487]. Deeply weathered Devonian conglomerates within the Crovie Sandstone Group constitute a minor resource of clayey bouldery gravel in the Turriff–Fyvie area and flat-lying spreads of waterlogged sand and gravel occur within the alluvium and alluvial terraces of the River Ythan and its headwater tributaries.

The deposits underlying the high-level terraces in the Deveron catchment vary considerably in composition. Till, with pockets of sand and gravel, underlies the terraces in the Deveron valley to the north of Turriff. Most of the glaciofluvial terraces around the town itself are composed of sand, with subordinate amounts of gravel, composed of psammitic, pelitic, and granitic clasts, as well as pebbles of sandstone. Gravels predominate upstream of Turriff and along the course of the Burn of King Edward. Clasts of psammitic metamorphic rocks and sandstone characterise the King Edward deposits; slate, feldspar-porphyry and decomposed basic igneous rock clasts are additional components of the gravels in the upper Deveron valley.

Gravels within the northern part of Towie Spillway are poorly sorted and composed of subrounded clasts of granite, sandstone, and chert, as well as of psammitic and pelitic metamorphic rocks. Towards Fyvie, the gravels are mostly composed of rounded clasts of quartzite and sandstone, which are largely derived from the underlying Devonian conglomerate bedrock. The gravel deposit at Windy Hills was evaluated in MAR 76 (Merritt, 1981) and is described in detail in Appendix 1 Site 13. More than 95 per cent (by weight) of this cobbly deposit is formed of water-worn clasts of white vein quartz and quartzite ((Figure A2.2)). The deposit at Delgaty contains a notable proportion of brown sandstone pebbles in addition to the predominant quartzose clasts, whereas the Dalgatty Wood deposit contains notable amounts of white sand; neither has been the subject of detailed assessment.

Sheet 87W Ellon

Most of the sand and gravel resources in this district (Map 6) form terraced glaciofluvial sheet deposits. Those in the northern portion of the sheet occur within the Ugie Water catchment and were included in the ‘Peterhead’ sand and gravel assessment (MAR 58). Deposits in the southern portion lie within the Ythan catchment and were evaluated as part of the ‘Ellon’ assessment (MAR 76). In the north, sands and gravels underlying discontinuous glaciofluvial terraces flanking the North Ugie Water downstream from Strichen constitute the most important resources; water-saturated gravel is also present beneath the floodplain. Smaller spreads of glaciofluvial sand and gravel flank the floodplain of the South Ugie Water, upstream of Old Deer, which is also built upon a glaciofluvial terrace. Similar terraced deposits extend along the valleys of the southern tributaries of the South Ugie Water, in the vicinity of Fordmouth [NK 012 457] and Stuartfield [NJ 973 457], but much of the latter resource is sterilised by buildings. Most of the readily available resources in this area were exploited during the 1970s and 1980s.

Glaciofluvial terrace deposits also form the most extensive resources in the Ythan catchment. These terraced deposits have been extensively worked downstream of Methlick. Most of the resources consist of coarse gravel (boulders over 1 m in diameter are common) lying above the water table. A less attractive resource of gravel, occurring almost entirely below the water table, underlies the river floodplain; it was formerly dredged at a pit [NJ 922 321] near Ardlethen. Terraced glaciofluvial sand and gravel associated with a former course of the Ythan, between Ellon and its present estuary, also constitute a significant resource of sand and gravel. It has been worked extensively in a pit to the north of Deep Heather [NJ 976 290]. An isolated deposit of sand and gravel was worked in a large pit (now filled-in) at Tillybrex [NK 002 348]. Moundy deposits in the vicinity of Cross Stone [NJ 954 279] are currently being worked.

The composition of gravels on Sheet 87W is illustrated in (Figure A2.2), which shows that psammitic clasts predominate, although granitic clasts constitute a significant component. Small amounts of deleterious material are generally present and pebbles of basic igneous rocks are numerous within the water-saturated gravels in the vicinity of Ardlethen. Most of the resources in the Ythan valley contain material suitable for a wide variety of end-uses, but those within the catchments of North Ugie Water and South Ugie Waters possibly suit fewer types of end-use. For example, only a limited amount of material matches the end-use criteria for coarse asphalt aggregate. However, fine aggregate potentially suitable for mortaring and concrete is present in some of the more sandy terraced deposits.

Sheet 87E Peterhead

As on Sheet 87W, most of the sand and gravel deposits in the northern part of this sheet (Map 7) were included in the ‘Peterhead’ resource survey; those in the southern portion were evaluated as part of the ‘Ellon’ assessment. The most extensive resources in the northern half of the sheet occur as terraced spreads of glaciofluvial gravel and sand around the confluence of the North and South Ugie Water and along the tributaries of the River Ugie, in the vicinity of Longside. There has been considerable exploitation of the terraced sand and gravel in the valley of the North Ugie Water, but a significant proportion of the resource remains; water-saturated gravel is also present beneath the floodplain. The terraced deposits to the south of Longside are generally more sandy and have not been worked as intensively.

Assessment boreholes sunk in alluvial deposits within the valley of the River Ugie indicated only minor resources of variable thickness and extent. The patchy deposits of sand and gravel that underlie degraded terraces flanking the Burn of Faichfield are probably more attractive resources, as are the glaciofluvial deposits around Blackhills [NK 050 527] and Artlaw [NK 073 501]; all have been worked in several places.

The coastal sand dunes north of Peterhead constitute a significant resource of medium-grained sand, with shell fragments locally abundant. Raised beach gravel is known to underlie the sand at several localities. Smaller spreads of blown sand are present inland of the Bay of Cruden and within the Sands of Forvie. The latter lie within a nature reserve, which precludes their exploitation.

In the southern part of Sheet 87E, the main deposits of glaciofluvial sand and gravel occur as relatively thin, discontinuous spreads, forming mounds and ridges on both sides of the boundary between the Logie-Buchan and East Grampian drift groups (Figure 50). Notable resources within the Logie-Buchan Drift Group occur in the vicinity of Hatton and also between Hillhead [NK 022 340] and Mains of Collieston [NK 032 292]. The latter include the Kippet Hills Esker (Plate 15), described in Appendix 1 (Site 16). Discontinuous resources of glaciofluvial sand and gravel and sandy glaciolacustrine deposits occur beneath clayey overburden between Hatton and Cruden Bay; waste partings of laminated silt and clay are commonly present within the sequences. Similar glaciofluvial–glaciolacustrine successions also extend from Lochlundie Moss to beyond the southern margin of Sheet 87E. Glaciofluvial sands and gravels are less widespread within the East Grampian Drift Group, though they have been worked near Moreseat [NK 054 404], North Aldie [NK 072 409] and Redleas [NK 091 428].

The Palaeogene–Neogene gravels of the Buchan Ridge Gravels Member crop out along the ‘Buchan Ridge’, between the Moss of Cruden and the Hill of Aldie, and beyond the ridge, at Den of Bodham. Unlike the quartzose gravels at Windy Hills, these deposits are primarily composed of flint cobbles in a matrix of kaolinitic clay and quartz sand (largely formed from highly decomposed clasts of igneous and metamorphic rocks). These gravels are regarded as being too coarse and clayey to be a particularly attractive source of aggregate, but they are a source of ornamental cobbles. These deposits are described further in Chapter 4 and Appendix 1 (Site 14).

Granite, decomposed to sand, has been exploited quite widely for making tracks and for use in foundations. It was dug on the Hill of Longhaven for bedding oil and gas pipelines.

The sand and gravel in the northern part of Sheet 87E are similar in composition to most of those on Sheet 87W and contain material suitable for fine asphalt aggregate, plastering/mortar sand and fine concrete aggregate. Clasts of psammitic and granitic rocks predominate (Figure A2.2) and a notable proportion of flint is present in the glaciofluvial gravels south-east of Longside. Much of the granitic material is probably derived from the Peterhead Granite; the flint is probably reworked from the Buchan Ridge Gravels Member.

In the southern part of the sheet a high proportion of platy clasts of limestone and calcareous siltstone occur within gravels of the Logie-Buchan Drift Group. The deposits were formerly burnt for the local production of lime. The calcareous clasts were probably derived from Mesozoic rocks in the adjacent offshore area. The gravels yield coarse- and fine-grained aggregates that appear to be suitable for most end uses.

Sheet 76E Inverurie

The most extensive spreads of sand and gravel on Sheet 76E (Map 8) occur within the valley of the River Don, between Kintore and Inverurie, and between Kemnay and Monymusk. Scattered moundy and terraced spreads of glaciofluvial sand and gravel occur also on the sides of tributary valleys such as the Kinnernie Burn (around Dunecht), and the Leuchar and Gormack burns farther south. Minor resources are also present in the valley of the River Urie (upstream of Inverurie), the valley of the Ton Burn and those of its tributaries (around Sauchen) and the Beltie Burn (around Torphins). The principal resources within the valley of the River Don were evaluated in Mineral Assessment Report 146 and those around Inverurie and Dunecht were evaluated as part of MAR 148. Resources within the remainder of the sheet are described briefly in Gould (1997).

The most attractive resources between Kintore and Monymusk occur within moundy and terraced spreads of glaciofluvial sand and gravel on the southern side of the Don floodplain. These merge with sands and gravels laid down at both ends of a major east–west-trending glacial meltwater channel (the Tom’s Forest Channel) which dissects the interfluve between Kintore and Kemnay. Much of the resource lies above the water table. The deposits vary in composition, from well-bedded sand and fine gravel, to clast-supported boulder gravel and silty sand. Well-bedded sand and gravel was formerly worked in a pit at Tavelty [NJ 780 173], north-west of Kintore. Similar deposits forming a kame-terrace, between Kintore and Inverurie, have been worked extensively also, but much of the remaining resource is sterilised by urban development. Moundy glaciofluvial deposits to the north of Kemnay have been worked in pits near Mill Farm [NJ 736 178], but significant resources remain to the south-east of the village.

Boulder gravels, which are not particularly attractive sources of aggregate, are typical of deposits that form esker ridges, such as the discontinuous Kemb Hills Esker. The latter, which extends for about 5 km to the south-west of Kemnay, generally stands about 6 m above the adjacent gravelly glaciofluvial terrace.

A large resource of gravel underlies the floodplain of the River Don, between Kintore and Inverurie, but it lies entirely below the water table. The deposit appears to thin and become more sandy downstream, where it occurs beneath an increasing thickness of alluvial clay and silt. Smaller resources of waterlogged gravel and sandy gravel underlie the Don floodplain between Burnhervie [NJ 732 194] and Port Elphinstone [NJ 777 202], where they abut fragmentary kame-terraces, and are composed of coarse gravel.

Waterlogged deposits of gravel and sand also continue beneath the Don floodplain between Kemnay and Monymusk. The floodplain is flanked by discontinuous river terraces, mainly composed of fine-grained sand and silt, as well as by terraced gravelly glaciofluvial deposits. The latter merge into glaciofluvial sheet deposits at the western end of a large glacial drainage channel. The sheet deposits were exposed in a pit [NJ 693 173] at Rothens, where they comprised clean, clast-supported gravel with subordinate beds of sand and clay-bound gravel. Thin, flat-lying spreads of water-saturated gravel are also present in the floor of the drainage channel, near Red Moss [NJ 705 179]; they abut thicker deposits of poorly sorted, matrix-rich gravel that have been worked north-east of Blairdaff [NJ 696 176]. The grading of potentially workable sand and gravel from sample points in the Kintore–Kemnay–Monumusk area is illustrated in Mineral Assessment Report 146 (fig.4).

The spreads of glaciofluvial sand and gravel that flank the valleys of the Kinnernie, Leuchar and Gormack burns, and their tributaries, are generally thinner and less homogeneous than those within the valley of the River Don. Nevertheless, resources of sandy gravel, lying above the water table are widespread in these smaller valleys. They occur as moundy deposits in the valley of the Kinnernie Burn, upstream of Dunecht, as mounds, esker ridges and flat-lying spreads in the valley of the Leuchar Burn, downstream of Garlogie [NJ 783 054], and as mounds, up to 10 m high, near Easter Finnercy [NJ 767 042]. Minor resources are also present underlying discontinuous river terraces of the River Urie, upstream of Inverurie.

The moundy glacial deposits that are widespread in the valleys of the western part of the sheet are generally too poorly sorted and bouldery to be considered attractive sources of aggregate. However, small resources of cobble gravel are present in the esker ridges associated with the moundy glacial deposits in the valley of the Ton Burn, and 5 m of well-bedded sand and gravel was exposed in a small pit [NJ 673 109] within terraced glaciofluvial deposits in the vicinity of Monyroads. Some of the glaciofluvial deposits around Sauchen and those in the valley of the Beltie Burn also contain notable amounts of sand and gravel lying above the water table.

Discrete areas of deeply decomposed granitic bedrock, notably on the flanks of Bennachie and around Easter Tolmauds [NJ 626 075], constitute sources of aggregate suitable for use as fill and in the construction of untarred roads.

All of the glaciofluvial deposits on Sheet 76E were laid down by meltwater associated with the decay of ice that moved from west to east across the area and eroded the underlying Dalradian metamorphic and Caledonian igneous bedrock. Consequently, the glaciofluvial gravels are predominantly composed of resistant clasts of metamorphic rock (mainly psammite) and igneous rock (mainly granite). Much of the sand is of granitic composition, being largely derived from local outcrops of deeply weathered granite. The alluvial deposits are of similar character, being largely composed of reworked material from the underlying glacial and glaciofluvial sediments. Numerous clasts of coarse- and fine-grained basic igneous rocks occur in gravels from the Urie catchment (Figure A2.2); they are largely derived from the Insch basic igneous intrusion, which crops out in the northern part of the area.

The aggregates appear to be suitable for most end uses, although there are some local variations. The sands and gravel in the catchments of the Kinnernie, Leuchar and Gormack burns, for example, are more sandy and better sorted than deposits in the valley of the River Don, and thus meet a greater range of end-use criteria (for details see Merritt et al., 1988). Sands for use in asphalting are known to be in relatively short supply in the Aberdeen market area, particularly as suitable resources close to the city are largely worked out. The sandy glaciofluvial terraced deposits around Kintore may provide alternative sources of medium- and fine-grained sand.

Sheet 77 Aberdeen

Significant spreads of sand and gravel are present on Sheet 77 (Map 9). Their proximity to Aberdeen, and the major market for building aggregate in north-east Scotland, has meant that many of the more attractive resources, such as those in the vicinity of Leuchlands [NJ 934 133], have been worked out; others have been sterilised by urban expansion. Mapping, undertaken as part of the Aberdeen sand and gravel assessment (MAR 146) and as part of an on-going programme of continuous revision, indicate that the resources on Sheet 77 are much less extensive than might be inferred from the 1:50 000 drift geological map published in 1980. The principal resources occur in three main areas.

Other resources are present within the sand dunes that extend southwards from Sands of Forvie to Bridge of Don. They constitute a potential resource of clean, medium- to fine-grained quartz sand, which, in places, overlies post-glacial raised beach deposits and glaciofluvial sands and gravels. Although the blown sand and underlying deposits have in the past been worked in pits at Blackdog Rifle Ranges [NJ 962 153] and Blackdog Rock [NJ 964 138], much of the remaining resource occurs within environmentally sensitive areas, such as nature reserves, or underlies golf courses.

Most of the sand and gravel deposits within the City of Aberdeen have either been worked or sterilised by urban and industrial development. These include the terraced spreads on the northern side of the Dee floodplain downstream of Peterculter, the moundy deposits near Nigg Bay [NJ 967 047], and those between Denmore [NJ 936 111] and Bridge of Don. Much of the urban development in the latter area has taken place since the summary assessment was made in 1988.

The sand and gravel between Newburgh and Denmore forms discontinuous mounds and ridges along the boundary between the Logie-Buchan and East Grampians drift groups. The potentially workable deposits are generally thin and sandy, but gravelly units are also present locally. Much of the resource is concealed beneath a variable thickness of overburden, comprising sandy till or a mixture of till and silty glaciolacustrine sediment; waste partings of till and silt are also common. The thickest gravel deposits form a steep-sided discontinuous ridge, 8 to 10 m high, between Hatterseat [NJ 982 215] and Newburgh; they have been worked in several pits near Drums [NJ 986 226]. The resources are less extensive than might be inferred from the 1:50 000 drift map, but a BGS borehole at the southern end of the ridge, near Hatterseat, proved 7.1 m of poorly sorted sandy gravel resting on bedrock (Auton and Crofts, 1986). Another BGS borehole, near Pitscaff Croft [NJ 989 235], proved 9.4 m of sandy gravel beneath 4.7 m of overburden. The gravels are composed mainly of clasts of quartzite, psammite and vein quartz.

Little remains of the former thick deposits of glaciofluvial sand and gravel between Corby Loch and the coast, which have been exploited for aggregate in 19 major workings (see Auton and Crofts, 1986, tables 24 and 25). Few of these pits are now active, but currently notable amounts of coarse- and fine-grained aggregate are being extracted at Lochhills (Bishops Loch) pit. The sands and gravels formed a complex pattern of mounds and ridges (see Munro, 1986, fig.36) but the exact distribution of sheet and moundy ice-contact deposits is now difficult to determine, owing to the extent of worked-out and reinstated ground. The workable material ranged from 1.0 to over 13.9 m, averaging about 6 m in thickness.

The deposits forming the flat-topped mounds generally coarsen upwards, from ripple-laminated fine-grained sand and silt, into well sorted coarse gravel and sand with planar and trough cross-bedding. These workable resources are commonly concealed by thin discontinuous spreads of sandy ‘flow till’. The sand and gravel forming the ridges is generally more poorly sorted and cobbly than that underlying the mounds and discontinuous waste partings of clayey gravel and till are common. Exposures in some of the sequences are illustrated in Aitken (1998).

The terraced spreads of sand and gravel in the valley of the River Don contain significant resources of coarse and fine aggregate, lying above the water table. Similar water-saturated material is present beneath the flood-plain. The terraced deposits were formerly worked in pits at Nether Kirkton [NJ 880 143]. Larger workings occur nearby, at Mill of Dyce (St Fergus), within moundy ice-contact sand and gravel that locally exceeds 16.2 m in thickness. These deposits, which are largely worked out, were seen to coarsen upwards, from silty and clayey sand into poorly sorted cobble and boulder gravel; they are described in detail in Appendix 1 (Site 18).

Attractive resources of aggregate lying above the water table are present within the dissected glaciofluvial terraces on the southern side of the valley of the River Dee, downstream of Templar’s Park [NJ 845 999]. The deposits average 4.3 m in thickness, but over 10 m of fine-grained silty sand were exposed in abandoned workings 200 m east of Templar’s Park, in 1999; up to 6 m of clast-supported gravel was recorded from working faces in Blairs Quarry [NJ 880 011] during 1988. Major resources of water-saturated gravel (with a mean thickness of 9.4 m) are also present beneath the river flood-plain, overlain by thin clayey alluvium.

Minor resources of moundy sand and gravel, which have been worked in places, are present on the Don–Dee interfluve near Borrowstone [NJ 849 078], Cairdhillock [NJ 848 068], East Brotherfield [NJ 844 043], Foggieton [NJ 867 038] and Easter Auguston [NJ 823 017], near Peterculter. Small deposits of sand and gravel forming mounds and esker ridges, were identified in the valley drained by the Green Burn, in the vicinity of Craibstone College of Agriculture [NJ 866 112], during surveys undertaken as part of the summary assessment study (Merritt et al., 1988). Up to 4 m of poorly sorted cobble gravel was exposed near the crest of the largest esker, and degraded faces in workings at the western end of the ridge indicate that the deposits have been exploited to depths of 12 to 15 m.

The grading of the most notable deposits of sand and gravel on Sheet 77 is illustrated in Mineral Assessment Report 146 (fig.4). This shows that gravels and sandy gravels predominate in the valleys of the rivers Don and Dee and their main tributaries, and that coarsening-upward successions dominate the more sandy sequences between Balmedie, Bridge of Don and Dyce.

The gravels on Sheet 77 are similar in composition to the bulk of those on the adjoining Sheet 76E Inverurie. Clasts of psammitic and granitic rocks predominate ((Figure A2.2), (Figure A2.3). Much of the granitic material is derived from the local granite intrusions at Clinterty, Crathes and Aberdeen; the clasts of basic and ultrabasic igneous rocks are probably derived mainly from the intrusion at Belhelvie. Small numbers of limestone and calcareous siltstone pebbles are present in gravels within the Logie-Buchan Drift Group along the coast north of Bridge of Don.

The gravels between Hatterseat and Newburgh and those within the valley of the River Dee, downstream of Peterculter, appear to be particularly suitable for coarse aggregate for road sub-base, asphalt and concrete. The workable deposits around Corby Loch, together with those in the valley of the River Don, between Mill of Dyce Pit and Aryburn [NJ 898 131] are potentially suitable for most end uses. The sandy deposits between Sheilhill [NJ 935 129] and Millden [NJ 971 163] were particularly suitable for use as asphalting sand, but the resource is largely worked out.

Sheet 66E Banchory

The principal resources of the northern part of Sheet 66E (Map 10) occur within glaciofluvial and alluvial deposits in the valleys of the River Dee, Burn of Sheeoch and Burn of Canny. Extensive spreads of sand and gravel are also present in the valley of the Water of Feugh, in the vicinity of Strachan. Minor resources occur within terraced alluvial deposits and moundy glaciofluvial deposits in Glen Dye and as flat-lying spreads of sand, in a small lake basin near Lochhead of Leys [NO 695 979] and in a larger basin around Loch of Park [NJ 770 990].

Some potentially workable coarse aggregate, mixed with bouldery diamicton, is present within the moundy glacial deposits forming moraines to the north and east of Loch of Park. Similar morainic sediments also occur in the valleys of the Water of Aven, the Burn of Melmannoch, the Burn of Knock, and on the southern side of the valley of the Water of Feugh, in the vicinity of Powlair [NO 620 912]. All are potential sources of low-grade aggregate. Discrete areas of deeply decomposed granitic bedrock, around Loch of Park, Powlair and the upper reaches of Garrol Burn also constitute sources of aggregate suitable for use as fill and road base.

In the southern part of Sheet 66E, the main spreads of sand and gravel form moundy topography on the northern margin of Strathmore, notably between Auchenblae and Drumlithie, and around Fettercairn. Less attractive deposits are present beneath the floodplains of the Bervie Water, Luther Water, Black Burn, Devilly Burn and Burn of Cauldcots, as well as in the discontinuous terraced spreads on the sides of the valleys. Deeply weathered exposures of Lower Devonian conglomerate north of the Glen of Drumtochty, constitute a resource of cobbly gravel which, in places, exceeds 3 m in thickness.

The sand and gravel resources in the Dee valley, downstream of Banchory and within the valley of the Burn of Sheeoch, were evaluated as part of Appendix 2 Bulk mineral resources Mineral Assessment Report 148. Those in the vicinity of Strachan and in Strathmore were included in MAR 149. The sands and gravels that crop out in the south-west corner of Sheet 66E have not been assessed in detail. They are described in Carroll (1995d).

The kettled terraces that flank the Dee, between Banchory and the eastern margin of Sheet 66E, contain extensive resources of well stratified gravel and sand, lying above the water table; they average 7.4 m in thickness. Gravelly deposits, beneath negligible overburden are also present underlying the floodplain; they average 10.4 m in thickness. The terraced glaciofluvial gravels have been worked to a depth of greater than 10 m in Park Quarry, near Gallow Hill [NO 806 978], where they merge into mounds and ridges of ice-contact gravel. The upper part of the sequence at Park Quarry is described in Brown (1993); nearby boreholes indicate that the workable material reaches a thickness of between 15.8 and 19.7 m.

The discontinuous spreads of glaciofluvial sand and gravel in the valley of the Burn of Sheeoch form flat-topped mounds and sinuous esker ridges trending north-eastwards. The sediments underlying the flat-topped mounds are typified by the coarsening upward deltaic sequence that was formerly exposed in Lochton Pit [NO 752 929]. Up to 5.4 m of sandy gravel was recorded overlying laminated glaciolacustrine silt and clay in the floor of the pit; the uppermost parts of the Lochton sequence are described in Brown (1994). The esker deposits are characterised by interdigitating sequences of poorly sorted gravel, fine-grained sand and laminated silt and clay. They are more heterogeneous and hence less attractive sources of aggregate than some of the deltaic deposits.

The most extensive spreads of sand and gravel in the valley of the Water of Feugh underlie the floodplain and river terraces upstream of Heugh-head [NO 687 928]. Much of this material lies close to or below the water table, but mounds and kettled terraces on the southern side of the river valley, upstream of its confluence with the Burn of Knock, contain thick deposits of potentially workable sand and gravel. These deposits are typified by the glaciofluvial deltaic sediments exposed in the working pit at Cammie Wood [NO 695 920], where 12.8 m of sand and gravel occurs above the water table (Plate 10d). The deposits coarsen upwards, from pebbly sand with partings of silt and clay into cross-bedded, fining-upward graded units of sandy gravel. The sandy gravel is overlain by up to 3 m of subhorizontally bedded coarse gravel that was laid down as the topset beds of the former ice-contact delta. The top of the sequence at Cammie Wood is described further in Brown (1994). Similar upward-coarsening sequences characterise many of the moundy glaciofluvial deposits between Scolly’s Cross [NO 642 877] and Pitdelphin Farm [NO 654 912], and also those forming large kames in the valley of the Devilly Burn, near Clatterin’ Brig [NO 665 782], in the southern part of Sheet 66E.

Small but attractive resources of coarse gravel, lying above the water table are present within esker systems in the vicinity of Strachan. Gravels forming the esker ridges, have been worked in pits north of Waulkmill [NO 647 923] and near Templeton [NO 671 916]; however, significant unworked resources are present, within discontinuous esker ridges between Powlair and Greendams [NO 649 901].

The sands and gravels that form the undulating topography between Auchenblae and Drumlithie, and around Fettercairn, were laid down by meltwaters draining between ice in Strathmore and ice in the upland area to the north, which retreated north-westwards. These deposits characteristically occur as coarsening-upward sequences, which were laid down as deltas and fans into bodies of standing water between the two ice masses. The sands and gravels commonly contain thin waste partings of laminated silt and clay and some workable deposits (notably those south-east of Drumlithie) extend beneath a thin overburden of sandy ‘flow till’. These glaciofluvial sediments are assigned to the Drumlithie Sand and Gravel Formation of the Mearns Drift Group (see Chapter 8). They contain a mixed assemblage of clasts with tough granitic and psammitic pebbles being derived from igneous and metamorphic terrain to the north and less resistant clasts of sandstone and andesitic volcanic rocks derived from the Old Red Sandstone succession in Strathmore.

The most laterally extensive resources lying above the water table in the vicinity of Auchenblae, occur as flat-topped mounds. These deposits were worked in a large pit at Drumsleed [NO 733 776], where 5 m of sandy gravel was formerly exposed, overlying 13 m of pebbly sand. Exposures in other disused workings indicate that some of the nearby flat-topped deposits are much thinner, and range in composition from clayey gravel, to medium- and coarse-grained sand. Farther to the east, the glaciofluvial deposits become more hummocky and form a series of east-trending ridges that are extensively dissected by glacial drainage channels. The workable deposits of sand and gravel within the ridges range from 1.8 m to over 23.1 m in thickness, and have a similar range of grading characteristics to the deposits forming the flat-topped mounds.

The discontinuous spreads of sand and gravel beneath the floodplains of the Bervie Water, Luther Water, Black Burn, Devilly Burn and Burn of Cauldcots are relatively unattractive sources of aggregate as most lie beneath the water table. However, the last two named burns dissect extensive deposits of sand and gravel of the Drumlithie Sand and Gravel Formation that form moundy and terraced spreads around Fettercairn. None of these deposits has been assessed in detail, but notable resources are thought to occur on the lower flanks of Tor Hill, between Thornhill [NO 630 725] and Fettercairn, around Stankeye [NO 643 745], Kincardine Castle [NO 671 751], and at Nether Thainston [NO 634 750].

Sands and gravels lying above the water table are present within ice-contact glaciofluvial sequences in two areas at the eastern edge of Sheet 66E. Significant resources of gravel and coarse-grained sand have been worked in several small pits in the valley of the Cowie Water, around Snob Cott

[NO 801 885]. They also extend downstream on to the adjoining Sheet 67 Stonehaven. The surrounding moundy glacial deposits have also been worked on a small scale, though the resource is commonly thin, bouldery and laterally discontinuous.

Mounds and ridges of ice-contact sand and gravel on the eastern side of the valley drained by the Carron Water, and the Forthie Water between Brenzieshill [NO 799 792] and Pitdrichie [NO 795 825], have been worked for gravel in several places. The deposits forming the Bridge of Fiddes Esker are largely worked out, whereas up to 18.7 m of gravel was recorded in 1989 from active workings in an esker ridge near Pitdrichie. In contrast, the Little Wards Esker, which forms a discontinuous ridge, extending south-south-westwards for a distance of 1.5 km from the vicinity of Bridgend [NO 803 789], had only been exploited in small piecemeal workings at the time of the assessment. Noteworthy resources were also present within the more extensive glaciofluvial spreads, but the deposits are generally thinner and more sandy than those forming the eskers.

The gravels in the northern portion of Sheet 66E are of similar composition to most of those in the southern part of Sheet 76E, to the north. Clasts of granite and felsite (microgranite) predominate (Figure A2.3). Most are derived from the Mount Battock Granite, which forms the high ground between the catchments of the Dee, Bervie Water, Carron Water and Cowie Water. Psammitic clasts become more numerous within the glaciofluvial and alluvial gravels in the Dee valley, downstream of Banchory and significant numbers of pebbles of basic igneous rocks are present in gravels forming south-east-trending eskers in the vicinity of Strachan.

The glaciofluvial deposits of the Drumlithie Sand and Gravel Formation in the southern part of Sheet 66E contain a high proportion of quartzite and andesite pebbles, many of which have been reworked from local Devonian conglomerates. Some granitic clasts are also present, but the gravels contain a much higher proportion of deleterious material than those in the northern part of Sheet 66E. The friable clasts, which are mainly Devonian mudstone and sandstone pebbles, significantly reduce the strength of the aggregate and its suitability for use in concrete and road construction (see aggregate test results below).

End-use suitability data are only available for the sand and gravel resources in the valley of the River Dee downstream of Banchory, the valley of Burn of Sheeoch, and the sandy lacustrine sediments near Lochhead of Leys and Loch of Park. The glaciofluvial deposits, downstream of Banchory appear to be suitable for most end uses and as ‘all in’ concrete aggregate. The resources in the valley of the Burn of Sheeoch are more variable, with the glaciofluvial sheet deposits near Mains of Blairydrine [NO 742 943] being the most attractive as a source of fine aggregate for asphalt, plaster, mortar and concrete. Some of the lacustrine sediments are also suitable for use as asphalt, plaster, mortar and concrete sand, but the deposits are generally rather thin and variable.

Sheet 67 Stonehaven

Limited spreads of sand and gravel occur on Sheet 67 (Map 11), but their proximity to good road links with the expanding Aberdeen market, has led to exploitation of much of the more attractive material. Most of the resources from Stonehaven northwards, including those in the lower reaches of the valleys of the Cowie Water and Carron Water, were evaluated as part of Mineral Assessment Report 148; most of those in the south were included in MAR 149. The deposits in the Dee valley, were included in MAR 146. The sands and gravels that crop out around Inverbervie have not been assessed in detail, but they are described in Carroll (1995b).

Apart from discontinuous terraced spreads in the valleys of the Cowie Water and the River Dee, most of the workable glaciofluvial deposits occur as isolated groups of mounds and ridges on the interfluves. Many of the Lower Devonian conglomerates which crop out to the south of Stonehaven are disaggregated to at least 5 m depth and form resources of cobble and boulder gravel. In places, the exposed conglomerates have been worked for use as fill, but extensive resources remain at the surface and beneath a thin overburden of till, which itself is very gravelly. Only minor discontinuous resources of coarse aggregate are present within the moundy glacial deposits in the northern half of the sheet.

Most of the thickest deposits of the sand and gravel in the valley of the Cowie Water are sterilised by urban development or are worked out. However, significant resources lying above the water table, do occur within fragmentary glaciofluvial terraces upstream of Ury House [NO 859 877] and in the moundy glaciofluvial deposits near Bogheadly [NO 812 888]. Thick, but laterally impersistent, resources of gravel and sand are also present as isolated mounds and flat-topped spreads on both sides of the boundary between the Mearns and East Grampian drift groups, to the north of Stonehaven. Within the Mearns Drift Group, upward-coarsening sequences of sand and fine-grained gravel, up to 12.4 m thick, have been recorded in boreholes on the flanks of the Burn of Ury. Similar deposits have been worked to depths of up to 8 m in a pit near Cantlayhills [NO 881 905]. Boulder gravel forming a glaciofluvial fan at the eastern end of a glacial drainage channel has also been worked in a pit near Logie [NO 886 888], its proximity to the A92 trunk road more than compensating for the large numbers of oversize clasts within the deposit.

Most of the spreads of glaciofluvial sand and gravel on the sides of the valley of the Carron Water contain less coarse aggregate than similar spreads in the valley of the Cowie Water. The former average 7.9 m in thickness and again typically form coarsening-upward sequences, separated by waste partings of reddish brown silt and clay. The deposits generally grade as fine sand and pebbly sand. Some of the larger spreads have been built on and a few of the remainder have been worked, notably at Braehead [NO 873 852] and north of Kirkton of Fetteresso [NO 853 857].

The most extensive spread of glaciofluvial sand and gravel in the southern part of Sheet 67 forms undulating topography between Denhead [NO 865 795] and Temple [NO 852 765]. The sand and gravel, up to 12.8 m thick, typically coarsens upwards and ranges in grade from sand to sandy gravel. Around Temple, at the southern end of the mapped spread, much of the workable sand and gravel appears to be formed of glacially transported conglomeratic material rather than outwash. The change from one type of deposit to the other is difficult to map with certainty. Other sand and gravel deposits of note occur near Dunnottar Mains Criggie [NO 875 838], Lindsayfield [NO 820 844], Briggs of Criggie [NO 843 824], Catterline [NO 879 782], Nether Craighill [NO 807 774], Largie [NO 835 760] and Pitcarry [NO 831 741]; significant resources have been sterilised within the Inverbervie built up area.

Clasts of granitic and psammitic rocks predominate within the terraced and moundy deposits in the valleys of the Dee and the Cowie Water (Figure A2.3). Almost all of the other gravels on Sheet 67 belong to the Drumlithie Sand and Gravel Formation and are of similar composition to the bulk of those in the southern part of Sheet 66E. However, they do contain a significantly larger proportion of nondurable clasts, mainly of mudstone, decomposed sandstone, and andesitic volcanic rocks. In general, the workable deposits within the valleys of the Cowie Water and Dee appear to be suitable for most end uses. The potentially workable deposits within the Drumlithie Sand and Gravel Formation are more variable. For example, the deposits on the flanks of the Burn of Ury and at Cantlayhills appear attractive sources of coarse and fine aggregate, whereas the boulder gravel near Logie is mainly suitable for coarse aggregate for asphalt and concrete. The deposits on the flanks of the Carron Water, are primarily attractive sources of fine sand for asphalt, plastering and mortar.

Aggregate testing results across the district

The strength and durability of aggregates from the district have been measured by a series of mechanical and physical tests. These were conducted in accordance with BS 812 (British Standards Institution, 1975), on the 10 to 14 mm gravel fraction taken from composite samples collected in each area of assessment. The tests included measurement of aggregate impact value (AIV), aggregate crushing value (ACV), ten per cent fines value, relative density (on both an oven-dried and surface dried basis), water absorption and measured concrete drying shrinkage. Aggregate impact value residue (AIVR) and aggregate crushing value residue (ACVR), as defined by Ramsay (1965) and Ramsay et al. (1973, 1974) were also calculated. Inferred concrete drying shrinkage values (which provide a crude estimate of concrete drying shrinkage) were determined, using the relationship between water absorption of gravel and concrete drying shrinkage established by Edwards (1970). The test results are summarised in (Table A2.2) and discussed more fully in each Mineral Assessment Report.

The AIV and water absorption results from the Garmouth assessment are the lowest from any aggregates tested during the assessments in north-east Scotland and suggest that these gravels are particularly durable. In general, the other mechanical and physical tests also show a smaller variation (range) of results than from any other assessment area and suggest that the resources are relatively homogeneous. The Garmouth results are probably representative of the sands and gravels in the lower Spey valley and reflect their high content of psammite and fresh granite clasts.

The results from the Ellon and Peterhead assessment areas are similar, but the AIV and water absorption results are higher than those from the Garmouth area, probably reflecting the higher proportion of nondurable rock types present in the gravels in the eastern parts of the district. The AIV figures are also higher than those obtained by Edwards (1970) from several gravel samples from north-east Scotland. However, they are comparable to values quoted by Edwards for ‘glacial mainly granite’ gravels in the district. Overall, most of the other testing results given in (Table A2.2) also suggest aggregates of poorer strength and much greater water absorption than inferred by Edwards. A probable explanation for these disparities is that Edwards tested processed samples from stockpiles at working pits, whereas most of the samples tested during the assessment studies also included unprocessed material, collected from boreholes and trial pits.

The mechanical and physical testing results from aggregates in the Aberdeen assessment area indicate that they are generally significantly weaker than those from the Ellon and Peterhead areas. The range of values is also more restricted, suggesting relatively homogeneous gravel resources within the Aberdeen assessment area as a whole. Apparent relative density values are comparable to those from all of the other assessments, but the low water absorption values are comparable to those from gravels in the Garmouth area. As the gravels from Garmouth and Aberdeen are both mainly composed of granitic and psammitic clasts, the relative weakness of the Aberdeen material may reflect the derivation of many of the granite clasts from weathered bedrock.

The testing results from the Inverurie–Stonehaven and Strachan–Auchenblae–Catterline assessment areas are similar to those from Aberdeen, though they both have higher water absorption values. These higher values are thought to reflect the presence of highly weathered volcanic clasts and porous clasts of mudstone and sandstone within the gravels of the Drumlithie Sand and Gravel Formation in Strathmore.

Despite the variations recorded across the district the test results indicate that much of the coarse aggregate appears suitable for use in concrete. Nearly all of the measured concrete drying shrinkage values for aggregates from Aberdeen, Inverurie–Stonehaven and Strachan–Auchenblae–Catterline areas are below 0.065 per cent, suggesting that the aggregate is suitable for inclusion in concrete for ‘most applications’ (Building Research Establishment, 1968; Smith and Collis, 1993). The lower AIV figures and comparable water absorption figures from the Garmouth, Peterhead and Ellon assessments indicate that coarse aggregate from these areas is also probably suitable for more specific types of concrete. This conclusion is supported by the inferred concrete drying shrinkage figures, which, although anomalously high when compared with actual measurements (see (Table A2.2)), are generally below 0.085 per cent. Hence, most aggregate from the district is at least suitable for making general purpose concrete.

Appendix 3 Results of shallow geophysical surveys

During surveys of the Cainozoic deposits in north-east Scotland various shallow geophysical techniques have been used to provide data on the nature, thickness and lateral extent of concealed sedimentary units (Chapter 2). Interpretation of electrical conductivity and resistivity measurements, in particular, has also provided insights into the lateral heterogeneity typical of the surface deposits of the district.

Conductivity survey

The methods employed and the results from a conductivity survey in the vicinity of the Houff of Ury [NO 856 889], near Stonehaven, and the resistivity soundings from the Inverurie–Stonehaven sand and gravel assessment area (MAR 148 on (Figure A2.1)), are fully described in Auton et al. (1988) and Auton (1992). Both techniques provided insights into the three dimensional distribution of workable deposits of sand and gravel, augmenting the surface mapping information. For example, interpretation of a contoured plot of ground conductivity values in the Houff of Ury area ((Figure A3.1)a), in conjunction with detailed geological mapping, resulted in a 60 per cent reduction in the estimated extent of sand and gravel, compared with that recorded during the primary geological survey of the area in 1884. It also led to the recognition of minor, but notable amounts of workable sand and gravel ((Figure A3.1)b) concealed beneath thin till overburden.

Resistivity surveys

Forty three resistivity depth soundings were taken at 25 sites during the Inverurie–Stonehaven sand and gravel resource survey. The soundings, together with data obtained from 173 sample points (trial pits, exposures and boreholes) were used to characterise the Quaternary sequences in the area. Eight of the resistivity sites were positioned close to sample points to allow calibration of resistivity values. The calibrated results enhanced extrapolation of the nature and thickness of deposits between sample points, particularly in the major river valleys, where the full thickness of water saturated material was commonly not proved by pitting and shallow drilling.

The soundings also provided data on the typical range of resistivities for each type of Quaternary deposit encountered (Figure A3.2). For example, resistivity values for ‘clayey’ materials, such as till and hummocky glacial deposits, were low, ranging from about 18 to 300 ohm m, whereas values for sand and gravel were high, ranging from 400 to 10 000 ohm m. The range of resistivity values obtained from 44 soundings taken in the adjacent Strachan-Auchenblae–Catterline assessment area (Auton et al., 1990) was somewhat larger. Values of 175 to 700 ohm m were obtained for tills, and 10 000 to 17 000 ohm m for sands and gravels of the East Grampian Drift Group in the Strachan area; values of 150 to 900 ohm m (till) and 300 to 6000 ohm m (sand and gravel) were recorded for deposits of the Mearns Drift Group, between Auchenblae and Catterline.

Differences between the ranges of values obtained from the two assessment areas were probably, in part, a reflection of moisture content, particularly in the permeable, sandy Quaternary strata, at the time of survey. The Inverurie–Stonehaven soundings were taken when the ground was wet, in early spring, whereas the Strachan–Auchenblae–Catterline soundings were made when the ground was dry, at the end of a notably dry summer. The resistivity values for tills of the East Grampian and Mearns drift groups are very similar in the Strachan–Auchenblae–Catterline area, and comparable to many of those recorded from tills in the Inverurie–Stonehaven area. The resistivity values from sands and gravels of the Mearns Drift Group are generally lower than those from the East Grampian Drift Group and probably reflect the finer grained and more silty nature of the former deposits.

The resistivity contrast between permeable sand and gravel, and less permeable deposits (till, glaciolacustrine silt and clay) was evident in both assessment surveys allowing clear subdivision of the Quaternary sequences, based on their geophysical signature. However, the contrast between Quaternary deposits and bedrock was generally less evident and in places accurate rockhead depth was difficult to determine without nearby borehole control. The resistivity values obtained for the Quaternary sediments in north-east Scotland are generally much higher than those in southern Britain (Auton, 1992). This reflects, in part, the crystalline or sandy nature of the resistant bedrock parent material, incorporated in the Quaternary deposits of the district, and also the high proportion of boulders and cobbles in the sediments. In southern England, in particular, much of the fine-grained material in glacigenic sediments is derived from mudstone bedrock and large clasts are normally less numerous.

Ground probing radar traverses

The two ground probing radar (GPR) traverses (Greenwood and Raines, 1994), each about 750 m in length, were undertaken in the Houff of Ury area (sited on (Figure A3.1)a). They provided data on the lateral extent and sedimentary architecture of the Quaternary deposits beyond what could be gleaned from surface mapping, conductivity measurements and the sinking of boreholes and trial pits (Greenwood et al., 1995). Gently inclined reflectors were interpreted as corresponding with large-scale foreset bedding typical of glaciofluvial deltaic deposits within the Drumlithie Sand and Gravel Formation in the area. These deltas form rounded and flat-topped mounds. Organised, quiet, layered patterns of reflectors, are thought to be typical of flat-lying glaciolacustrine deposits (Ury Silts Formation) infilling shallow ice-scoured rock basins. The glaciolacustrine sediments typically comprise silty, fine-grained sand and sandy silt. Multiple diffractions, typical of pebble and cobble gravel, were seen beneath topographic highs. These correspond to gravelly deltaic topset beds; the gently inclined reflectors indicating foreset bedding are developed on the flanks of the topographic highs. Lower amplitude diffraction patterns, indicating sediments containing scattered boulders and cobbles, were seen to be characteristic of till overlying Dalradian metamorphic bedrock.

Four GPR traverses ((Figure A3.3)a) were made across the type area of the Buchan Ridge Gravel Member at Moss of Cruden, during 1994, to image the internal geometry of the deposit and the form of its basal contact with the underlying bedrock. A short traverse was also undertaken along a track through a forestry plantation at Moss of Auquharney [NJ 023 397], some 150 m south of the type area. The resulting profiles (Greenwood and Raines, 1994), together with data from trial pits, boreholes, and resistivity soundings provided important evidence bearing on the origin of the gravel (Chapter 4; Appendix 1 Site 14).

The most informative GPR profile (line 1) is almost coincident with the pitting transect A1–A of Hall and Jarvis (1994, fig.3), across the northern slope of the ridge at Moss of Cruden. Part of the north-west to south-east profile, (162–364 m) is shown in (Figure A3.3)b. The series of shallow-dipping, well-ordered reflections (without diffractions) indicates undisturbed Palaeogene to Neogene gravel (Buchan Ridge Gravel Member), with low-angle cross-stratification dipping towards the axis of the ridge. Farther up-slope (276–342 m), the GPR reflectors are distorted by diffractions that may indicate periglacial and glacitectonic disturbance of the upper part of the gravel. The profile shows the cross-stratified gravel unit thinning north-westwards, confirming that its feather edge rests on weathered Lower Cretaceous sandstone and Caledonian granitic bedrock.

The Cretaceous Moreseat Sandstone, which is largely decomposed to soft sandy silt and silty sand, is characterised by a diminution in radar reflectivity. This was only clearly imaged toward the north-western end of the profile (Greenwood et al.,1995, fig.5, 80–100 m). The resistivity depth probes and radar indicate that around 25 m of Buchan Ridge Gravel lies beneath the crest of the ridge, infilling a channel running parallel to the line of the ridge.

Interpretation of resistivity data from Moss of Cruden has allowed modelling of resistivity values for parts of the sequence encountered in the GPR profile 1 ((Figure A3.3)c). The topsoil is characterised by high resistivity values (2675–5940 ohm m) and the upper parts of the Buchan Ridge Gravel display a range of lower resistivities (447–2534 ohm m). This is thought to be due, in part, to variations in the proportions of kaolinised granitic clasts throughout the unit, the presence of interbeds of clayey pebbly sand (known from the upper part of the gravel sequence in Borehole NK 04 SW3) and the disturbance of bedding, indicated by distorted diffractions in the GPR profile. A layer, about 10 m thick, displaying relatively uniform low resistivities (307–355 ohm m) occurs below the ‘disturbed’ gravel at sounding sites MOSS and M1-285; it reaches the surface at M1-225. Its uniform resistivity response is taken to indicate that this gravel (with its shallow-dipping, well-ordered GPR reflections) is little disturbed and generally contains more kaolinitic material, in the form of decomposed granite cobbles and as fine-grained matrix, than the overlying ‘disturbed’ material. The absence of the latter in sounding M1-225, suggests that the upper ‘disturbed’ units are only well preserved where the gravel is thickest (towards the crest of the ridge); they have probably been removed on the flanks of the ridge by glacial erosion.

The presence of Moreseat Sandstone, infilling a topographic depression in deeply decomposed granite bedrock (between M1-225 and MOSS), is suggested by a 120 ohm m layer at the base of sounding M1-285 and corroborated by the trial pitting results reported by Hall and Jarvis (1994). A value of 120 ohm m is lower than might be expected for sandstone, but is comparable to values of 76–97 ohm m obtained for decomposed Devonian sandy siltstone from the Stonehaven–Auchenblae area (Auton et al., 1990). The very low resistivity (22 ohm m) layer from about 112 to 95 m above OD in sounding M-225, corresponds to grussified granitic bedrock recorded at the bottom of trial pit 15 in Hall and Jarvis (1994, fig.3). Similar (29.5 ohm m) deeply weathered material, some 23 m thick, overlies fresh granite bedrock (3640 ohm m) in sounding MOSS on the ridge crest.

Five GPR traverses were also made (in 1994) across the type area of the Windy Hills Gravel Member (Greenwood et al., 1995). The interpreted GPR data are incorporated in the description of the Windy Hills site given in Appendix 1 (see also Chapter 4).

References

Most of the references listed below are held in the Library of the British Geological Survey, Murchison House, Edinburgh, and at Keyworth, Nottingham. Copies of the references can be purchased subject to current copyright legislation. BGS library catalogue can be searched online at: http://geolib.bgs.ac.uk.

AALBERSBERG, G, and LITT, T. 1998. Multiproxy climate reconstructions for the Eemian and Early Weichselian. Journal of Quaternary Science, Vol. 13, 367–390.

ABER, J S, CROOT, D G, and FENTON, F M. 1989. Glaciotectonic landforms and structures. (Dordrecht, The Netherlands: Kluwer Academic Publishers.)

AITKEN, A M. 1983. A preliminary study of the sand and gravel deposits of Strathmore (1:25 000 sheets NN72, 82, 92 and parts of 71, 81 and 91: sheets NO03, 14, 24, 25, 34, 35, 45, 55, 56, 64, 65, 66, 67, 75, 76, 77, 86, 87, 88 and parts of NO02, 04, 12, 13, 23, 36, 44, 46, 54 and 78). Institute of Geological Sciences, Industrial Mineral Assessment Unit, Open File Report, WF/SC/83/5.

AITKEN, A M, MERRITT, J W, and SHAW, A J. 1979. The sand and gravel resources of the country around Garmouth, Grampian Region. Description of 1:25 000 resource Sheet NJ36. Mineral Assessment Report of the Institute of Geological Sciences, No. 41.

AITKEN, J F. 1991. Sedimentology and palaeoenvironmental significance of late Devensian to mid Holocene deposits in the Don valley, north-east Scotland. Unpublished PhD thesis, University of Aberdeen.

AITKEN, J F. 1993. A re-appraisal of supposed iceberg ‘dump’ and ‘grounding’ structures from Pleistocene glaciolacustrine sediments, Aberdeenshire. Quaternary Newsletter, Vol. 71, 1–10.

AITKEN, J F. 1995. Lithofacies and depositional history of a Late Devensian ice-contact deltaic complex, northeast Scotland. Sedimentary Geology, Vol. 99, 111–130.

AITKEN, J F. 1998. Sedimentology of Late Devensian glaciofluvial outwash in the Don Valley, Grampian Region. Scottish Journal of Geology, Vol. 34, 97–117.

ALLEY, R B, and MACAYEAL, D R. 1994. Ice-rafted debris associated with binge-purge oscillations of the Laurentide Ice Sheet. Palaeoceanography, Vol. 9, 503–511.

ALLEY, R B, and ten others. 1993. Abrupt increase in Greenland snow accumulation at the end of the Younger Dryas event. Nature, Vol. 362, 527–529.

ALVERSON, K, and OLDFIELD, F. 2000. Past global changes and their significance for the future: an introduction. Quaternary Science Reviews, Vol. 19, 3–7.

ANDERSON, J G C. 1943. Sands and gravels of Scotland. Quarter-inch Sheet 9 Elgin–Banff–Aberdeen. Wartime Pamphlet of the Geological Survey of Great Britain, No. 30.

ANDERSON, J G C. 1945. Sands and gravels of Scotland: Quarter-Inch Sheet 12, Stonehaven–Perth–Dundee. Wartime Pamphlet of the Geological Survey of Great Britain, No. 30.

ANDERTON, R, BRIDGES, P H, LEEDER, M R, and SELLWOOD, B W. 1979. A dynamic stratigraphy of the British Isles. (London: Allen and Unwin.)

ANDREWS, I J, and six others. 1990. United Kingdom offshore regional report: the geology of the Moray Firth. (London: HMSO for the British Geological Survey.)

ANSARI, M H. 1992. Stratigraphy and palaeobotany of Middle Pleistocene interglacial deposits in the North Sea. Unpublished PhD thesis, University Wales; Bangor.

ARMSTRONG, M, and PATERSON, I B. 1985. Some recent discoveries of ice-wedge cast networks in north-east Scotland — a comment. Scottish Journal of Geology, Vol. 21, 107–108.

ARMSTRONG, M, PATERSON, I B, and BROWNE, M A E. 1985. Geology of the Perth and Dundee district. Memoir of the British Geological Survey, Sheets 48W, 48E and 49 (Scotland).

ASHCROFT, W A, and WILSON, C D V. 1976. A geophysical survey of the Turriff basin of the Old Red Sandstone, Aberdeenshire. Journal of the Geological Society of London, Vol. 132, 27–43.

ASHCROFT, W A, and MUNRO, M. 1978. The structure of the eastern part of the Insch mafic intrusion, Aberdeenshire. Scottish Journal of Geology, Vol. 14, 55–79.

ASHWORTH, J R. 1975. The sillimanite zones of the Huntly–Portsoy area in the north-east Dalradian, Scotland. Geological Magazine, Vol. 112, 113–136.

ATKINSON, T C, BRIFFA, K R, and COOPE, G R. 1987. Seasonal temperatures in Britain during the last 22 000 years, reconstructed using beetle remains. Nature, London, Vol. 325, 587–593.

AUSTIN, W E N, and KROON, D. 1996. Late glacial sedimentology, foraminifera and stable isotope stratigraphy of the Hebrides continental shelf, northwest Scotland. 187–213 in Late Quaternary Palaeoceanography of the North Atlantic Margins.

ANDREWS, J T, and AUSTIN, W E N, BERGSTEN, H, and JENNINGS, A E (editors). Geological Society, Special Publication, No. 111.

AUTON, C A. 1992. The utility of conductivity surveying and resistivity sounding in evaluating sand and gravel deposits and mapping drift sequences in northeast Scotland. Engineering Geology, Vol. 32, 11–28.

AUTON, C A. 2000. The Buchan Gravels Formation. 79–81 in The Quaternary of the Banffshire coast and Buchan: Field Guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

AUTON, C A, and CROFTS, R G. 1986. The sand and gravel resources of the country around Aberdeen, Grampian Region. Description of 1:25 000 resource sheets NJ71, 80, 81 and 91, with parts of NJ61, 90 and 92, and with parts of NO89 and 99. Mineral Assessment Report, British Geological Survey, No. 146.

AUTON, C A, MERRITT, J W, and ROSS, D L. 1988. The sand and gravel resources of the country around Inverurie and Dunecht, and between Banchory and Stonehaven, Grampian Region. Mineral Assessment Report 148. British Geological Survey Technical Report, WF/88/1.

AUTON, C A, THOMAS, C W, and MERRITT, J W. 1990. The sand and gravel resources of the country around Strachan and between Auchenblae and Catterline, Grampian Region. Mineral Assessment Report 149. British Geological Survey Technical Report, WF/90/7.

AUTON, C A, GORDON, J E, MERRITT, J W, and WALKER, M J C. 2000. The glacial and interstadial sediments at the Burn of Benholm, Kincardineshire: evidence for onshore pre-Devensian ice movement in northeast Scotland. Journal of Quaternary Science, Vol. 15, 141–156.

BAKER, A, SMART, P L, and EDWARDS, R L. 1995. Paleoclimate implications of mass spectrometric dating of a British flowstone. Geology, Vol. 23, 309–312.

BALLANTYNE, C K, and HARRIS, C. 1994. The periglaciation of Great Britain. (Cambridge: Cambridge University Press.)

BALLANTYNE, C K, and five others. 1998. High-resolution reconstruction of the last ice sheet in NW Scotland. Terra Nova, Vol. 10, 63–67.

BARD, E, and five others. 1987. Retreat velocity of the North Atlantic polar front during the last deglaciation determined by 14C accelerator mass spectrometry. Nature, London, Vol. 328, 791–794.

BARNE, J H, ROBSON, C F, KAZNAKOVA, S S, DOODY, J P, and DAVIDSON, N C (editors). 1996. Coasts and seas of the United Kingdom: North-east Scotland: Cape Wrath to St Cyrus. (Peterborough: Nature Conservation Committee.)

BASHAM, I R. 1968. Deeply weathered rock and associated soils of the Insch and Boganclough masses, Aberdeen. Unpublished PhD thesis, University of Aberdeen.

BASHAM, I R. 1974. Mineralogical changes associated with deep weathering of gabbro in Aberdeenshire. Clay Minerals, Vol. 10, 189–202.

BAUMANN, K H, and six others. 1995. Reflection of Scandinavian ice sheet fluctuations in Norwegian sea sediments during the past 150 years. Quaternary Research, Vol. 43, 185–197.

BEHRE, K E. 1989. Biostratigraphy of the last glacial period in Europe. Quaternary Science Reviews, Vol. 8, 25–44.

BELL, D. 1895. The shelly clays and gravels of Aberdeenshire considered in relation to the question of submergence. Quaterly Journal of the Geological Society of London, Vol. 51, 472–479.

BENN, D I, and EVANS, D J A. 1996. The interpretation and classification of subglacially-deformed sediments. Quaternary Science Reviews, Vol. 15, 23–52.

BENN, D I, and EVANS, D J A. 1998. Glaciers and glaciation. (London: Arnold.)

BENN, D I, LOWE, J J, and WALKER, M J C. 1992. Glacier response to climatic change during the Loch Lomond Stadial and early Flandrian: geomorphological and palynological evidence from the Isle of Skye, Scotland. Journal of Quaternary Science, Vol. 7, 125–144.

BENT, A J A. 1986. Aspects of Pleistocene glaciomarine sequences in the northern North Sea. Unpublished PhD thesis, University of Edinburgh.

BERSTAD, S, and DYPVIK, H. 1982. Sedimentological evolution and natural radioactivity of Tertiary sediments from the central North Sea. Journal of Sedimentary Petrology, Vol. 5, 77–88.

BIRKS, H J B. 1977. The Flandrian forest history of Scotland: a preliminary synthesis. 119–135 in British Quaternary studies: recent advances. SHOTTON, F W (editor). (Oxford: Clarendon Press.)

BIRKS, H J B, and RANSOM, M E. 1969. An interglacial peat at Fugla Ness, Scotland.

New Phytologist, Vol. 68, 777–796.

BISHOP, W W, and COOPE, G R. 1977. Stratigraphical and faunal evidence for Lateglacial and early Flandrian environments in south-west Scotland. 61–88 in Studies in the Scottish Lateglacial. GRAY, J M, and LOWE, J J (editors). (Oxford and New York: Pergamon.)

BOND, G, and LOTTI, R. 1995. Iceberg discharges into the North Atlantic on millennial time scales during the last glaciation. Science, Vol. 267, 1005–1010.

BOND, G, and six others. 1993. Correlation between climate records from North Atlantic sediments and Greenland ice. Nature, Vol. 365, 143–147.

BOULTON, G S. 1968. Flow tills and related deposits on some Vestspitsbergen glaciers. Journal of Glaciology, Vol. 7, No. 51, 391–412.

BOULTON, G S. 1970. On the deposition of subglacial and melt-out tills at the margins of certain Svalbard glaciers. Journal of Glaciology, Vol. 9, 231–245.

BOULTON, G S. 1979. Processes of glacier erosion on different substrata.

Journal of Glaciology, Vol. 23, 15–38.

BOULTON, G S. 1987. A theory of drumlin formation by subglacial deformation. 25–80 in Drumlin Symposium. MENZIES, J, and ROSE, J (editors). (Rotterdam: Balkema.)

BOULTON, G S, and DEYNOUX, M. 1981. Sedimentation in glacial environments and the identification of tills and tillites in ancient sedimentary sequences. Precambrian Research, Vol. 15, 397–422.

BOULTON, G S, and DOBBIE, K E. 1993. Consolidation of sediments by glaciers: relations between sediment geotechnics, soft-bed glacier dynamics and subglacial groundwater flow. Journal of Glaciology, Vol. 39, 26–44.

BOULTON, G S, and PAUL, M A. 1976. The influence of genetic processes on some geotechnical properties of glacial tills. Quarterly Journal of Engineering Geology, Vol. 9, 159–193.

BOULTON, G S, DENT, D L, and MORRIS, E M. 1974. Subglacial shearing and crushing, and the role of water pressure in tills from SE Iceland. Geografiska Annaler, Series A Physical Geography, Vol. 56A, No. 314, 135–145.

BOULTON, G S, JONES, A S, CLAYTON, K M, and KENNING, M J. 1977. A British ice-sheet model and patterns of glacial erosion and deposition in Britain 231–246 in British Quaternary studies: recent advances. SHOTTON, F W (editor). (Oxford: Claredon Press.)

BOULTON, G S, SMITH, G D, JONES, A S, and NEWSOME, J. 1985. Glacial geology and glaciology of the last mid-latitude ice sheets. Journal of the Geological Society of London, Vol. 142, 447–474.

BOULTON, G S, PEACOCK, J D, and SUTHERLAND, D G. 1991. Quaternary. 503–543 in Geology of Scotland. (Third edition). CRAIG, G Y (editor). (London: The Geological Society.)

BOWEN, D Q. 1989. The last interglacial-glacial cycle in the British Isles. Quaternary International, Vol. 3, Pt 4, 41–47.

BOWEN, D Q, (editor). 1999. A revised correlation of Quaternary deposits in the British Isles. Geological Society Special Report, No. 23.

BOWEN, D Q, ROSE, J, MCCABE, A M, and SUTHERLAND, D G. 1986. Correlation of Quaternary glaciations in England, Ireland, Scotland and Wales. 299–340 in Quaternary Glaciations in the Northern Hemisphere. SIBRAVA, V, BOWEN, D Q, and RICHMOND, G M (editors). Quaternary Science Reviews, Vol. 5.

BOWEN, D Q, and SYKES, G A. 1988. Correlation of marine events and glaciations on the northeast Atlantic margin. Philosophical Transactions of the Royal Society of London, Vol. B, 318, 619–635.

BRADY, G S, CROSSKEY, H W, and ROBERSTON, D. 1874. A monograph of the post-Tertiary Entomostraca of Scotland. Palaeontographical Society Monograph, 1–232.

BREMNER, A. 1912. The physical geology of the Dee valley. (Aberdeen: The University Press.)

BREMNER, A. 1915. The capture of the Geldie by the Feshie. Scottish Geographical Magazine, Vol. 31, 589–596.

BREMNER, A. 1916. Problems in the glacial geology of northeast Scotland and some fresh facts bearing on them. Transactions of the Edinburgh Geological Society, Vol. 10, 334–347.

BREMNER, A. 1917. Low-level kettle-holes in and near Aberdeen. Transactions of the Edinburgh Geological Society, Vol. 11, 23–24.

BREMNER, A. 1919. A geographical study of the high plateau of the south-eastern Highlands. Scottish Geographical Magazine, Vol. 35, 331–351.

BREMNER, A. 1920a. The glacial geology of the Stonehaven District. Transactions of the Edinburgh Geological Society, Vol. 11, 25–41.

BREMNER, A. 1920b. Limits of valley glaciation in the basin of the Dee. Transactions of the Edinburgh Geological Society, Vol. 11, 61–68.

BREMNER, A. 1921. The physical geology of the Don basin. (Aberdeen: The University Press.)

BREMNER, A. 1922. Deeside as a field for the study of geology. The Deeside Field, Vol. 1, 33–36.

BREMNER, A. 1928. Further problems in the glacial geology of north-eastern Scotland. Transactions of the Edinburgh Geological Society, Vol. 12, 147–164.

BREMNER, A. 1931. The valley glaciation in the district round Dinnet, Cambus o’May and Ballater. The Deeside Field, Vol. 5, 15–24.

BREMNER, A. 1934a. The glaciation of Moray and ice movements in the north of Scotland. Transactions of the Edinburgh Geological Society, Vol. 13, 17–56.

BREMNER, A. 1934b. Meltwater drainage channels and other glacial phenomena of the Highland Border Belt from Cortachy to the Bervie Water. Transactions of the Edinburgh Geological Society, Vol. 13, 174–175.

BREMNER, A. 1936. The glaciation of Glenesk. Transactions of the Edinburgh Geological Society, Vol. 13, 378–382.

BREMNER, A. 1938. The glacial epoch in north-east Scotland. The Deeside Field, Vol. 8, 64–68.

BREMNER, A. 1939. Notes on the glacial geology of east Aberdeenshire. Transactions of the Edinburgh Geological Society, Vol. 13, 474–475.

BREMNER, A. 1942. The origins of the Scottish river system. Scottish Geographical Magazine, Vol. 58, 15–20, 54–59, 99–103.

BREMNER, A. 1943. The glacial epoch in the north-east. 10–30 in The Book of Buchan (Jubilee Volume). TOCHER, J F (editor). (Aberdeen: P Scrogie Ltd & Aberdeen University Press.)

BRIDGLAND, D R. 2000a. Flint-rich gravels in Aberdeenshire. 96–101 in The Quaternary of the Banffshire coast and Buchan: Field Guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

BRIDGLAND, D R. 2000b. Discussion: the characteristics, variations and likely origin of the Buchan Ridge Gravel. 139–143 in The Quaternary of the Banffshire coast and Buchan: Field Guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

BRIDGLAND, D R, and SAVILLE, A. 2000. Den of Boddam. 102–115 in The Quaternary of the Banffshire coast and Buchan: Field Guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

BRIDGLAND, D R, SAVILLE, A, and SINCLAIR, J M. 1997. New evidence for the origin of the Buchan Ridge Gravel, Aberdeenshire. Scottish Journal of Geology, Vol. 33, 43–50.

BRIDGLAND, D R, SAVILLE, A, and SINCLAIR, J. 2000. Skelmuir Hill. 126–138 in The Quaternary of the Banffshire coast and Buchan: Field Guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

BRITISH GEOLOGICAL SURVEY. 1977. Moray–Buchan Sheet 57°N–04°W. Solid geology. 1:250 000. (Southampton: Ordnance Survey for British Geological Survey.)

BRITISH GEOLOGICAL SURVEY. 1992. Peterhead, Scotland Sheet 87E. Solid and Drift. 1:50 000. (Keyworth: British Geological Survey.)

BRITISH GEOLOGICAL SURVEY. 2002. Lexicon of named rocks units [online]. Keyworth, Nottingham: British Geological Survey. Last revised 05/22/2002. Available from http://www.bgs.ac.uk.

BRITISH GEOLOGICAL SURVEY. 1998. United Kingdom Minerals Yearbook 1997. (Keyworth, Nottingham: British Geological Survey.)

BRITISH STANDARDS INSTITUTION. 1975. BS 812: Methods for sampling and testing of mineral aggregates, sands and fillers. (London: British Standards Institution.)

BROECKER, W S. 1984. Terminations. 687–698 in Milankovitch and Climate. Part 2. BERGER, A, IMBRIE, J, HAYS, J, KUKLA, G, and SALTZMAN, B (editors). (Norwell, Mass: D Reidel.)

BROECKER, W S. 1994. Massive iceberg discharges as triggers for global climate change. Nature, Vol. 372, 421–424.

BROECKER, W S, and DENTON, G H. 1990. What drives glacial cycles? Scientific American, Vol. 262, 43–50.

BROWN, A G. 1987. Long-term sediment storage in the Severn and Wye catchments. 307–332 in Palaeohydrology in practice. GREGORY, K J, LEWIN, J, and THORNES, J B (editors). (Chichester: Wiley.)

BROWN, I M. 1993. Pattern of deglaciation of the last (Late Devensian) Scottish ice sheet: evidence from ice-marginal deposits in the Dee valley, north-east Scotland. Journal of Quaternary Science, Vol. 8, 235–250.

BROWN, I M. 1994. Former glacial lakes in the Dee valley: origin, drainage and significance. Scottish Journal of Geology, Vol. 30, 147–158.

BUCHAN, A. 1935. Investigations of the glacial and post-glacial deposits of Loch Spynie. Unpublished PhD thesis, University of Aberdeen.

BUCHARDT, B. 1978. Oxygen isotope palaeotemperatures from the Tertiary period in the North Sea area. Nature, London, Vol. 275, 121–123.

BUILDING RESEARCH ESTABLISHMENT. 1968. Shrinkage of natural aggregates in concrete. Digest Building Research Establishment, No. 35, Series 2.

CABRERA, D E. 1976. A study of the soils and their underlying deposits at Cruden Bay, Grampian. Unpublished MSc thesis, University of Aberdeen.

CAMBRIDGE, P G. 1982. A note on supposed crag shells from the Kippet Hills, Aberdeenshire. Bulletin of the Geological Society of Norfolk, Vol. 32, 37–38.

CAMERON, T D J, STOKER, M S, and LONG, D. 1987. The history of Quaternary sedimentation in the UK sector of the North Sea Basin. Journal of the Geological Society, London, Vol. 144, Pt 1, 43–58.

CAMERON, T D J, and six others. 1992. United Kingdom Offshore Regional Report: the Geology of the Southern North Sea. (London: HMSO for the British Geological Survey, London.)

CAMERON, D G, and seven others. 1998. Directory of Mines and Quarries, 1998. (5th edition). (Keyworth, Nottingham: British Geological Survey.)

CAMPBELL, R. 1934. On the occurrence of shelly boulder clay and interglacial deposits in Kincardineshire. Transactions of the Edinburgh Geological Society, Vol. 13, 176–182.

CARR, S. 1998. Last glacial maximum in the North Sea Basin. Unpublished PhD thesis, Royal Holloway, University of London.

CARR, S. 1999. The micromorphology of Last Glacial Maximum sediments in the Southern North Sea. Catena, Vol. 35, 123–145.

CARROLL, S. 1995a. Geology of the Stonehaven district. 1:10 000 sheets NO88NW, NO88NE, NO88SW and NO88SE (South of the Highland Boundary Fault). British Geological Survey Technical Report, WA/94/19.

CARROLL, S. 1995b. Geology of the Inverbervie and Catterline District. 1:10 000 sheets NO87NW, NO87NE and NO87SW. British Geological Survey Technical Report, WA/94/20.

CARROLL, S. 1995c. Geology of the Laurencekirk District. 1:10 000 sheets NO77SW and NO77SE. British Geological Survey Technical Report, WA/95/81.

CARROLL, S. 1995d. Geology of the Fettercairn District. 1:10 000 sheets NO67NW, NO67NE, NO67SW, NO67SE and NO68SE (South of the Highland Boundary Fault). British Geological Survey Technical Report, WA/95/91.

CASELDINE, C J, and EDWARDS, K J. 1982. Interstadial and last interglacial deposits covered by till in Scotland: comments and new evidence. Boreas, Vol. 11, 119–123.

CHAMBERS, R. 1857. Geological notes on Banffshire. Proceedings of the Royal Society of Edinburgh, Vol. 3, 332–333.

CHAPMAN, M R, SHACKLETON, N J, and DUPLESSY, J-C. 2000. Sea surface temperature variabillity during the last glacial–interglacial cycle: assessing the magnitude and pattern of climate change in the North Atlantic. Palaeogeography, Palaeoclimatology, Palaeoecology, Vol. 157, 1–25.

CHARLESWORTH, J K. 1926. The re-advance, marginal kame-moraine of the south of Scotland, and some later stages of retreat. Transactions of the Royal Society of Edinburgh, Vol. 55, 25–50.

CHARLESWORTH, J K. 1956. The Late-glacial history of the Highlands and Islands of Scotland. Transactions of the Royal Society of Edinburgh, Vol. 62, 769–928.

CLAPPERTON, C M. 1977. Buchan. 37–39 in INQUA X Congress 1977. Guidebook for Excursions A10 and C10. The Northern Highlands of Scotland. CLAPPERTON, C M (editor). (Norwich: Geo Abstracts.)

CLAPPERTON, C M. 1997. Greenland ice cores and North Atlantic sediments: implications for the last glaciation in Scotland. 45–58 in Reflections on the ice age in Scotland. GORDON, J E (editor). (Glasgow: Scottish Association of Geography Teachers and Scottish Natural Heritage.)

CLAPPERTON, C M, and GEMMELL, A M D. 1998. Windy Hills SSSI, Aberdeenshire: site survey and management phase 1. Scottish Natural Heritage Commissioned Report, F98LF10 (Unpublished report).

CLAPPERTON, C M, and SUGDEN, D E. 1975. The glaciation of Buchan —a reappraisal. 19–22 in Quaternary studies in north east Scotland. A M D GEMMELL (editor). (Aberdeen: Department of Geography, University of Aberdeen.)

CLAPPERTON, C M, and SUGDEN, D E. 1977. The Late Devensian glaciation of north-east Scotland. 1–13 in Studies in the Scottish Lateglacial environment. GRAY, J M, and LOWE, J J (editors). (Oxford: Pergamon Press.)

CLAYTON, K M. 1974. Zones of glacial erosion. Institute of British Geographers Special Publication, Vol. 7, 163–176.

CLIMAP PROJECT MEMBERS. 1976. The surface of the ice age Earth. Science, Vol. 191, 1131–1136.

CONNELL, E R. 1984a. Kirkhill Quarry. Site. 59 in Buchan field guide. HALL, A M (editor). (Cambridge: Quaternary Research Association.)

CONNELL, E R. 1984b. Kirkhill Quarry. Deposits between the lower and upper palaeosols. 62–64 in Buchan field guide. HALL, A M (editor). (Cambridge: Quaternary Research Association.)

CONNELL, E R. 1984c. Kirkhill Quarry. Deposits above the upper palaeosol. 64–69 in Buchan field guide. HALL, A M (editor). (Cambridge: Quaternary Research Association.)

CONNELL, E R. 1984d. Sandhole (Denhead) Sand Pit. 83–86 in Buchan field guide. HALL, A M (editor). (Cambridge: Quaternary Research Association).

CONNELL, E R. 2000a. Tofthead, Fochabers. 38–39 in The Quaternary of the Banffshire coast and Buchan: field guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

CONNELL, E R. 2000b. Sandford Bay, Peterhead. 57–61 in The Quaternary of the Banffshire coast and Buchan: field guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

CONNELL, E R, and HALL, A M. 1984a. Kirkhill Quarry. Deposits beneath the lower palaeosol. 60–62 in Buchan field guide. HALL, A M (editor). (Cambridge: Quaternary Research Association).

CONNELL, E R, and HALL, A M. 1984b. Boyne Bay Quarry. 97–101 in Buchan field guide. HALL, A M (editor). (Cambridge: Quaternary Research Association).

CONNELL, E R, and HALL, A M. 1984c. Sandford Bay. 106–107 in Buchan field guide. HALL, A M (editor). (Cambridge: Quaternary Research Association).

CONNELL, E R, and HALL, A M. 1987. The periglacial stratigraphy of Buchan. 277–285 in Periglacial processes and landforms in Britain and Ireland. BOARDMAN, J (editor). (Cambridge: Cambridge University Press).

CONNELL, E R, and ROMANS, J C C. 1984. Kirkhill Quarry. Palaeosols. 70–76 in Buchan field guide. HALL, A M (editor). (Cambridge: Quaternary Research Association.)

CONNELL, E R, EDWARDS, K J, and HALL, A M. 1982. Evidence for two pre-Flandrian palaeosols in Buchan, Scotland. Nature, London, Vol. 297, 570–572.

CONNELL, E R, HALL, A M, SHAW, D, and RILEY, L A. 1985. Palynology and significance of radiocarbon dated organic materials from Cruden Bay Brick Pit, Grampian Region, Scotland. Quaternary Newsletter, Vol. 47, 19–25.

CRAIG, G Y (editor). 1991. Geology of Scotland. (Third edition). (London: The Geological Society.)

CRAMPTON, C B, and CARRUTHERS, R G. 1914. The geology of Caithness. Memoir of the Geological Survey of Great Britain, Sheets 110, 116 with parts of 109, 115 and 117 (Scotland).

CROFTS, R S. 1975. Sea bed topography off north east Scotland. Scottish Geographical Magazine, Vol. 91, 52–64.

CULLINGFORD, R A, and SMITH, D E. 1980. Late Devensian raised shorelines in Angus and Kincardineshire, Scotland. Boreas, Vol. 9, 21–38.

CULLINGFORD, R A, SMITH, D E, and FIRTH, C R. 1991. The altitude and age of the Main Postglacial Shoreline in eastern Scotland. Quaternary International, Vol. 9, 39–52.

CUMMING, G A, and BATE, P A. 1933. The Lower Cretaceous erratics of the Fraserburgh district, Aberdeenshire. Geological Magazine, Vol. 70, 397–413.

DALE, B. 1985. Dinoflagellate cyst analysis of Upper Quaternary sediments in core GIK 15530–4 from the Skagerrak. Norsk Geologisk Tidsskrift, Vol. 65, 97–102.

DANSGAARD, W, and ten others. 1993. Evidence for general instability of past climate from a 250-kyr ice-core record. Nature, Vol. 364, 218–220.

DAWSON, A G, LONG, D, and SMITH D E. 1988. The Storegga Slides: evidence from eastern Scotland of a possible tsunami. Marine Geology, Vol. 82, 271–276.

DEPARTMENT OF AGRICULTURE AND FISHERIES FOR SCOTLAND. 1962. Scottish Peat. (Second Report of the Scottish Peat Committee). (Edinburgh: HMSO.)

DEPARTMENT OF AGRICULTURE AND FISHERIES FOR SCOTLAND. 1964. Scottish Peat Surveys. (South West Scotland). (Edinburgh: HMSO.)

DEPARTMENT OF AGRICULTURE AND FISHERIES FOR SCOTLAND. 1965a. Scottish Peat Surveys. (Central Scotland). (Edinburgh: HMSO.)

DEPARTMENT OF AGRICULTURE AND FISHERIES FOR SCOTLAND. 1965b. Scottish Peat Surveys. (Western Highlands and Islands). (Edinburgh: HMSO.)

DEPARTMENT OF AGRICULTURE AND FISHERIES FOR SCOTLAND. 1968. Scottish Peat Surveys. (Caithness, Shetland and Orkney). (Edinburgh: HMSO.)

DONNER, J J. 1957. The geology and vegetation of Late-glacial retreat stages in Scotland. Transactions of the Royal Society of Edinburgh, Vol. 63, 221–264.

DONNER, J J. 1960. Pollen analysis of the Burn of Benholm peat bed, Kincardineshire. Societas Scientiarum Fennica, Commentationes Biologicae, Vol. 22, 1–13.

DONNER, J J. 1963. The Late- and Post-Glacial raised beaches in Scotland. Annales Academiae Scientiarum Fennicae, Vol. 68, Pt II, Series A III Geologica-Geographica, 1–13.

DONNER, J J. 1979. The Early or Middle Devensian peat at Burn of Benholm, Kincardineshire. Scottish Journal of Geology, Vol. 15, 247–250.

DOWDESWELL, J A, ELVERHØI, A, ANDREWS, J T, and HEBBELN, D. 1999. Asynchronous deposition of ice-rafted layers in the Nordic seas and North Atlantic Ocean. Nature, London, Vol. 400, 348–351.

DREIMANIS, A, and VAGNERS, U J. 1971. The effect of lithology on the texture of tills. 66–82 in Research Methods in Pleistocene Geomorphology. YATSU, and FALCONER (editors). (Norwich: Geo Abstracts.)

DROOP, G T R, and CHARNLEY, N. 1985. Comparative geobarometry of pelitic hornfelses associated with the newer gabbros: a preliminary study. Journal of the Geological Society of London, Vol. 142, 53–62.

DRYBURGH, P M. 1978. Scotland’s peat resources: an introduction to their potential. (Edinburgh: School of Engineering, University of Edinburgh.)

DULLER, G A T, WINTLE, A F, and HALL, A M. 1995. Luminescence dating and its application to key pre-Late Devensian sites in Scotland. Quaternary Science Reviews, Vol. 14, 495–519.

DURNO, S E. 1956. Pollen analysis of peat deposits in Scotland. Scottish Geographical Magazine, Vol. 72, 177–187.

DURNO, S E. 1957. Certain aspects of vegetational history in north-east Scotland. Scottish Geographical Magazine, Vol. 73, 176–184.

DURNO, S E. 1970. Pollen diagrams from three buried peats in the Aberdeen area. Transactions of the Botanical Society of Edinburgh, Vol. 41, 43–50.

EDMOND, J M, and GRAHAM, J D. 1977. Peterhead power station cooling water intake tunnel: an engineering case study. Quarterly Journal of Engineering Geology, Vol. 10, 281–301.

EDWARDS, A G. 1970. Scottish aggregates: their suitability for concrete with regard to rock constituents. Current Paper of the Building Research Station, No. 28/70.

EDWARDS, K J. 1978. Palaeoenvironmental and archaeological investigations in the Howe of Cromar, Grampian Region, Scotland. Unpublished PhD thesis, University of Aberdeen.

EDWARDS, K J. 1979. Environmental impact in the prehistoric period. 27–42 in Early Man in the Scottish Landscape. Scottish Archaeological Forum. THOMS, L M (editor). Vol. 9. (Edinburgh: Edinburgh University Press.)

EDWARDS, K J, and CONNELL, E R. 1981. Interglacial and interstadial sites in north-east Scotland. Quaternary Newsletter, Vol. 33, 22–28.

EDWARDS, K J, and ROWNTREE, K M. 1980. Radiocarbon and palaeoenvironmental evidence for changing rates of erosion at a Flandrian stage site in Scotland. 207–223 in Timescales in Geomorphology. CULLINGFORD, R A, DAVIDSON, D A, and LEWIN, J (editors). (Chichester and New York: Wiley.)

EDWARDS, K J, CASELDINE, C J, and CHESTER, D K. 1976. Possible interstadial and interglacial pollen floras from Teindland, Scotland. Nature, Vol. 264, 742–744.

EHLERS, J. 1988. Skandinavische Geschiebe in Großbritannien. Der Geschiebesammler, Vol. 22, 49–64.

EHLER, J, and WINGFIELD, R. 1991. The extension of the Late Weichselian/Late Devensian ice sheets in the North Sea Basin. Journal of Quaternary Science, Vol. 6, 313–326.

EHLERS, J, GIBBARD, P L, and ROSE, J. 1991. Glacial deposits of Britain and Europe: general overview. 493–501 in Glacial deposits in Great Britain and Ireland. EHLERS, J, GIBBARD, P L, and ROSE, J (editors). (Rotterdam: Balkema.)

EMILIANI, C. 1954. Depth habitats of some species of pelagic foraminifera as indicated by oxygen isotope ratios. American Journal of Science, Vol. 252, 149.

ENGLAND, J. 1986. Glacial erosion of a high Arctic valley. Journal of Glaciology, Vol. 32, 60–64.

EVANS, D, CHESHER, J A, DEEGAN, C E, and FANNIN, N G T. 1981. The offshore geology of Scotland in relation to the IGS shallow drilling programme, 1970–1978. Report of the Institute of Geological Sciences, No. 81/12.

EVANS, D, MORTON, A C, WILSON, S, JOLLEY, D, and BARREIRO, B A. 1997. Palaeoenvironmental significance of marine and terrestrial Tertiary sediments on the NW Scottish Shelf in BGS borehole 77/7. Scottish Journal of Geology, Vol. 33, 31–42.

EYLES, N, MCCABE, A M, and BOWEN, D Q. 1994. The stratigraphic and sedimentological significance of Late Devensian Ice Sheet surging in Holderness, Yorkshire, UK. Quaternary Science Reviews, Vol. 13, 727–759.

EYLES, V A, ANDERSON, J G C, BONNELL, D G R, and BUTTERWORTH, A R I C. 1946. Brick clays of north-east Scotland. Part I —Description of occurrences; Part II — Report on analyses and physical tests. Geological Survey of Great Britain: Wartime Pamphlet, No. 47.

FIRTH, C R, and HAGGART, B A. 1989. Loch Lomond Stadial and Flandrian shorelines in the inner Moray Firth area, Scotland. Journal of Quaternary Science, Vol. 4, 37–50.

FITZPATRICK , E A. 1956. An indurated soil horizon formed by permafrost. Soil Science, Vol. 7, 248–254.

FITZPATRICK, E A. 1958. An introduction to the periglacial geomorphology of Scotland. Scottish Geographical Magazine, Vol. 74, 28–36.

FITZPATRICK, E A. 1963. Deeply weathered rock in Scotland, its occurrence, age and contribution to the soils. Journal of Soil Science, Vol. 14, 33–43.

FITZPATRICK, E A. 1965. An interglacial soil at Teindland, Morayshire. Nature, Vol. 207, 621–622.

FITZPATRICK, E A. 1969. Some aspects of soil evolution in north-east Scotland. Soil Science, Vol. 107, 403–408.

FITZPATRICK, E A. 1972. The principal Tertiary and Pleistocene events in north-east Scotland. 1–4 in North-East Scotland Geographical Essays. CLAPPERTON, C M (editor). (Aberdeen: Department of Geography, University of Aberdeen.)

FITZPATRICK, E A. 1975a. Errollston Farm. 19–21 in Aberdeen Field Excursion Guide. (Informal Publication: Quaternary Research Association.)

FITZPATRICK, E A. 1975b. Particle size distribution and stone orientation patterns in some soils of north-east Scotland. 49–60 in Quaternary Studies in North East Scotland. GEMMELL, A M D (editor). (Aberdeen: Department of Geography, University of Aberdeen.)

FITZPATRICK, E A. 1975c. Windy Hills. 23–25 in Aberdeen Field Excursion Guide. (Informal Publication: Quaternary Research Association.)

FITZPATRICK, E A. 1987. Periglacial features in the soils of north east Scotland. 153–162 in Periglacial processes and landforms in Britain and Ireland. BOARDMAN, J (editor). (Cambridge: Cambridge University Press.)

FLETCHER, T P, and five others. 1996. Geology of the Fortrose and eastern Inverness district. Memoir of the British Geological Survey, Sheet 84W (Scotland).

FLETT, J S, and READ, H H. 1921. Tertiary gravels of the Buchan district of Aberdeenshire. Geological Magazine, Vol. 58, 215–225.

FRASER, G K. 1943. Peat deposits of Scotland. Part I — General account. Wartime Pamphlet of the Geological Survey of Great Britain: Scotland, No. 36.

FRASER, G K. 1948. Peat deposits of Scotland. Part II — Peat mosses of Aberdeenshire, Banffshire and Morayshire (area of quarter-inch sheet 9). Wartime Pamphlet of the Geological Survey of Great Britain: Scotland, No. 36.

FRIIS, H. 1976. Weathering of a Neogene fluviatile fining-upwards sequence at Voervadsbro, Denmark. Bulletin of the Geological Society of Denmark, Vol. 25, 99–105.

FRONVAL, T, JANSEN, E, BLOEMENDAL, J, and JOHNSEN, S. 1995. Oceanic evidence for coherent fluctuations in Fennoscandian and Laurentide ice sheets on millennium timescales. Nature, Vol. 374, 443–446.

FUNNELL, B M. 1995. Global sea-level and the (pen-)insularity of late Cenozoic Britain. 3–13 in Island Britain: a Quaternary Perspective. PREECE, R C (editor). Geological Society of London, Special Publication, No. 96.

GALLOWAY, R W. 1958. Periglacial phenomena in Scotland. Unpublished PhD thesis, University of Edinburgh.

GALLOWAY, R W. 1961a. Solifluction in Scotland. Scottish Geographical Magazine, Vol. 77, 75–87.

GALLOWAY, R W. 1961b. Periglacial phenomena in Scotland. Geografiska Annaler, Vol. 43, 348–353.

GALLOWAY, R W. 1961c. Ice wedges and involutions in Scotland. Biuletyn Periglacjalny, Vol. 10, 169–193.

GATLIFF, R W, and eleven others. 1994. United Kingdom offshore regional report: the geology of the central North Sea. (London: HMSO for the British Geological Survey.)

GEIKIE, A. 1878. On the Old Red Sandstone of western Europe. Transactions of the Royal Society of Edinburgh, Vol. 28, 345–452.

GEMMELL, A M D. 1975. The Kippet Hills. 14–19 in Aberdeen Field Excursion Guide. (Informal Publication: Quaternary Research Association.)

GEMMELL, A M D, and AUTON, C A. 2000. Windy Hills. 81–92 in The Quaternary of the Banffshire coast and Buchan: field guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

GEMMELL, A M D, and KESEL, R H. 1979. Developments in the study of the Buchan flint deposits. 66–77 in Early Man in the Scottish landscape. THOMS, L M (editor). (Edinburgh: Edinburgh University Press.)

GEMMELL, A M D, and KESEL, R H. 1982. Letters to the editors. The ‘Pliocene’ Gravels of Buchan: a reappraisal: reply. Scottish Journal of Geology, Vol. 18, 333–335.

GEMMELL, A M D, and RALSTON, I B M. 1984. Some recent discoveries of ice-wedge cast networks in north-east Scotland. Scottish Journal of Geology, Vol. 20, 115–118.

GEMMELL, A M D, and RALSTON, I B M. 1985. Ice wedge polygons in north-east Scotland: a reply. Scottish Journal of Geology, Vol. 21, 109–111.

GEMMELL, A M D, and STOVE, G C. 1999. Windy Hills SSSI, Aberdeenshire: site survey and management phase 2. Scottish Natural Heritage Commissioned Report, F98LF11 (Unpublished report).

GEMMELL, S, HANSOM, J, and HOEY, T B. 1996. Geomorphology conservation and management of the River Spey and Spey Bay SSSIs, Moray. Scottish Natural Heritage Commissioned Report.

GLASSER, N F, and HALL, A M. 1997. Calculating Quaternary erosion rates in North East Scotland. Geomorphology, Vol. 20, 29–48.

GLENTWORTH, R. 1954. The soils of the country around Banff, Huntly and Turriff (Sheets 86 and 96). (Memoir of the Soil Survey of Great Britain: Scotland). ( Edinburgh: Her Majesty’s Stationery Office for Department of Agriculture and Fisheries for Scotland.)

GLENTWORTH, R, and MUIR, J W. 1963. The soils of the country around Aberdeen, Inverurie and Fraserburgh. Sheets 77, 76 and 87/97. (Memoir of the Soil Survey of Great Britain: Scotland). (Edinburgh: Her Majesty’s Stationery Office for Department of Agriculture and Fisheries for Scotland.)

GLENTWORTH, R, MITCHELL, W A, and MITCHELL, B D. 1964. The red glacial drift deposits of north-east Scotland. Clay Minerals Bulletin, Vol. 5, 373–381.

GODWIN, H, and WILLIS, E H. 1959. Radiocarbon dating of the Late-glacial Period in Britain. Proceedings of the Royal Society of London, Series B, Vol. 150, 199–215.

GOODWIN, I D. 1998. Did changes in Antarctic ice volume influence Late Holocene sea-level lowering? Quaternary Science Reviews, Vol. 17, 319–332.

GORDON, D, and six others. 1989. Dating of Late Pleistocene Interglacial and Interstadial Periods in the United Kingdom from Speleothem Growth Frequency. Quaternary Research, Vol. 31, 14–26.

GORDON, J E. 1993a. Boyne Quarry. 233–236 in Quaternary of Scotland. GORDON, J E, and SUTHERLAND, D G (editors). Geological Conservation Review Series, No. 6 (London: Chapman and Hall.)

GORDON, J E. 1993b. Kippet Hills. 242–245 in Quaternary of Scotland. GORDON, J E, and SUTHERLAND, D G (editors). Geological Conservation Review Series, No. 6 (London: Chapman and Hall.)

GORDON, J E. 1993c. Nigg Bay. 479–482 in Quaternary of Scotland. GORDON, J E, and SUTHERLAND, D G (editors). Geological Conservation Review Series, No. 6 (London: Chapman and Hall.)

GORDON, J E. 1993d. Burn of Benholm. 482–485 in Quaternary of Scotland. GORDON, J E, and SUTHERLAND, D G (editors). Geological Conservation Review Series, No. 6 (London: Chapman and Hall.)

GORDON, J E, (editor). 1997. Reflections on the Ice Age in Scotland: an update on Quaternary studies. (Glasgow: Scottish Association of Geography Teachers and Scottish Natural Heritage).

GORDON, J E, and SUTHERLAND, D G (editors). 1993a. Quaternary of Scotland. Geological Conservation Review Series, No. 6 (London: Chapman and Hall.)

GORDON, J E, and SUTHERLAND, D G. 1993b. Windy Hills. 216–218 in Quaternary of Scotland. GORDON, J E, and SUTHERLAND, D G. Geological Conservation Review Series, No. 6 (London: Chapman and Hall.)

GORNITZ, V. 1995. Sea-level rise: a review of recent past and near-future trends. Earth Surface Processes and Landforms, Vol. 20, 7–20.

GOULD, D. 1996. Geology of the Findochty and Bin of Cullen areas. 1:10 000 sheets NJ46NE, NJ46SE and part of NJ45NE. British Geological Survey Technical Report, WA/96/68.

GOULD, D. 1997. Geology of the country around Inverurie and Alford. Memoir of the British Geological Survey, Sheet 76W and 76E (Scotland).

GOULD, D. 2001. Geology of the Aboyne district. Memoir of the British Geological Survey, Sheet 66W (Scotland).

GRAHAM, D K, HARLAND, R, GREGORY, D M, LONG, D, and MORTON, A C. 1990. The biostratigraphy and chronostratigraphy of BGS borehole 78/4, north Minch. Scottish Journal of Geology, Vol. 26, 65–75.

GRANT, R. 1960. The soils of the country round Elgin. Memoir of the Soil Survey (Scotland).

GRAY, J M, and LOWE, J J (editors). 1977. Studies in the Scottish Lateglacial Environment. (Oxford: Pergamon Press.)

GREENWOOD, P G, and RAINES, M G. 1994. Ground probing radar surveys in Scotland and East Anglia. Part 2 — Ground Probing Radar Sections. British Geological Survey Technical Report, WN/94/12.

GREENWOOD, P G, AUTON, C A, and RAINES, M G. 1995. Ground probing radar surveys in Scotland and East Anglia. Part 1 — Ground Probing Radar Report. British Geological Survey Technical Report, WN/95/6.

GREGORY, D M, and BRIDGE, V A. 1979. On the Quaternary foraminiferal species Elphidium ?ustulatum Todd 1957: its stratigraphic and palaeontological implications. Journal of Foraminiferal Research, Vol. 9, 70–75.

GROUSSET, F E, PUJOL, C, LABEYRIE, L, AUFFRET, G, and BOELAERT, A. 2000. Were the North Atlantic Heinrich events triggered by the behavior of the European ice sheets? Geology, Vol. 28, 123–126.

GUNSON, A R. 1975. The vegetation history of north-east Scotland. 61–72 in Quaternary studies in north-east Scotland. GEMMELL, A M D (editor). (Aberdeen: Department of Geography, University of Aberdeen.)

HALL, A M. 1982. The Pliocene gravels of Buchan: a reappraisal: discussion. Scottish Journal of Geology, Vol. 18, 336–338.

HALL, A M. 1983. Deep weathering and landform evolution in north-east Scotland. Unpublished PhD thesis, University of St Andrews.

HALL, A M. 1984a. Buchan field guide. (Cambridge: Quaternary Research Association.)

HALL, A M. 1984b. Oldmill. 106 in Buchan Field Guide. HALL, A M (editor). (Cambridge: Quaternary Research Association).

HALL, A M. 1985. Cenozoic weathering covers in Buchan, Scotland and their significance. Nature, Vol. 315, 392–395.

HALL, A M. 1986. Deep weathering patterns in north-east Scotland and their geomorphological significance. Zeitschrift für Geomorphologie, Vol. NF 30, 407–422.

HALL, A M. 1987. Weathering and relief development in Buchan, Scotland. Part II. 991–1005 in International Geomorphology 1986. GARDINER, V (editor). (Chichester: John Wiley.)

HALL, A M. 1991. Pre-Quaternary landscape evolution in the Scottish Highlands. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol. 8, 1–26.

HALL, A M. 1993a. Moss of Cruden. 218–221 in Quaternary of Scotland. GORDON, J E, and SUTHERLAND, D G (editors). Geological Conservation Review Series, No. 6 (London: Chapman and Hall.)

HALL, A M. 1993b. Pittodrie. 221–223 in Quaternary of Scotland. GORDON, J E, and SUTHERLAND, D G (editors). Geological Conservation Review Series, No 6 (London: Chapman and Hall.)

HALL, A M. 1993c. Hill of Longhaven. 223–225 in Quaternary of Scotland. GORDON, J E, and SUTHERLAND, D G (editors). Geological Conservation Review Series, No. 6 (London: Chapman and Hall.)

HALL, A M. 1995. Was all of Lewis glaciated in the Late Devensian? Quaternary Newsletter, Vol. 76, 1–7.

HALL, A M. 1997. Quaternary stratigraphy: the terrestrial record. 59–71 in Reflections on the Ice Age in Scotland. GORDON, J E (editor). (Glasgow: Scottish Association of Geography Teachers and Scottish Natural Heritage).

HALL, A M. 2000. The Teindland Interglacial site. 34–37 in The Quaternary of the Banffshire coast and Buchan: field guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

HALL, A M, and BENT, A J A. 1990. The limits of the last British Ice Sheet in northern Scotland and the adjacent shelf. Quaternary Newsletter, Vol. 61, 2–12.

HALL, A M, and CONNELL, E R. 1982. Recent excavations at the Greensand locality of Moreseat, Grampian Region. Scottish Journal of Geology, Vol. 18, 291–296.

HALL, A M, and CONNELL, E R. 1986. A preliminary report on the Quaternary sediments at Leys gravel pit, Buchan. Quaternary Newsletter, Vol. 48, 17–28.

HALL, A M, and CONNELL, E R. 1991. The glacial deposits of Buchan, northeast Scotland. 129–136 in Glacial Deposits in Great Britain and Ireland. EHLERS, J, GIBBARD, P L, and ROSE, J (editors). (Rotterdam: Balkema.)

HALL, A M, and CONNELL, E R. 2000. Howe of Byth Quarry. 72–78 in The Quaternary of the Banffshire coast and Buchan: field guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

HALL, A M, and JARVIS, J. 1989. A preliminary report on the Late Devensian glaciomarine deposits around St Fergus, Grampian Region. Quaternary Newsletter, Vol. 59, 5–7.

HALL, A M, and JARVIS, J. 1993a. Kirkhill. 225–230 in Quaternary of Scotland. GORDON, J E, and SUTHERLAND, D G (editors). Geological Conservation Review Series, No. 6 (London: Chapman and Hall.)

HALL, A M. and JARVIS, J. 1993b. Bellscamphie. 230–233 in Quaternary of Scotland. GORDON, J E, and SUTHERLAND, D G (editors). Geological Conservation Review Series, No. 6 (London: Chapman and Hall.)

HALL, A M, and JARVIS, J. 1994. A concealed Lower Cretaceous outlier at Moss of Cruden, Grampian Region. Scottish Journal of Geology, Vol. 30, 163–166.

HALL, A M, and JARVIS, J. 1995. A multiple till sequence near Ellon, Grampian Region: T. F. Jamieson’s ‘indigo boulder clay’ re-examined. Scottish Journal of Geology, Vol. 31, 53–59.

HALL, A M, and PEACOCK, J D. 2000. King Edward. 43–45 in The Quaternary of the Banffshire coast and Buchan: field guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

HALL, A M, and SUGDEN, D E. 1987. Limited modification of mid-latitude landscapes by ice sheets. Earth Surface Processes and Landforms, Vol. 12, 531–542.

HALL, A M, and WHITTINGTON, G. 1989. Late Devensian glaciation of southern Caithness. Scottish Journal of Geology, Vol. 25, 307–324.

HALL, A M, MELLOR, T, and WILSON, J. 1989. The clay mineralogy and age of deeply weathered rock in north-east Scotland. Zeitschrift der Geomorphologie Supplement Bund, Vol. 72, 97–108.

HALL, A M, WHITTINGTON, G, DULLER, G A T, and JARVIS, J. 1995a. Late Pleistocene environments in lower Strathspey, Scotland. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol. 85, 253–273.

HALL, A M, DULLER, G, JARVIS, J, and WINTLE, A G. 1995b. Middle Devensian ice-proximal gravels at Howe of Byth, Grampian Region. Scottish Journal of Geology, Vol. 31, 61–64.

HALL, A R. 1980. Late Pleistocene deposits at Wing, Rutland. Philosophical Transactions of the Royal Society of London, Vol. B289, 135–164.

HALLAM, A, and SELLWOOD, B. 1976. Middle Mesozoic sedimentation in relation to tectonics in the British area. Journal of Geology, Vol. 84, 302–321.

HAMBLIN, R J O. 1973. The Haldon Gravels of south Devon. Proceedings of the Geologists’ Association, Vol. 84, 459–476.

HANCOCK, J M. 1975. The petrography of the Chalk. Proceedings of the Geologists’ Association, Vol. 86, 499–536.

HARDE, K W. 1981. A field guide in colour to beetles. (London: Octopus Books.) HARKNESS, D D, and WILSON, H W. 1979. Scottish Universities Research and Reactor Centre radiocarbon measurements III. Radiocarbon, Vol. 21, 203–256.

HARKNESS, D D, and WILSON, H W. 1981. Scottish Universities Research and Reactor Centre radiocarbon measurements IV. Radiocarbon, Vol. 23, 252–304.

HARLAND, R. 1983. Distribution maps of recent dinoflagellate cysts in bottom sediments from the North Atlantic Ocean and adjacent seas. Palaeontology, Vol. 26, 321–387.

HARLAND, R. 1988. Quaternary dinoflagellate cyst biostratigraphy of the North Sea. Palaeontology, Vol. 31, 877–903.

HARLAND, R. 1992. Dinoflagellate cysts of the Quaternary System. 253–266 in A stratigraphic index of dinoflagellate cysts. POWELL, A J (editor). (London: Chapman and Hall.)

HARLAND, R. 1993. Dinoflagellate cyst analysis of the Quaternary ‘shelly till’ at the Burn of Benholm, Scottish Sheet 57A. British Geological Survey Technical Report, WH93/173R.

HARRIS, A L, HASELOCK, P J, KENNEDY, M J, and MENDUM, J R. 1994. The Dalradian Supergroup in Scotland, Shetland and Ireland. 33–53 in A revised correlation of Precambrian rocks in the British Isles. GIBBONS, W, and HARRIS, A L (editors). Special Report of the Geological Society of London, No. 22.

HART, J K, and BOULTON, G S. 1991. The interrelation of glacitectonic and glaciodepositional processes within the glacial environment. Quaternary Science Reviews, Vol. 10, 335–350.

HART, R. 1941. Soil studies in relation to geology in an area in north-east Scotland. Part 1. The mineralogy of the soil parent materials. Journal of Agricultural Science, Vol. 31, 438–447.

HEDBERG, H D (editor). 1976. International stratigraphic guide. (New York: John Wiley & Sons).

HEINRICH, H. 1988. Origin and consequences of cyclic ice rafting in the north-east Atlantic Ocean during the past 130 000 years. Quaternary Research, Vol. 29, 142–152.

HICOCK, S R, LIAN, O B, and MATHEWES, R W. 1999. ‘Bond cycles’ recorded in terrestrial Pleistocene sediments of southwestern British Columbia, Canada. Journal of Quaternary Science, Vol. 14, 443–449.

HILL, M O, and eight others. 1999. Climate changes and Scotland’s natural heritage: an environmental audit. Scottish Natural Heritage Research and Monitoring Report, No. 132.

HINXMAN, L W, and TEALL, J J H. 1896. The geology of West Aberdeenshire, Banffshire, parts of Elgin and Inverness. Memoir of the Geological Survey (Scotland).

HINXMAN, L W, and WILSON, J S G. 1902. The geology of Lower Strathspey. Memoir of the Geological Survey, Scotland, Sheet 85 (Scotland).

HOLMES, R. 1977. Quaternary deposits of the central North Sea, 5. The Quaternary geology of the UK sector of the North Sea between 56° and 58°N. Report of the Institute of Geological Sciences, No. 77/14.

HOLMES, R. 1991. Foula Sheet 60°N–04°W. Quaternary Geology. 1:250 000 Map Series. British Geological Survey, Edinburgh.

HOLMES, R. 1997. Quaternary stratigraphy: the offshore record. 72–94 in Reflections on the ice age in Scotland. GORDON, J E (editor). (Glasgow: Scottish Association of Geography Teachers and Scottish Natural Heritage.)

HOLMES, R, JEFFREY, D H, RUCKLEY, N A R, and WINGFIELD, R. 1993. Quaternary sediments around the United Kingdom, North Sheet, 1:1 000 000. British Geological Survey, Edinburgh.

HUIJZER, B, and VANDENBERGHE, J. 1998. Climatic reconstruction of the Weichselian pleniglacial in northwestern and central Europe. Journal of Quaternary Science, Vol. 13, 391–417.

HULL, E. 1895. The glacial deposits of Aberdeenshire. Geological Magazine, Vol. 2, 450–452.

HYDRAULICS RESEARCH. 1997. Coastal cells in Scotland. Scottish Natural Heritage Research, Survey and Monitoring Report, No. 56.

IMBRIE, J, and IMBRIE, K P. 1979. Ice Ages: solving the mystery. (London: MacMillan.)

IMBRIE, J, and eight others. 1984. The orbital theory of Pleistocene climate: support from a revised chronology of the marine 18O record. 269–305 in Milankovitch and climate: understanding the response to astronomical forcing. BERGER, A L, and four others (editors). (Dordrecht: Reidel).

JAMIESON, T F. 1858. On the Pleistocene deposits of Aberdeenshire. Quarterly Journal of the Geological Society of London, Vol. 14, 509–532.

JAMIESON, T F. 1859. On an outlier of Lias in Aberdeenshire. Quarterly Journal of the Geological Society of London, Vol. 15, 131–133.

JAMIESON, T F. 1860a. On the occurrence of crag strata beneath boulder clay in Aberdeenshire. Quarterly Journal of the Geological Society of London, Vol. 16, 371–373.

JAMIESON, T F. 1860b. On the drift and rolled gravel of the north of Scotland. Quarterly Journal of the Geological Society of London, Vol. 16, 347–371.

JAMIESON, T F. 1862. On the ice-worn rocks of Scotland. Quarterly Journal of the Geological Society of London, Vol. 18, 164–184.

JAMIESON, T F. 1865. On the history of the last geological changes in Scotland. Quarterly Journal of the Geological Society of London, Vol. 21, 161–203.

JAMIESON, T F. 1866. On the glacial phenomena of Caithness. Quarterly Journal of the Geological Society of London, Vol. 22, 261–281.

JAMIESON, T F. 1874. On the last stage of the glacial period in North Britain. Quarterly Journal of the Geological Society of London, Vol. 30, 317–338.

JAMIESON, T F. 1882a. On the crag shells of Aberdeenshire and the gravel beds containing them. Quarterly Journal of the Geological Society, Vol. 38, 145–159.

JAMIESON, T F. 1882b. On the Red Clay of the Aberdeenshire coast and the direction of ice-movement in that quarter. Quaterly Journal of the Geological Society of London, Vol. 38, 160–177.

JAMIESON, T F. 1906. The glacial period in Aberdeenshire and the southern border of the Moray Firth. Quarterly Journal of the Geological Society of London, Vol. 62, 13–39.

JAMIESON, T F. 1910. On the surface geology of Buchan. 19–25 in The Book of Buchan. TOCHER, J F, (editor). (Aberdeen and Peterhead: Aberdeen University Press and P. Scrogie Ltd).

JAMIESON, T F, JUKES-BROWNE, A J, and MILNE, J. 1897. Cretaceous fossils in Aberdeenshire. Report of the British Association for the Advancement of Science, 333–342.

JANSEN, J H F. 1976. Late Pleistocene and Holocene history of the northern North Sea, based on acoustic reflection records. Netherlands Journal of Sea Research, Vol. 10, 1–43.

JAPSEN, P. 1997. Regional Neogene exhumation of Britain and the western North Sea. Journal of the Geological Society of London, Vol. 154, 239–247.

JAPSEN, P. 1998. Regional velocity-depth anomalies, North Sea Chalk: a record of overpressure and Neogene uplift and erosion. American Association of Petroleum Geologists, Vol. 82, 2031–2074.

JARDINE, W G, and five others. 1988. A late Middle Devensian interstadial site at Sourlie, near Irvine, Strathclyde. Scottish Journal of Geology, Vol. 24, 288–295.

JEFFERY, D H, and LONG, D. 1989. Early Pleistocene sedimentation and geographic change in the UK sector of the North Sea. Terra Abstracts, Vol. 1, 425.

JENSEN, K A, and KNUDSEN, K L. 1988. Quaternary foraminiferal stratigraphy in boring 81/29 from the central North Sea. Boreas, Vol. 17, 273–287.

JOHNSON, H, RICHARDS, P C, LONG, D, and GRAHAM, C C. 1993. United Kingdom Offshore Regional Report: the Geology of the Northern North Sea. (London: HMSO for the British Geological Survey.)

JONES, L H P, and MILNE, A A. 1956. Birnessite, a new manganese oxide mineral from Aberdeenshire, Scotland. Mineralogical Magazine, Vol. 31, 283–288.

JORDT, H, FALEIDE, J I, BJØLYKKE, K, and IBRAHAM, M T. 1998. Cenozoic sequence stratigraphy of the central and northern North Sea basin: tectonic development, sediment distribution and provenance areas. Marine and Petroleum Geology, Vol. 12, 845–879.

KARLLSON, W, VOLLSET, J, BJORLYKKE, K, and JORGENSEN, P. 1979. Changes in the mineralogical composition of Tertiary sediments from North Sea wells. 281–289 in Proceedings of the 6th International Clay Conference. Development in sedimentology, No. 27. MORTLAND, M M, and FARMER, V C (editors). (Amsterdam, Netherlands: Elsevier.)

KESEL, R H, and GEMMELL, A M D. 1981. The ‘Pliocene’ gravels of Buchan: a reappraisal. Scottish Journal of Geology, Vol. 17, 185–203.

KNUDSEN, K L, and SEJRUP, H P. 1993. Pleistocene stratigraphy in the Devils Hole area, central North Sea: foraminiferal and amino-acid evidence. Journal of Quaternary Science, Vol. 8, 1–14.

KOPPI, A J. 1977. Weathering of Tertiary gravels, a schist and a metasediment in north-east Scotland. Unpublished PhD thesis, University of Aberdeen.

KOPPI, A J, and FITZPATRICK, E A. 1980. Weathering in Tertiary gravels in NE Scotland. Journal of Soil Science, Vol. 31, 525–532.

KOTILAINEN, A T, and SHACKLETON, N J. 1995. Rapid climate variability in the North Pacific Ocean during the past 95 000 years. Nature, London, Vol. 377, 323–326.

KROON, D, AUSTIN, W E N, CHAPMAN, M R, and GANSSEN, G M. 1997. Deglacial surface circulation changes in the northeastern Atlantic: temperature and salinity records off NW Scotland on a century scale. Palaeoceanography, Vol. 12, 755–763.

LAGERKLINT, I M, and WRIGHT, J D. 1999. Late glacial warming prior to Heinrich event 1: the influence of ice rafting and large ice sheets on the timing of initial warming. Geology, Vol. 27, 1099–1102.

LAMB, H H. 1977. Climate: present, past and future. Vol. 2. Climatic history and the future. (London: Methuen.)

LAMBECK, K. 1991. A model for Devensian and Flandrian glacial rebound and sea-level change in Scotland. 33–62 in Glacial isostacy, sea level and mantle rheology. SABADINI, R, LAMBECK, K, and BOSCHI, E, (editors). (Dordrecht: Kluwer Academic Publishers.)

LAMBECK, K. 1993a. Glacial rebound of the British Isles, I; preliminary model results. Geophysical Journal International, Vol. 115, 941–959.

LAMBECK, K. 1993b. Glacial rebound of the British Isles, II; a high resolution, high precision model. Geophysical Journal International, Vol. 115, 960–990.

LAMBECK, K. 1995. Late Devensian and Holocene shorelines of the British Isles and North Sea from models of glacio-hydro-isostatic rebound. Journal of the Geological Society of London, Vol. 152, 437–448.

LARSEN, E, and five others. 1987. Cave stratigraphy in western Norway; multiple Weichselian glaciations and interstadial vertebrate fauna. Boreas, Vol. 16, 267–292.

LAW, G R. 1962. The sub-surface geology of the Bay of Nigg–Tullos area. Unpublished MSc thesis, University of Aberdeen.

LAWSON, T J. 1984. Reindeer in the Scottish Quaternary. Quaternary Newsletter, Vol. 42, 1–7.

LAWSON, T J, and ATKINSON, T C. 1995. Quaternary chronology. 12–18 in The Quaternary of Assynt and Coigach: field guide. LAWSON, T J (editor). (Cambridge: Quaternary Research Association.)

LAXTON, J L. 1992. A particle-size classification of sand and gravel deposits as a basis for end-use assessment. Engineering Geology, Vol. 32, 29–37.

LEHMAN, S. 1996. True grit spells double trouble. Nature, London, Vol. 382, 25–26.

LESLIE, A G. 1984. Field relations in the north-eastern part of the Insch mafic igneous mass, Aberdeenshire. Scottish Journal of Geology, Vol. 20, 225–235.

LEWIN, J, and WEIR, M J C. 1977. Morphology and recent history of the Lower Spey. Scottish Geographical Magazine, Vol. 93, 45–51.

LIDMAR-BERGSTROM, K. 1982. Pre-Quaternary geomorphological evolution in southern Fennnoscandia. Sveriges Geologiska Undersökning, Vol. 785, 1–202.

LINDROTH, C H. 1985. The Carabidae (Coleoptera) of Fennoscandia and Denmark. In Fauna Entomologica Scandinavica. BRILL, E J (editor). Part 1. Vol. 15. (Copenhagen: Scandanavian Science Press Ltd.)

LINTON, D L. 1951. Problems of Scottish scenery. Scottish Geographical Magazine, Vol. 67, 65–85.

LINTON, D L. 1963. The forms of glacial erosion. Transactions of the Institute of British Geographers, Vol. 33, 1–28.

LONG, D, SMITH, D E, and DAWSON, A G. 1989. A Holocene tsunami deposit in eastern Scotland. Journal of Quaternary Science, Vol. 4, 61–66.

LOWE, J J. 1984. A critical evaluation of pollen-stratigraphic investigations of pre-Late Devensian sites in Scotland. Quaternary Science Reviews, Vol. 3, 405–432.

LOWE, J J, and eight others. 1999. The chronology of palaeoenvironmental changes during the last Glacial-Holocene transition: towards an event stratigraphy for the British Isles. Journal of the Geological Society of London, Vol. 156, 397–410.

LUBINSKY, I. 1980. Marine bivalve molluscs of the Canadian central and eastern Arctic: faunal composition and zoogeography. Canadian Bulletin of Fisheries and Aquatic Sciences, Vol. 207, 1–111.

LUMSDEN, G I. 1958. The glacial and post-glacial history of the lower reaches of the River Don, Aberdeenshire. Transactions of the Geological Society of Edinburgh, Vol. 17, 141–155.

MACKIE, W. 1905. Some notes on the distribution of erratics over eastern Moray. Transactions of the Edinburgh Geological Society, Vol. 8, 91–97.

MACKIE, W. 1923. The principles that regulate the distribution of particles of heavy minerals in sedimentary rocks, as illustrated by the sandstones of the north-east of Scotland. Transactions of the Edinburgh Geological Society, Vol. 11, 138–164.

MACPHERSON, E. 1971. The marine molluscs of arctic Canada. (Publications in Biological Oceanography, No 3). (Ottawa: National museums of Canada.)

MADGETT, P A, and CATT, J A. 1978. Petrography, stratigraphy and weathering of Late Pleistocene tills in East Yorkshire, Lincolnshire and north Norfolk. Proceedings of the Yorkshire Geological Society, Vol. 42, 55–108.

MAIZELS, J K, and AITKEN, J F. 1991. Palaeohydrological change during deglaciation in upland Britain: a case study from northeast Scotland. 105–145 in Temperate Palaeohydrology. STARKEL, L, GREGORY, K J, and THORNES, J B (editors). (Chichester: John Wiley.)

MANGERUD, J, and six others. 1981. A Middle Weichselian ice-free period in Western Norway: the Ålesund interstadial. Boreas, Vol. 10, 447–462.

MANGERUD, J, JANSEN, E, and LANDVIK, J.Y. 1996. Late Cenozoic history of the Scandinavian and Barents Sea ice sheets. Global and Planetary Change, Vol. 12, 11–26.

MASLIN, M A, and five others. 1995. Northwest Pacific Site 882: the initiation of Northern Hemisphere glaciation. Ocean Drilling Program, College Station, TX. Proceedings of the Ocean Drilling Program, Scientific Reports, Vol. 145, 315–329.

MASLIN, M A, LI, X S, LOUTRE, M-F, and BERGER, A. 1998. The contribution of orbital forcing to the progressive intensification of Northern Hemisphere glaciation. Quaternary Science Reviews, Vol. 17, 411–426.

MAUZ, B. 1998. The onset of the Quaternary: a review of new findings in the Pliocene–Pleistocene chronostratigraphy. Quaternary Science Reviews, Vol. 17, 357–364.

MAYLE, F E, and nine others. 1999. Climate variations in Britain during the last Glacial-Holocene transition (15.0–11.5 cal ka BP): comparison with the GRIP ice-core record. Journal of the Geological Society of London, Vol. 156, 411–423.

MCCABE, A M. 1996. Dating and rhythmicity from the last deglacial cycle in the British Isles. Journal of the Geological Society of London, Vol. 153, 499–502.

MCCABE, A M, and CLARK, P U. 1998. Ice-sheet variability around the North Atlantic Ocean during the last deglaciation. Nature, London, Vol. 392, 373–377.

MCCABE, A M, KNIGHT, J, and MCCARRON, S. 1998. Evidence for Heinrich event 1 in the British Isles. Journal of Quaternary Science, Vol. 13, 549–568.

MCDONALD, B C, and SHILTS, W W. 1975. Interpretation of faults in glaciofluvial sediments. 123–131 in Glaciofluvial and glaciolacustrine sedimentation. JOPLING, A V, and MCDONALD, B C (editors). Society for Economic Paleontologists and Mineralogists (Society for Sedimentary Geology) Special Publication, No. 23.

MCEWEN, L J. 1997. Geomorphological change and fluvial landscape evolution in Scotland during the Holocene. 116–129 in Reflections on the Ice Age in Scotland. GORDON, J E (editor). (Glasgow: Scottish Association of Geography Teachers and Scottish Natural Heritage.)

MCGOWN, A. 1971. The classification for engineering purposes of tills from moraines and associated landforms. Journal of Engineering Geology, Vol. 4, 115–130.

MCGOWN, A, and DERBYSHIRE, E. 1977. Genetic influences on the properties of tills. Quarterly Journal of Engineering Geology, Vol. 10, 389–410.

MCGREGOR, D M, and WILSON, J S G. 1967. Gravity and magnetic surveys of the younger gabbros of Aberdeenshire. Quarterly Journal of the Geological Society of London, Vol. 123, 99–123

MCLEAN, F. 1977. The glacial sediments of a part of Aberdeenshire. Unpublished PhD thesis, University of Aberdeen.

MCMILLAN, A A, and AITKEN, A M. 1981. The sand and gravel resources of the country west of Peterhead, Grampian Region. Description of 1:25 000 sheet NK04 and parts of NJ94, 95 and NK 05, 14 and 15. Mineral Assessment Report of the Institute of Geological Sciences, No. 58.

MCMILLAN, A A, and MERRITT, J W. 1980. A reappraisal of the ‘Tertiary’ deposits of Buchan, Grampian Region. Report of the Institute of Geological Sciences, No. 80/1, 18–25.

MCMILLAN, A A, AND HAMBLIN, R J O. 2000. A mapping related lithostratigraphical framework for the Quaternary of the UK. Quaternary Newsletter, No. 92, 21–34.

MENKE, B, and BEHRE, K E. 1973. History of vegetation and biostratigraphy. Eiszeitalter und Gegenwart, Vol. 23, 251–267.

MERRITT, J W. 1981. The sand and gravel resources of the country around Ellon, Grampian Region. Description of 1:25 000 resource sheets NJ93 with parts of NJ82, 83 and 92, and NK03 and parts of NK02 and 13. Mineral Assessment Report of the Institute of Geological Sciences, No. 76.

MERRITT, J W. 1992a. A critical review of methods used in the appraisal of onshore sand and gravel resources in Britain. Engineering Geology, Vol. 32, 1–9.

MERRITT, J W. 1992b. The high-level marine shell-bearing deposits of Clava, Invernessshire, and their origin as glacial rafts. Quaternary Science Reviews, Vol. 11, 759–779.

MERRITT, J W, and CONNELL, E R. 2000. Oldmill Quarry. 68–71 in The Quaternary of the Banffshire coast and Buchan: field guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

MERRITT, J W, and MCMILLAN, A A. 1982. Letters to the editors. The ‘Pliocene’ gravels of Buchan: a reappraisal. Scottish Journal of Geology, Vol. 18, 329–332.

MERRITT, J W, and PEACOCK, J D. 1983a. A preliminary study of the sand and gravel deposits around Aberdeen (1:25 000 sheets NJ71, 80, 81, 90, 91 and parts of 61 and 92; NO69 and 79 and parts of 89). Institute of Geological Sciences, Industrial Mineral Assessment Unit, Open File Report, WA/HI/83/4.

MERRITT, J W, and PEACOCK, J D. 1983b. A preliminary study of the sand and gravel deposits around Inverness and Nairn, Highland Region and Forres and Elgin, Grampian Region (1:25 000 sheets NH63, 64, 73, 75, 84, 85, 94, 95 and 96; NJ05, 06, 15, 16, 17, 25, 26, 27, 35 and part of 04). Institute of Geological Sciences, Industrial Mineral Assessment Unit, Open File Report, WA/HI/83/5.

MERRITT, J W, and PEACOCK, J D. 2000. Glacial meltwater channels in Banffshire and Buchan with particular reference to the area around the Tore of Troup. 46–48 in The Quaternary of the Banffshire coast and Buchan: field guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

MERRITT, J W, AUTON, C A, and ROSS, D L. 1988. Summary assessment of the sand and gravel resources of northeast Scotland. British Geological Survey Technical Report, WF/88/2.

MERRITT, J W, AUTON, C A, and FIRTH, C R. 1995. Ice-proximal glaciomarine sedimentation and sea-level change in the Inverness area, Scotland: a review of the deglaciation of a major ice stream of the British Late Devensian ice sheet. Quaternary Science Reviews, Vol. 14, 289–329.

MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R. 2000. The Quaternary of the Banffshire coast and Buchan: field guide. (London: Quaternary Research Association.)

MERRITT, J W, CONNELL, E R, HALL, A M, and PEACOCK, J D. 2000. An introduction to the Cainozoic geology of coastal Morayshire, Banffshire and Buchan. 1–20 in The Quaternary of the Banffshire coast and Buchan: field guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

MILLER, G H. 1990. Additional dating methods. 405–413 in Geomorphological Techniques. GOUDIE, A (edited for the British Geomorphological Research Group). (London: Unwin Hyman.)

MILLER, G H, and MANGERUD, J. 1985. Aminostratigraphy of European marine interglacial deposits. Quaternary Science Reviews, Vol. 4, 215–278.

MILLER, G H, BRIGHAM, J K, and CLARK, P. 1982. Alteration of the Total aIle/Ile ratio by different methods of sample preparation. 9–20 in Amino acid geochronology laboratory. January 1981 through May 1982. Report of current activities. MILLER, G H (editor). Institute of Arctic and Alpine Research and Department of Geological Sciences, University of Colorado.

MILLER, G H, SEJRUP, H P, MANGERUD, J, and ANDERSEN, B G. 1983. Amino acid ratios in Quaternary molluscs and foraminifera from western Norway: aminostratigraphy and paleotemperature estimates. Boreas, Vol. 12, 107–124.

MILLER, G H, and seven others. 1987. Racemization-derived late Devensian temperature reduction in Scotland. Nature, London, Vol. 326, 593–595.

MILLER, H. 1859. Sketchbook of popular geology. (Edinburgh: Constable.)

MILNE, J. 1892a. Geology of Buchan. Transactions of the Buchan Field Club, Vol. 2, 23–38.

MILNE, J. 1892b. Drift rocks in Buchan. Transactions of the Buchan Field Club, Vol. 2, 181–198.

MITCHELL, G F, PENNY, L F, SHOTTON, F W, and WEST, R G. 1973. A correlation of Quaternary deposits in the British Isles. Geological Society of London Special Report, No. 4.

MUNRO, M. 1970. A re-asssessment of the ‘younger’ basic igneous rocks between Huntly and Portsoy based on new borehole evidence. Scottish Journal of Geology, Vol. 6, 41–52.

MUNRO, M. 1986. Geology of the country around Aberdeen. Memoir of the British Geological Survey, Sheet 77 (Scotland).

MUNRO, M, and GALLACHER, J W. 1984. Cumulate relations in the ‘Younger Basic’ masses of the Huntly–Portsoy area, Grampian Region. Scottish Journal of Geology, Vol. 20, 361–382.

MURDOCH, W M. 1975. The geomorphology and glacial deposits of the area around Aberdeen. 14–18 in Quaternary Studies in North East Scotland. GEMMELL, A M D (editor). (Aberdeen: University of Aberdeen.)

MURDOCH, W M. 1977. The glaciation and deglaciation of south east Aberdeenshire. Unpublished PhD thesis, University of Aberdeen.

MUSSON, R M W. 1994. A catalogue of British earthquakes. British Geological Survey Technical Report, WL/94/04.

MUSSON, R M W. 1996. The seismicity of the British Isles. Annali di Geofisica, Vol. 39, 463–469.

MUSSON, R M W, NEILSON, G, and BURTON, P W. 1986. Macroseismic reports on historical British earthquakes X11: 1927 January 24 North Sea. British Geological Survey Global Seismology Unit Report, WL/GS/86/281.

MYKURA, W. 1983. Old Red Sandstone. 205–251 in Geology of Scotland. (second edition). CRAIG, G Y (editor). (Edinburgh: Scottish Academic Press.)

NATURAL ENVIRONMENT RESEARCH COUNCIL. 1998. UKDMAP CD-ROM. United Kingdom Digital Marine Atlas. (Third edition). (Bidston Observatory, Birkenhead: British Oceanographic Data Centre.)

NESJE, A, and SEJRUP, H P. 1988. Late Weichselian/Late Devensian ice sheets in the North Sea and adjacent areas. Boreas, Vol. 17, 371–384.

NICOL, J. 1860. On the geological structure of the vicinity of Aberdeen and the north-east of Scotland. Report of the British Association for the Advancement of Science, 29th meeting, 116–119.

NORTON, P E P, and SPAINK, G. 1973. The earliest occurrence of Macoma balthica (L.) as a fossil in the North Sea deposits. Malacologia, Vol. 14, 33–37.

OCKELMANN, W K. 1954. On the interrelationship and the zoogeography of northern species of Yoldia Møller, S. Str. (Mollusca, fam. Ledidae). Meddelelser om Grønland, Vol. 107, Nr. 7.

OGILVIE, A G. 1923. The physiography of the Moray Firth coast. Transactions of the Royal Society of Edinburgh, Vol. 53, 377–404.

OLSEN, L. 1997. Rapid shifts in glacial extension characterise a new concepted model for glacial variations during the Mid and Late Weichselian in Norway. Norges Geologiste Undersokelse, Bulletin, 433, 54–55.

OSBORNE, P J. 1988. A late Bronze Age insect fauna from the River Avon, Warwickshire, England: its implications for the terrestrial and fluvial environment and for climate. Journal of Archaeological Science, Vol. 15, 715–727.

OWEN, D E. 1987. Commentary: usage of stratigraphic terminology in papers, illustrations, and talks. Journal of Sedimentary Petrology, Vol. 57, 363–372.

OWENS, B, and MARSHALL, J, (compilers). 1978. Micropalaeontological biostratigraphy of samples from around the coasts of Scotland. Report of the Institute of Geological Sciences, No. 78/20.

PATERSON, I B. 1974. The supposed Perth Re-advance in the Perth district. Scottish Journal of Geology, Vol. 10, 53–66.

PAUL, M A, and EYLES, N. 1990. Constraints on the preservation of diamict facies (melt-out tills) at the margins of stagnant glaciers. Quaternary Science Reviews, Vol. 9, 51–69.

PAUL, M A, and LITTLE, J A. 1991. Geotechnical properties of glacial deposits in lowland Britain. 389–403 in Glacial deposits in Great Britain and Ireland. EHLERS, J, GIBBARD, P L, and ROSE, J (editors). (Rotterdam: Balkema.)

PEACOCK, J D. 1966. Note on the drift sequence near Portsoy, Banffshire. Scottish Journal of Geology, Vol. 2, 35–37.

PEACOCK, J D. 1971. A re-interpretation of the coastal deposits of Banffshire and their place in the late-Glacial history of NE Scotland. Bulletin of the Geological Survey of Great Britain, Vol. 37, 81–89.

PEACOCK, J D. 1975. Scottish Late- and Post-glacial marine deposits. 45–48 in Quaternary studies in north east Scotland. GEMMELL, A M D (editor). (Aberdeen: University of Aberdeen.)

PEACOCK, J D. 1979. Mill of Dyce. 203–256 in Scottish Universities Research and Reactor Centre radiocarbon measurements III. HARKNESS, D D, and WILSON, H W (editors). Radiocarbon, Vol. 21.

PEACOCK, J D. 1980. An overlooked record of interglacial or interstadial sites in north-east Scotland. Quaternary Newsletter, Vol. 32, 14–15.

PEACOCK, J D. 1983. Planning for development: Peterhead project. British Geological Survey Technical Report, WA/HI/83/1.

PEACOCK, J D. 1984. Errolston. 108–109 in Buchan field guide. HALL, A M (editor). (Cambridge: Quaternary Research Association.)

PEACOCK, J D. 1985. Comments on the Quaternary deposits and landforms of Scotland and the neighbouring shelves: a review. Quaternary Science Reviews, Vol. 4, i–ii.

PEACOCK, J D. 1995. Late Devensian to early Holocene palaeoenvironmental changes in the Viking Bank area, northern North Sea. Quaternary Science Reviews, Vol. 14, 1029–1042.

PEACOCK, J D. 1996. Marine molluscan proxy data applied to Scottish late glacial and Flandrian sites: strengths and limitations. 215–228 in Late Quaternary Palaeoceanography of the North Atlantic Margins. ANDREWS, J T, AUSTIN, W E N, BERGSTEN, H, and JENNINGS, A E (editors). Special Publication of the Geological Society of London, No. 111.

PEACOCK, J D. 1997. Was there a re-advance of the British ice sheet into the North Sea between 15 ka and 14 ka BP? Quaternary Newsletter, Vol. 81, 1–8.

PEACOCK, J D. 1999. The pre-Windermere Interstadial (Late Devensian) raised marine strata of eastern Scotland and their macrofauna: a review. Quaternary Science Reviews, Vol. 18, 1655–1679.

PEACOCK, J D. 2000. A glaciated pavement and crossing striations between Fraserburgh and Rattray Head. 56 in The Quaternary of the Banffshire coast and Buchan: field guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

PEACOCK, J D, and BROWNE, M A E. 1998. Radiocarbon dates from the Errol Beds (pre-Windermere Interstadial raised marine deposits) in eastern Scotland. Quaternary Newsletter, Vol. 86, 1–7.

PEACOCK, J D, and CONNELL, E R. 2000. Glacial lakes in the valleys of the North Ugie and South Ugie. 49–52 in The Quaternary of the Banffshire coast and Buchan: field guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

PEACOCK, J D, and HARKNESS, D D. 1990. Radiocarbon ages and the full-glacial to Holocene transition in seas adjacent to Scotland and southern Scandinavia: a review. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol. 81, 385–396.

PEACOCK, J D, and MERRITT, J W. 1997. Glacigenic rafting at Castle Hill, Gardenstown, and its significance for the glacial history of northern Banffshire, Scotland. Journal of Quaternary Science, Vol. 12, 283–294.

PEACOCK, J D, and MERRITT, J W. 2000a. Glacial deposits at the Boyne Limestone Quarry, Portsoy, and the late-Quaternary history of coastal Banffshire. Journal of Quaternary Science, Vol. 15, 543–555.

PEACOCK, J D, and MERRITT, J W. 2000b. Boyne Bay Limestone Quarry. 25–32 in The Quaternary of the Banffshire coast and Buchan: field guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

PEACOCK, J D, and MERRITT, J W. 2000c. So-called ‘pre-Glacial’ raised beach deposits deposits on the Banffshire coast. 33 in The Quaternary of the Banffshire coast and Buchan: Field Guide. (London: Quaternary Research Association.)

PEACOCK, J D, and MERRITT, J W. 2000d. Castle Hill, Gardenstown. 40–42 in The Quaternary of the Banffshire coast and Buchan: field guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

PEACOCK, J D, and MERRITT, J W. 2000e. Exposures in the Blackhills Sand and Gravel Formation and related deposits. 53–55 in The Quaternary of the Banffshire coast and Buchan: field guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Assocation.)

PEACOCK, J D, BERRIDGE, N G, HARRIS, A L, and MAY, F. 1968. The geology of the Elgin district. Memoir of the Geological Survey, Sheet 95 (Scotland).

PEACOCK, J D, and seven others. 1977. Sand and gravel resources of the Grampian region. Report of the Institute of Geological Sciences, No. 77/2.

PEACOCK, J D, GRAHAM, D K, and WILKINSON, I P. 1978. Late-glacial and post-glacial marine environments at Ardyne, Scotland, and their significance in the interpretation of the history of the Clyde sea area. Report of the Institute of Geological Sciences, No. 78/17.

PEACOCK, J D, and five others. 1992. Late Devensian and Flandrian palaeoenvironmental changes in the Scottish continental shelf west of the Outer Hebrides. Journal of Quaternary Science, Vol. 7, 145–161.

PEACOCK, J D, HALL, A M, and CONNELL, E R. 2000. St Fergus. 62–64 in The Quaternary of the Banffshire coast and Buchan: field guide. (London: Quaternary Research Association.)

PEARS, N V. 1968. Post-glacial tree-lines of the Cairngorm Mountains. Transactions of the Botanical Society of Edinburgh, 40, 361–394.

PEARS, N V. 1970. Post-glacial tree-lines of the Cairngorm Mountains; some modifications based on radiocarbon dating. Transactions of the Botanical Society of Edinburgh, Vol. 40, 536–544.

PEARSON, D G, EMELEUS, C H, and KELLEY, S P. 1996. Precise 40Ar/39Ar age for the initiation of Palaeogene volcanism in the Inner Hebrides and its regional significance. Journal of the Geological Society of London, Vol. 153, 815–818.

PENNEY, D. 1990. Quaternary ostracod chronology of the central North Sea: the record from BH 81/29. Courier Forschungsinstitut Senckenberg, Vol. 123, 97–109.

PHILLIPS, L. 1976. Pleistocene vegetational history and geology in Norfolk. Philosophical Transactions of the Royal Society of London, Vol. B275, 215–286.

PIRAZZOLI, P.A. 1993. Global sea-level changes and their measurement. Global and Planetary Change, Vol. 8, 135–148.

PONS, A, GUIOT, J L, BEAULIEU, J L, DE, and REILLE, M. 1992. Recent contributions to the climatology of the last glacial–interglacial cycle based on French pollen sequences. Quaternary Science Reviews, Vol. 11, 439–448.

POPE, R D. 1977. Revised version of ‘A check list of British insects’ by G S Kloet, and W D Hinks. Second edition, part 3. Royal Entomological Society of London, Handbooks for the identification of British Insects, Volume 11, Part 3.

POWELL, A J. 1992. Dinoflagellate cysts of the Tertiary System. 253–266 in A stratigraphic index of dinoflagellate cysts. POWELL, A J (editor). (London: Chapman and Hall.)

PRESTWICH, J. 1837. Observations on the Ichthyolites of Gamrie in Banffshire, and on the accompanying red conglomerates and sandstones. Transactions of the Geological Society of London, Vol. 5, 139–148.

PRICE, R J. 1983. Scotland’s environment during the last 30 000 years. (Edinburgh: Scottish Academic Press.)

QUATERNARY RESEARCH ASSOCIATION. 1975. Field excursion guide: Aberdeen. (Informal publication: Quaternary Research Association.)

RAMSAY, D M. 1965. Factors influencing aggregate impact value in rock aggregate. Quarry Manager’s Journal, Vol. 49, 129–134.

RAMSAY, D M, DHIR, R K, and SPENCE, I M. 1973. The reproducibility of results in the aggregate impact test: the influence of non-geological factors. ‘Quarry Managers’ Journal, Vol. 57, 179–181.

RAMSAY, D M, DHIR, R K, and SPENCE, I M. 1974. The role of rock and clast fabric in the physical performance of crushed-rock aggregate. Engineering Geology, Vol. 8, 267–285.

RAMSAY, K. 1999. Geotechnical properties of glacial tills in the Aberdeen area. Unpublished MSc thesis, Heriot-Watt University, Edinburgh.

RAYMO, M E. 1994. The initiation of Northern Hemisphere glaciation. Annual Review of Earth and Planetary Sciences, Vol. 22, 353–383.

RAYMO, M E. 1997. The timing of major climate terminations. Paleoceanography, Vol. 12, 577–585.

RAYMO, M E, and RUDDIMAN, W F. 1992. Tectonic forcing of late Cenozoic climate. Nature, London, Vol. 359, 117–122.

READ, H H. 1923. The geology of the country around Banff, Huntly and Turriff (Lower Banffshire and north-west Aberdeenshire). Memoir of the Geological Survey, Scotland, sheets 86 and 96.

READ, H H, BREMNER, A, CAMPBELL, R, and GIBB, A W. 1923. Records of the occurrence of boulders of Norwegian rocks in Aberdeenshire and Banffshire. Transactions of the Edinburgh Geological Society, Vol. 11, 230–231.

RIDGWAY, J M. 1982. Common clay and shale. Mineral Dossier Mineral Resources Consultative Committee, No. 22.

RIDING, J B. 1998. A palynological investigation of a Quaternary sample from Boyne Bay, Banffshire (Scottish 1:50 000 Sheet 96). British Geological Survey: Edinburgh, Technical Report, WH/98/17R.

RITCHIE, W. 1992. Scottish landform examples — 4. Coastal parabolic dunes of the Sands of Forvie. Scottish Geographical Magazine, Vol. 108, 39–44.

RITCHIE, W, SMITH, J S, and ROSE, N. 1978. Beaches of northeast Scotland. (Aberdeen: Department of Geography, University of Aberdeen.)

ROBINS, N S. 1990. Hydrogeology of Scotland. (London: HMSO for the British Geological Survey.)

ROMANS, J C C. 1977. Stratigraphy of buried soil at Teindland Forest, Scotland. Nature, London, Vol. 268, 622–623.

ROSE, J. 1985. The Dimlington Stadial/Dimlington Chronozone: a proposal for the naming of the main glacial episode of the Late Devensian in Britain. Boreas, Vol. 14, 225–230.

RUDDIMAN, W F, and KUTZBACH, J E. 1991. Plateau uplift and climatic change. Scientific American, Vol. 264, 66–75.

RUDDIMAN, W F, and MCINTYRE, A. 1973. Time-transgressive deglacial retreat of polar waters from the North Atlantic. Quaternary Research, Vol. 3, 117–130.

RUDDIMAN, W F, and MCINTYRE, A. 1976. North-east Atlantic palaeoclimatic changes over the past 600 000 years. 111–146 in Investigation of late Quaternary palaeoceanography and palaeoclimatology. CLINE, R M, and HAGS, J D (editors). Geological Society of America Memoir, 145.

RUDDIMAN, W F, and MCINTYRE, A. 1981. The mode and mechanism of the last deglaciation: oceanic evidence. Quaternary Research, Vol. 16, 125–134.

RUDDIMAN, W F, and RAYMO, M E. 1988. Northern Hemisphere climate régimes during the past 3 Ma: possible tectonic connections. Philosophical Transactions of the Royal Society of London, Vol. B318, 411–430.

RUDDIMAN, W F, SANCETTA, C D, and MCINTYRE, A. 1977. Glacial/interglacial response rate of subpolar North Atlantic waters to climatic change: the record in oceanic sediments. Philosophical Transactions of the Royal Society of London, Vol. B280, 119–142.

RUDDIMAN, W F, RAYMO, M E, MARTINSON, D G, CLEMENT, B M, and BACKMAN, J. 1989. Pleistocene evolution: northern hemisphere ice sheets and North Atlantic Deep-Sea Core V23–82: correlation with the terrestrial record. Palaeoceanography, Vol. 4, 353–412.

SALTER, J W. 1857. On the Cretaceous fossils of Aberdeenshire. Quarterly Journal of the Geological Society of London, Vol. 13, 83–89.

SAVILLE, A. 2000. The NMS Project on flint mining in Scotland. 93–95 in The Quaternary of the Banffshire coast and Buchan: field guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

SCHRAG, D P. 2000. Of ice and elephants. Nature, London, Vol. 404, 23–24.

SCOTT, D. 1890. Excursion to Links of St Fergus. Transactions of the Buchan Field Club, Vol. 1, 86–92.

SEJRUP, H P, and KNUDSEN, K L. 1993. Paleoenvironments and correlations of interglacial sediments in the North Sea. Boreas, Vol. 22, 223–235.

SEJRUP, H P, and nine others. 1987. Quaternary stratigraphy of the Fladen area, central North Sea: a multidisciplinary study. Journal of Quaternary Science, Vol. 2, 35–58.

SEJRUP, H P, and six others. 1994. Late Weichselian glaciation history of the northern North Sea. Boreas, Vol. 23, 1–13.

SEJRUP, H P, and seven others. 1995. Quaternary of the Norwegian Channel; paleoceanography and glaciation history. Norsk Geologisk Tidsskrift, Vol. 75, 65–87.

SEJRUP, H P, and five others. 1998. The Jaeren area: a border zone of the Norwegian Channel Ice Stream. Quaternary Science Reviews, Vol. 17, 801–812.

SEJRUP, H P, and five others. 2000. Quaternary glaciations in southern Fennoscandia: evidence from southwestern Norway and the northern North Sea region. Quaternary Science Reviews, Vol. 19, 667–685.

SELBY, I. 1989. Quaternary geology of the Hebridean continental margin. Unpublished PhD thesis, University of Nottingham.

SHACKLETON, N J, and OPDYKE, N D. 1973. Oxygen isotope and palaeomagnetic stratigraphy of equatorial Pacific core V28–238: oxygen isotope temperatures and ice volumes on a 105 year and 106 year scale. Quaternary Research, Vol. 3, 39–55.

SHACKLETON, N J, and sixteen others. 1984. Oxygen isotope calibration of the onset of ice-rafting and history of glaciation in the North Atlantic region. Nature, London, Vol. 307, 620–623.

SHACKLETON, N J, BERGER, A, and PELTIER, W R. 1990. An alternative astronomical calibration of the lower Pleistocene time based on ODP Site 677. Transactions of the Royal Society of Edinburgh for Earth Sciences, Vol. 81, 251–261.

SHAW, J. 1979. Genesis of the Sveg tills and rogen moraines of Central Sweden: a model of basal meltout. Boreas, Vol. 8, 409–426.

SHENNAN, I. 1989. Holocene crustal movements and sea-level changes in Britain. Journal of Quaternary Science, Vol. 4, 77–89.

SIMPSON, S. 1948. The glacial deposits of Tullos and Bay of Nigg, Aberdeen. Transactions of the Royal Society of Edinburgh, Vol. 61, 687–687.

SIMPSON, S. 1955. A re-interpretation of the drifts of north-east Scotland. Transactions of the Edinburgh Geological Society, Vol. 16, 189–199.

SISSONS, J B. 1958. Supposed ice-dammed lakes in Britain with particular reference to the Eddleston Valley, southern Scotland. Geografiska Annaler, Vol. 40, 159–187.

SISSONS, J B. 1961a. A subglacial drainage system by the Tinto Hills, Lanarkshire. Transactions of the Edinburgh Geological Society, Vol. 18, 175–193.

SISSONS, J B. 1961b. Some aspects of glacial drainage channels in Britain. Scottish Geographical Magazine, Vol. 77, 15–36.

SISSONS, J B. 1965. Quaternary. 467–503 in The geology of Scotland. CRAIG, G Y, (editor). (Edinburgh: Oliver and Boyd).

SISSONS, J B. 1967. The evolution of Scotland’s scenery. (Edinburgh & London: Oliver & Boyd.)

SISSONS, J B. 1974a. A lateglacial ice-cap in the central Grampians. Transactions of the Institute of British Geographers, Vol. 62, 95–114.

SISSONS, J B. 1974b. Lateglacial marine erosion in Scotland. Boreas, Vol. 3, 41–48.

SISSONS, J B. 1974c. The Quaternary in Scotland: a review. Scottish Journal of Geology, Vol. 10, 311–337.

SISSONS, J B. 1976. The geomorphology of the British Isles: Scotland. (London: Methuen.)

SISSONS, J B. 1979. The Loch Lomond Stadial in the British Isles. Nature, London, Vol. 280, 199–203.

SISSONS, J B. 1981. The last Scottish ice-sheet: facts and speculative discussion. Boreas, Vol. 10, 1–17.

SISSONS, J B. 1982. Interstadial and last interglacial deposits covered by till in Scotland: a reply. Boreas, Vol. 11, 123–124.

SISSONS, J B. 1983. Quaternary. 399–424 in The geology of Scotland. (second edition). G Y CRAIG (editor). (Edinburgh: Scottish Academic Press.)

SKEMPTON, A W, and NORTHEY, R D. 1952. The sensitivity of clays. Géotechnique, Vol. 3, 30–53.

SKINNER, A C, MCELVENNEY, E, RUCKLEY, N, RISE, L, and ROKOENGEN, K. 1986. Cormorant Sheet 61°N–00°. Quaternary geology. 1:250 000. British Geological Survey.)

SKINNER, B J, and PORTER, S C. 1995. The Dynamic Earth. (Third edition). (New York: Wiley.)

SMITH, C G. 1983. Planning for development: Aberdeen project. British Geological Survey Technical Report, WA/HI/86/1.

SMITH, D E. 1984. Lower Ythan Valley. 47–58 in Buchan field guide. HALL, A M (editor). (Cambridge: Quaternary Research Association.)

SMITH, D E. 1993. Philorth Valley. 251–254 in Quaternary of Scotland. GORDON, J E, and SUTHERLAND, D G (editors). Geological Conservation Review Series, No. 6 (London: Chapman and Hall.)

SMITH, D E, and CULLINGFORD, R A. 1985. Flandrian relative sea-level changes in the Montrose Basin area. Scottish Geographical Magazine, Vol. 101, 91–105.

SMITH, D E, and DAWSON, A G. 1990. Tsumani waves in the North Sea. New Scientist, Vol. 127, 46–49.

SMITH, D E, CULLINGFORD, R A, and SEYMOUR, W P. 1982. Flandrian relative sea-level changes in the Philorth Valley, north-east Scotland. Transactions of the Institute of British Geographers (New Series), Vol. 7, 321–336.

SMITH, D E, CULLINGFORD, R A, and BROOKS, C L. 1983. Flandrian relative sea level changes in the Ythan Valley, north-east Scotland. Earth Surface Processes and Landforms, Vol. 8, 423–438.

SMITH, D E, FIRTH, C R, BROOKS, C L, ROBINSON, M, and COLLINS, P E F. 1999. Relative sea-level rise during the Main Postglacial Transgression in NE Scotland, U.K. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol. 90, 1–27.

SMITH, J, SCOTT, T, and STEEL, J. 1904. The post-drift fossils of the Clyde drainage area at low levels. 528–545 in The geology and palaeontology of the Clyde drainage area. MURDOCH, J B (editor). (Glasgow: Geological Society of Glasgow.)

SMITH, J S, MATHER, A S, and GEMMELL, A M D. 1977. A landform inventory of Grampian Region. (Aberdeen: Department of Geography, University of Aberdeen.)

SMITH, M R, and COLLIS, L (editors). 1993. Aggregates. Sand, gravel and crushed rock aggregates for construction purposes. (Second Edition.) Special Publication of the Geological Society of London, No. 9.

SMITH, S M, and HEPPELL, D. 1991. Check list of British marine Mollusca. National Museums of Scotland Information Series (Natural History), No. 11.

SOIL SURVEY OF SCOTLAND. 1983. East of Scotland. Land capability for agriculture. Sheet 5. 1:250 000. (Aberdeen: The Ordnance Survey, Southampton for the Macaulay Institute for Soil Research.)

STACE, C. 1991. New Flora of the British Isles. (Cambridge: Cambridge University Press.)

STEERS, A. 1937. The Culbin Sands and Burghead Bay. Geological Journal, Vol. 90, 498 (Second Edition.) 528.

STEPHENSON, D, and GOULD, D. 1995. British regional geology: the Grampian Highlands. (fourth edition). (London: HMSO for the British Geological Survey.)

STEVENSON, A S. 1991. Miller Sheet 61° N–02°W. Quaternary Geology. 1:250 000 Map Series. British Geological Survey.

STEWART, F S. 1991. A reconstruction of the eastern margin of the late Weichselian ice sheet in northern Britain. Unpublished PhD thesis, University of Edinburgh.

STEWART, F S, and STOKER, M S. 1990. Problems associated with seismic facies analysis of diamicton-dominated, shelf glacigenic sequences. Geo-Marine Letters, Vol. 10, 151–156.

STOKER, M S. 1988. Pleistocene ice-proximal glaciomarine sediments in boreholes from the Hebrides shelf and Wyville-Thomson Ridge, NW UK continental shelf. Scottish Journal of Geology, Vol. 24, 249–262.

STOKER, M S. 1990. Glacially influenced sedimentation on the Hebridean slope, northwestern United Kingdom continental margin. 349–362 Special Publication of the Geological Society of London, No. 53.

STOKER, M S, and BENT, A J A. 1985. Middle Pleistocene glacial and glaciomarine sedimentation in the west central North Sea. Boreas, Vol. 14, 325–332.

STOKER, M S, and BENT, A J A. 1987. Lower Pleistocene deltaic and marine sediments in boreholes from the central North Sea. Journal of Quaternary Science, Vol. 2, 87–96.

STOKER, M S, and GRAHAM, C. 1985. Pre-Late Weichselian submerged rock platforms off Stonehaven. Scottish Journal of Geology, Vol. 21, 205–208.

STOKER, M S, and HOLMES, R. 1991. Submarine end-moraines as indicators of Pleistocene ice limits of NW Britain. Journal of the Geological Society of London, Vol. 148, 431–434.

STOKER, M S, SKINNER, A C, FYFE, J A, and LONG, D. 1983. Palaeomagnetic evidence for early Pleistocene in the central and northern North Sea. Nature, London, Vol. 304, 332–334.

STOKER, M S, LONG, D, and FYFE, J A. 1985. A revised Quaternary stratigraphy for the central North Sea. Report of the British Geological Survey, No. 17/2.

STOKER, M S, HITCHEN, K, and GRAHAM, C C. 1993. United Kingdom offshore regional report: the geology of the Hebrides and West Shetland shelves, and adjacent deepwater areas. (London: HMSO for the British Geological Survey.)

STOKER, M S, and eight others. 1994. A record of late Cenozoic stratigraphy, sedimentation and climate change from the Hebrides Slope, NE Atlantic Ocean. Journal of the Geological Society of London, Vol. 151, 235–249.

STOKER, M S, HOWE, J A, and STOKER, S J. 1999. Late Vendian–?Cambrian glacially influenced deep-water sedimentation, Macduff Slate Formation (Dalradian), N E Scotland. Journal of the Geological Society of London, Vol. 156, 55–61.

STUIVER, M, and REIMER, P J. 1993. Extended 14C data base and revised calib 3.0 14C age calibration program. Radiocarbon, Vol. 35, 215–230.

SUGDEN, D E. 1986. Pre-glacial weathered rock in till, Tyrebagger, Aberdeen and its implications. 397–401 in Essays for Professor R E H Mellor. RITCHIE, W, STONE, J C, and MATHER, A S (editors). (Aberdeen: University of Aberdeen.)

SUGDEN, D E. 1989. Modification of old land surfaces by ice sheets. Zeitschrift fur Geomorphologie N F Supplemunt-Bunt, Vol. 72, 163–172.

SUGDEN, D E, and JOHN, B S. 1976. Glaciers and landscape: a geomorphological approach. (London: Arnold.)

SUTHERLAND, D G. 1981. The high-level marine shell beds of Scotland and the build-up of the last Scottish ice sheet. Boreas, Vol. 10, 247–254.

SUTHERLAND, D G. 1984a. The Quaternary deposits and landforms of Scotland and the neighbouring shelves: a review. Quaternary Science Reviews, Vol. 3, 157–254.

SUTHERLAND, D G. 1984b. Gardenstown (Gamrie). 89–93 in Buchan field guide. HALL, A M (editor). (Cambridge: Quaternary Research Association.)

SUTHERLAND, D G. 1984c. King Edward. 93–96 in Buchan field guide. HALL, A M (editor). (Cambridge: Quaternary Research Association.)

SUTHERLAND, D G. 1993a. Teindland Quarry. 236–240 in The Quaternary of Scotland. GORDON, J E, and SUTHERLAND, D G (editors). Geological Conservation Review Series, No. 6 (London: Chapman and Hall.)

SUTHERLAND, D G. 1993b. Castle Hill. 240–242 in The Quaternary of Scotland. GORDON, J E, and SUTHERLAND, D G (editors). Geological Conservation Review Series, No. 6 (London: Chapman and Hall.)

SUTHERLAND, D G. 1999. Scotland. 99–114 in A revised correlation of Quaternary deposits in the British Isles. BOWEN, D Q, (editor). Geological Society of London Special Report, No. 23.

SUTHERLAND, D G, and GORDON, J E. 1993. The Quaternary in Scotland. 13–47 in The Quaternary of Scotland. GORDON, J E, and SUTHERLAND, D G (editors). Geological Conservation Review Series, No. 6 (London: Chapman and Hall.)

SYKES, G. 1991. Amino acid dating. 161–176 in Quaternary dating methods —a user’s guide. Technical Guide 4. SMART, P L, and FRANCES, P D (editors). (Cambridge: Quaternary Research Association.)

SYNGE, F M. 1956. The glaciation of north-east Scotland. Scottish Geographical Magazine, Vol. 72, 129–143.

SYNGE, F M. 1963. The Quaternary succession round Aberdeen, north-east Scotland. 353–361 in Report on the 6th International Congress on the Quaternary. (Warsaw, 1961: International Union for Quaternary Research, Lodz.)

THOMAS, G S P. 1984. Sedimentation of a sub-aqueous esker-delta at Strabathie, Aberdeenshire. Scottish Journal of Geology, Vol. 20, 9–20.

THOMAS, G S P, and CONNELL, R, J. 1985. Iceberg drop, dump, and grounding structures from Pleistocene glacio-lacustrine sediments, Scotland. Journal of Sedimentary Petrology, Vol. 55, 243–249.

THOMSON, K, and HILLIS, R R. 1995. Tertiary structuration and erosion of the inner Moray Firth. 249–269 in The tectonics, sedimentation and palaeoceanography of the North Atlantic Region. SCRUTTON, R A, STOKER, M S, SHIMMIELD, G B, and TUDHOPE, A W (editors). Geological Society of London Special Publication, No. 90.

TIPPING, R M, and MILBURN, P. 2000. Palaeoenvironmental evidence from peat beneath the floor of the Den of Boddam. 116–125 in The Quaternary of the Banffshire coast and Buchan: field guide. MERRITT, J W, CONNELL, E R, and BRIDGLAND, D R (editors). (London: Quaternary Research Association.)

TRAVERSE, A, and GINSBURG, R N. 1966. Palynology of the surface sediments of the Great Bahama Bank, as related to water movement and sedimentation. Marine Geology, Vol. 4, 417–459.

TURNER, C. 1970. The Middle Pleistocene deposits at Marks Tey, Essex. Philosophical Transactions of the Royal Society of London, Vol. B257, 373–440.

TURNER, W. 1870. On the bones of a seal found in red clay near Grangemouth, with remarks on the species. Proceedings of the Royal Society of Edinburgh, Vol. 7, 105–114.

VALEN, V, LARSEN, E, and MANGERUD, J. 1995. High resolution paleomagnetic correlation of Middle Weichselian ice-dammed lake sediments in two coastal caves, western Norway. Boreas, Vol. 24, 141–153.

VAN AMERONGEN, J C. 1976. The study of two soils in an area near Mintlaw, North-East Scotland. Unpublished MSc thesis, University of Aberdeen.

VAN DEN BROEK, J M M, and VAN DER WAALS, L. 1967. The late Tertiary peneplain of southern Limburg. Geologie en Mijnbouw, Vol. 46, 318–332.

VASARI, Y. 1977. Radiocarbon dating of the Late Glacial and Early Flandrian vegetational succession in the Scottish Highlands and the Isle of Skye. 143–163 in Studies in the Scottish Lateglacial environment. GRAY, J M, and LOWE, J J (editors). (Oxford: Pergamon Press.)

VASARI, Y, and VASARI, A. 1968. Late- and post-glacial macrophytic vegetation in the lochs of northern Scotland. Acta Botanica Fennica, Vol. 80, 1–120.

WALKER, A B. 1998. UK earthquake monitoring 1997/1998: British Geological Survey Seismic Monitoring and Information Service Nineth Annual Report. British Geological Survey Technical Report, WL/98/03.

WALKER, A D. 1973. The age of the Cuttie’s Hillock Sandstone (Permo–Triassic) of the Elgin area. Scottish Journal of Geology, Vol. 9, 177–183.

WALKER, A D, and six others. 1982. Soil and land capability for agriculture: Eastern Scotland. (Handbook of the Soil Survey of Scotland, No. 5). (Aberdeen: Macauley Institute for Soil Research.)

WALKER, M J C, and six others. 1992. Allt Odhar and Dalcharn: two pre-Late Devensian (Late Weichselian) sites in northern Scotland. Journal of Quaternary Science, Vol. 7, 69–86.

WALKER, M J C, and six others. 1999. Isotopic ‘events’ in the GRIP ice core: a stratotype for the Late Pleistocene. Quaternary Science Reviews, Vol. 18, 1143–1150.

WALTON, K. 1956. Rattray: a study in coastal evolution. Scottish Geographical Magazine, Vol. 72, 85–96.

WALTON, K. 1959. Ancient elements in the coastline of northeast Scotland. 93–109 in Geographical essays in memory of Alan G Ogilvy, Edinburgh. MILLER, R, and WATSON, J W (editors). (London: Nelson).

WHITTAKER, A. and eleven others. 1991. A guide to stratigraphical procedure. Journal of the Geological Society, London, Vol. 148, 813–824.

WHITTINGTON, G. 1994. Bruckenthalia spiculfolia (Salisb.) Reichenb. (Ericaceae) in the Late Quaternary of western Europe. Quaternary Science Reviews, Vol. 13, 761–768.

WHITTINGTON, G, HALL, A M, and JARVIS, J. 1993. A pre-Late Devensian pollen site from Camp Fauld, Buchan, north-east Scotland. New Phytologist, Vol. 125, 867–874.

WHITTINGTON, G, and six others. 1998. Devensian organic interstadial deposits and ice sheet extent in Buchan, Scotland. Journal of Quaternary Science, Vol. 13, 309–324.

WILKINSON, I P. 1993. Calcareous micropalaeontological analysis of the Quaternary ‘shelly till’ at Burn of Benholm. British Geological Survey Technical Report, WH93/232R.

WILKINSON, I P. 1998. Pleistocene Microfossils from Boyne Bay, Banffshire. British Geological Survey Technical Report, WH97/224R.

WILSON, J S G. 1882. Northern Aberdeenshire and eastern Banffshire. Memoirs of the Geological Survey, Sheet 97 (Scotland).

WILSON, J S G. 1886. North-east Aberdeenshire with detached portions of Banffshire. Memoirs of the Geological Survey, Scotland, Sheet 87.

WILSON, J S G, and HINXMAN, L W. 1890. Geology of central Aberdeenshire. Memoir of the Geological Survey of Scotland, Sheet 76 (Scotland).

WILSON, M J, and TAIT, J M. 1977. Halloysite in some soils from north-east Scotland. Clay Minerals, Vol. 12, 59–66.

WILSON, M J, BAIN, D C, and MITCHELL, W A. 1968. Saponite from the Dalradian metalimestones of north-east Scotland. Clay Minerals, Vol. 7, 343–349.

WILSON, M J, and five others. 1981. A swelling hematite/layer sillicate complex in weathered granite. Clays and Clay Minerals, Vol. 16, 261–278.

WILSON, M J, RUSSELL, J D, TAIT, J M, CLARK, D R, and FRASER, A R. 1984. Macaulayite, a new mineral from north-east Scotland. Mineralogical Magazine, Vol. 48, 127–129.

WORSLEY, P. 1991. Possible early Devensian glacial deposits in the British Isles. 47–51 in Glacial deposits in Great Britain and Ireland. EHLERS, J, GIBBARD, P L, and ROSE, J (editors). (Rotterdam: A A Balkema.)

WRIGHT, J S. 1997. Deep weathering profiles (saprolites) in north east Scotland. Scottish Geographical Magazine, Vol. 113, 189–194.

ZENKEVITCH, L. 1963. Biology of the seas of the U.S.S.R. (London: Allen and Unwin.)

ZIEGLER, P A. 1981. Evolution of sedimentary basins of north-west Europe. (London: Heyden.)

Figures, plates, tables and maps

Figures

(Figure 1a) Ground surface created from a digital terrain model made by the Centre for Ecology and Hydrogeology, Wallingford, from Ordnance Survey panorama data.

(Figure 1b) Topography of the district and locality map showing the boundaries of the eleven geological sheets. Additional topographical details are given on maps 1 to 11.

(Figure 2) Solid geology of the district.

(Figure 3) Models of the British Main Late Devensian ice sheet at its maximum extent. a Boulton et al. (1977) b Boulton et al. (1985) c Boulton et al.(1991) d Lambeck (1991) e Jansen (1976) f Ehlers and Wingfield (1991)

(Figure 4) Generalised flow-lines of ice during the Main Late Devensian glaciation and profile map of the five groups of glacigenic deposits mapped in north-east Scotland.

(Figure 5) Key to the most common morpho-lithogenetic symbols used on recently published drift maps of the district.

(Figure 6) British and north-west European chronostratigraphy (after Gordon, 1997; Bowen et al., 1999) and a representative oxygen isotope and geomagnetic polarity record (ODP 677) (after Shackleton et al., 1990; Bowen et al., 1999). See also (Figure 33)

(Figure 7) The ‘SPECMAP’ oxygen isotope curve for the last glacial–interglacial cycle (after Imbrie et al., 1984) with climato-stratigraphical stages and selected events in north-east Scotland.

(Figure 8) Major sand and gravel workings in north-east Scotland (1999).

(Figure 9) Principal former brick-clay localities in north-east Scotland.

(Figure 10) Shallow groundwater occurrence.

(Figure 11) Grain size plots for four of the drift groups in north-east Scotland.

(Figure 12) Plots of Plasticity Index versus Liquid Limit for four drift groups in north-east Scotland.

(Figure 13) Histograms of Liquid Index of four drift groups in north-east Scotland.

(Figure 14) Histograms of Undrained Shear Strength for the Banffshire Coast and Logie-

(Figure 15) Erosion surfaces in north-east Scotland (after Hall, 1986).

(Figure 16) Devonian unroofing and post-Devonian erosion of the Bennachie granite mass (after Hall, 1987).

(Figure 17) Summary of the main geomorphological events in the shaping of north-east Scotland (after Hall, 1987). HT humid tropical; A arid; ST subtropical; Te temperate; K kaolinitic; S sandy

(Figure 18) Palaeogene drainage and sediment transport (after Hall, 1991).

(Figure 19) Neogene drainage patterns (after Hall, 1991).

(Figure 20) Summary of Cainozoic relief development in central Buchan (after Hall, 1987).

(Figure 21) A comparison of the patterns of weathering and glacial erosion in eastern Aberdeenshire (after Hall and Sugden, 1987).

(Figure 22) Distribution of Miocene land surface and major topographical features (after Hall, 1985).

(Figure 23) Deep weathering sites in north-east Scotland (after Hall, 1986).

(Figure 24) Slope development and the progressive removal of different types of weathering profile (after Hall, 1983).

(Figure 25) Vertical changes in a gruss weathering profile in quartz-biotite psammite, Northseat [NJ 930 408] (after Hall, 1986).

(Figure 26) Scarp-foot weathering zones and their progressive exploitation (after Hall, 1986).

(Figure 27) Regional weathering patterns (after Hall, 1986). ZONE 1 Deep and continuous saprolites with few fresh outcrops. Saprolite thicknesses generally are at least 3 m and commonly exceed 10 m. Fresh outcrops are confined to highly resistant rocks ZONE 2 Thinner and discontinuous saprolites with fresh outcrops. Pockets of deep alteration remain but saprolites are generally less than 3 m thick. Fresh outcrops are common, especially on steep slopes and along valley floors ZONE 3 Fresh rocks with numerous pockets of weathering. Fresh rocks underlie most of the area. Weathering generally is found only in ice-lee locations on interfluves and along valleys, and is common only in tributary valleys and basins ZONE 4 Fresh rocks with rare pockets of weathering. Fresh rock predominates and local pockets of weathering usually relate to fracture zones. A distinction is made between areas of gentle slopes and/or rocks of low to moderate resistance in which weathering was probably extensive prior to glaciation (Zone 4a) and areas of steep slopes and/or resistant rocks in which weathering was probably thinly or sporadically developed prior to glaciation (Zone 4b)

(Figure 28) Granulometry of weathered rocks.

(Figure 29) Oxygen isotope curve (Pacific Site 849) representing ice volume change over the past 1.2 million years and a curve predicting summer solar insolation at 65°N (after Raymo, 1997).

(Figure 30) Major thermohaline circulation cells that make up the global conveyor system (after Skinner and Porter, 1995). The cells are driven by exchange of heat and moisture between the atmosphere and ocean.

(Figure 31) Greenland (GRIP Summit) oxygen isotope record (after Dansgaard et al.,1993) and predicted former ice cover in Scotland (after Clapperton,1997). The GRIP timescale was determined by counting annual ice layers back to 14.5 ka, and beyond by estimation based on ice-flow modelling. Peaks in ice rafted detritus (Heinrich events H0-H6) are placed after Bond et al. (1993).

(Figure 32) Proxy climate record of the last termination based on a graph of Late-glacial snow accumulation on the Greenland ice sheet summit (after Alley et al., 1993; Clapperton, 1997). GS-1 Greenland Stadial 1; GI-1 Greenland Interstadial 1; GS-2 Greenland Stadial 2 (after Walker et al., 1999). H0 and H1 are approximate positions of Heinrich events so-named.

(Figure 33) Quaternary stratigraphy of the central North Sea (from Gatliff et al., 1994).

(Figure 34) General stratigraphical relationships of Quaternary formations in the central North Sea north of 56°N (from Gatliff et al., 1994).

(Figure 36) Patterns of ice flow deduced by Jamieson (1906). a Deposition of Indigo Till b Deposition of Lower Grey Boulder Clay c Deposition of Dark Blue Boulder Clay and Red Clay Series d Aberdeen Re-advance

(Figure 35) Generalised sequence of Quaternary deposits in the Moray Firth (from BGS Moray–Buchan Sheet 57°N–04°W). Ages uncertain, possibly spanning Middle Devensian to early Flandrian.

(Figure 36) Patterns of ice flow deduced by Jamieson (1906). a Deposition of Indigo Till b Deposition of Lower Grey Boulder Clay c Deposition of Dark Blue Boulder Clay and Red Clay Series d Aberdeen Re-advance

(Figure 37) Patterns of ice flow deduced by Bremner (1943).

(Figure 38) Patterns of ice flow deduced by Synge (1956) (after Clapperton and Sugden, 1977). a Greater Highland Glaciation b Moray Firth-Strathmore Glaciation c Aberdeen Re-advance d Dinnet Re-advance

(Figure 39) Pattern of glacial re-advances following the Main Late Devensian maximum, when Scottish and Scandinavian ice sheets coalesced, as deduced by Sissons (1967).

(Figure 40) Pattern of ice flow deduced by Clapperton and Sugden (1977).

(Figure 41) Reconstruction of the maximum extent of the Main Late Devensian–Weichselian ice sheets (28–22 ka BP) showing possible location of former ice streams (modified from Sejrup et al., 2000).

(Figure 42) Tentative reconstructions of former proglacial lakes in north-east Scotland. a Creation of ‘50 m’ Lake Ugie shortly after the maximum of the second major expansion of the Main Late Devensian ice sheet (after 18 ka BP) and following the earlier ponding of the 80 m’ lake. Widespread glacial over-riding of glaciolacustrine deposits occured east of Lake Ugie. b Parting of East Grampian and ‘Logie-Buchan’ ice with the formation of Lake Ythan. c Diachronous ponding along the Banffshire coast and glaciomarine incursion around St Fergus at about 15 ka BP

(Figure 43) Devensian–Weichselian events in Britain and south-west Fennoscandia (after Peacock and Merritt, 1997; Sejrup et al., 2000). Norwegian data from Baumann et al. (1995) and Mangerud et al. (1981). D/L amino-acid ratios corrected according to Miller and Mangerud (1985). British data from Baker et al. (1995), Bowen (1989), Gordon et al. (1989), Lawson and Atkinson (1995) and Miller et al. (1987).

(Figure 44) Tentative reconstruction of ice margins at the maximum stage of the second major expansion of the Main Late Devensian ice sheet (after Hall and Bent, 1990 and Sejrup et al., 1987). This stage is correlated with the maximum of the ‘Dimlington Advance’, 18.5–15.1 ka BP (Sejrup et al., 1994).

(Figure 45) Map showing the transport paths of some indicator erratics across the southern part of the Moray Firth (after Mackie, 1905; Sissons, 1967).

(Figure 46) Contoured rockhead surface beneath the City of Aberdeen (after Munro, 1986).

(Figure 47) Transport paths of some indicator erratics in north-east Scotland (after Read, 1923; Bremner, 1928; Synge, 1956; Sissons, 1967; Ehlers, 1988).

(Figure 48) Inferred sea-level change over the past 15 000 years, based on height-age relationships of raised shorelines. Curves a and c from Lambeck, 1993a; b from Smith et al., 1999 with speculative Late-glacial part of curve extrapolated from St Fergus.

(Figure 49) Isobases for the Main Postglacial Shoreline in eastern Scotland (after Cullingford et al., 1991).

(Figure 50) Distribution of the Logie-Buchan Drift Group and related features.

(Figure 51) Coverage of large-scale Quaternary geological maps of the district.

(Figure 52) Distribution of site investigation records held by BGS in 1999.

(Figure A1.1) Locations of sites described in Appendix 1.

(Figure A1.2) Location of sites in the vicinity of Teindland (after Hall et al., 1995a).

(Figure A1.3) Lithostratigraphy at Red Burn (after Hall et al., 1995a).

(Figure A1.4) The original Teindland pollen site (after Hall et al., 1995a).

(Figure A1.5) The location of the Boyne Limestone Quarry (after Peacock and Merritt, 2000a).

(Figure A1.6) The glacigenic sequence at locality 1, Boyne Limestone Quarry (after Peacock and Merritt, 2000a).

(Figure A1.7) Map of Castle Hill, Gardenstown (after Peacock and Merritt, 1997).

(Figure A1.8) Lithological log of Section S1 at Castle Hill (after Peacock and Merritt, 1997).

(Figure A1.9) Crossbrae Farm site (after Whittington et al., 1998).

(Figure A1.10) Graphic logs and lithostratigraphy of representative sections at the Howe of Byth sand and gravel quarry (after Hall et al., 1995b).

(Figure A1.11) Generalised lithological sequences at Kirkhill and Leys quarries.

(Figure A1.12) Generalised transect along the lower Philorth valley showing radiocarbon and pollen sites used in the construction of the local Holocene sea level curve (after Smith et al., 1982).

(Figure A1.13) Landforms and deposits in and adjacent to the Ugie catchment. A, Ardglassie; B, Blackhills; BB, Ballus Bridge; G, Greenhill; K, Kirkhill; L, Lumbs; Ly, Leys; O, Oldmill; R, Red Loch; SL, Sinclair Hills; W, Upper Waughtonhill; WD, West Dens channel.

(Figure A1.14) Glaciofluvial deposits and landforms in the vicinity of Auchlee, north of Longside. A-C are localities where fossil tundra polygons are visible on air photos; D-G and I sites mentioned in the text.

(Figure A1.15) Localities at St Fergus (after Peacock, 1999). See (Figure A1.13) for general location of the St Fergus Silt Formation.

(Figure A1.16) Schematic stratigraphical relationships in the Sandford Bay area, based on cliff sections near Sandford Lodge [NK 125 434] and in the Invernettie road diversion [NK 120 447].

(Figure A1.17) The principle outcrops of the Windy Hills Gravel Member of the Buchan Gravels Formation near Fyvie.

(Figure A1.18) Investigations on the Buchan Ridge (after Hall and Jarvis, 1994).

(Figure A1.19) Camp Fauld site (after Whittington et al., 1993).

(Figure A1.20) Schematic logs and correlations in the Ellon area (after Hall and Jarvis, 1995).

(Figure A1.21) Quaternary deposits and landforms of the Kippet Hills area.

(Figure A1.22) Mill of Dyce sand and gravel quarry and the deglaciation of the lower Don valley.

(Figure A1.23) A palaeoenvironmental model of the Strabathie subaqueous esker-delta (after Thomas, 1984).

(Figure A1.24) Iceberg dump and grounding structures at Strabathie (after Thomas, 1984). a–d dump structures; e grounding structure

(Figure A1.25) Lithostratigraphy at the Nether Daugh site.

(Figure A1.26) Absolute pollen diagram of selected taxa from Rothens (after Aitken, 1991).

(Figure A1.27) Stratigraphical relationships of deposits at the Nigg Bay cliff section in 1977–80 (not to scale and partially schematic).

(Figure A1.28) Loch of Park locality, near Banchory.

(Figure A1.29) Stratigraphy and structure of the Quaternary sediments at Balnakettle.

(Figure A1.30) Location of Knockhill Wood site.

(Figure A1.31) Sketch transect of the Knockhill Wood exposure showing the relationship between the organic and glacigenic sediments.

(Figure A1.32) Graphic log of the organic sequence at Knockhill Wood.

(Figure A1.33) Postulated form of the landslip Knockhill Wood.

(Figure A1.34) Excavations and measured sections at the Burn of Benholm.

(Figure A2.1) Sketch map showing the location of published sand and gravel resource assessment sheets and administrative areas in north-east Scotland.

(Figure A2.2) Composition of workable gravel deposits between Elgin and Aberdeen.

(Figure A2.3) Composition of workable gravel deposits south of Aberdeen.

(Figure A3.1) Houff or Ury area. a Contoured ground conductivity values and location of GPR traverses in the Houff of Ury area b Geological map of the contact between the East Grampian and Mearns drift groups in the Houff of Ury area, based on geological mapping, conductivity surveying and ground resistivity soundings

(Figure A3.2) Frequency distribution of interpreted resistivity values from the Banchory–Stonehaven sand and gravel resource sheet area (Auton et al., 1988).

(Figure A3.3) Buchan Ridge. a Location of GPR traverses and resistivity soundings on the northern slope of the ‘Buchan Ridge’ at the Moss of Cruden b Interpreted GPR profile (50mHz) along part of line 1 (162–342 m)

(Figure A3.3c) Buchan Ridge. Resistivity model along line 1 ( 200-392 m) and line 2

Plates

(Plate 1) Bennachie (D4331).View of Bennachie taken from the north-east showing the two tops, Mither Tap (left) and Craigshannoch (right), capped by tors formed of the Bennachie Granite. The lower ground is underlain by sandy till of the East Grampian Drift Group resting on norites of the Insch Basic Intrusions.

(Plate 2) Mill of Dyce sand and gravel pit in glaciofluvial deposits on the southern side of the valley of the River Don north-west of Aberdeen (D4795A).

(Plate 3) Back scar of a landslip in the Kirk Burn Silt Formation near Macduff (P104100).

(Plate 4a) Buchan Gravels Formation at Windyhills. a Quartz-quartzite gravel of the Windy Hills Gravel Member at its type locality (P104101)

(Plate 4b) Buchan Gravels Formation at Windyhills. b Cryoturbated till with vertically aligned pebbles capping the Windy Hills Gravel Member (P104102) Scale: hammer is 32 cms long

(Plate 5a) Flint clasts recovered from the base of the Buchan Ridge Gravel Member in a trial pit on the Moss of Cruden [NK 0253 4024]. Little-worn flints such as these were either derived directly from a former cover of Chalk in the area or from a remanié flint deposit. This evidence suggests that the present terrain lies close to the level of the sub-Cenomanian surface. (a P104122, b P104121).

(Plate 5b) Flint clasts recovered from the base of the Buchan Ridge Gravel Member in a trial pit on the Moss of Cruden [NK 0253 4024]. Little-worn flints such as these were either derived directly from a former cover of Chalk in the area or from a remanié flint deposit. This evidence suggests that the present terrain lies close to the level of the sub-Cenomanian surface. (a P104122, b P104121).

(Plate 6) Weathered megacrystic Crathes Granite in Littletown quarry [NJ 6988 0969], west of Dunecht (P104104). Scale: hammer is 40 cms long.

(Plate 7) Corestones developed in weathered gabbroic rock near Maud [NJ 8897 5053]. (P104105). Scale: spade is 0.9 m long.

(Plate 8a) Till. a Subglacial till of the Banchory Till Formation containing angular cobbles of granite and Dalradian metamorphic rocks, exposed in the bank of the Cowie Water [NO 8421 8845] (Z00410). Scale: compass shown in top right

(Plate 8b) Till. b Mill of Forest Till containing rounded cobbles of Devonian rocks at its type locality [NO 8620 8540] (Z00411). Scale: lens cap in centre

(Plate 8c) Till. c Till of the Banffshire Coast Drift Group at Oldmill [NK 0245 4407], showing downward transition from weathered to unweathered diamicton. (P104106)

(Plate 8d) Till. d Till of the Banffshire Coast Drift Group overlying cryoturbated gravel with vertical clasts [NK 0130 5491], near Kirkhill. (P104107). Section shown is about 1.5 m high.

(Plate 9) Hummocky glacial deposits: an extremely poorly sorted, heterogeneous morainic deposit at Blairdaff [NJ 7001 1797], near Kemnay. The deposit comprises bouldery gravel and sandy diamicton that was laid down as debris flows at the margin of the receding East Grampian ice sheet. (P104108). Scale: spade is 0.9 m long.

( Plate 10a) Glaciofluvial sand and gravel. a Deltaic foresets dipping eastwards, capped by poorly stratified topset gravel at Lochinvar pit [NJ 185 613], near Elgin (P104109). Scale: spade is 0.9 m long

(Plate 10b) Glaciofluvial sand and gravel. b Climbing ripple-drift cross-lamination in deltaic sands at Kirkmyres pit [NJ 926 609], near Fraserburgh. (P104110). Face shown is about 2 m high

(Plate 10c) Glaciofluvial sand and gravel. c Channel-fill cross-bedded gravels forming glaciofluvial sheet deposits in Danshillock pit, south of Macduff (D6152). Hammer is 40 cms long

(Plate 10d) Glaciofluvial sand and gravel. d Topset gravels overlying sandy deltaic foresets in the Lochton Sand and Gravel Formation, Cammie Wood sand and gravel pit, Feughside (D4987)

(Plate 11) Convolute lamination in tectonised glaciolacustrine silts from the Blackhills Sand and Gravel Formation, Tipperty sand pit, south-west of Banff. (D6140). Hammer is 32 cm long.

(Plate 12a) Head deposits. a Rubbly clast-supported diamicton formed mainly of angular fragments of local frost-shattered metagreywacke and more far-travelled rounded clasts of metaquartzite [NJ 8214 5122]. Such deposits are widespread in central Buchan, typically about 0.5 m thick and are partially matrix-supported (P104111)

(Plate 12b) Head deposits. b A gravelly head deposit with periglacial festoon structures overlying, and locally penetrating into, cryoturbated grussified granite in a pit near Kemnay [NJ 7502 1377]. The gravel locally truncates ice-wedge casts in the decomposed granite that are filled with weathered sandy diamicton of a pre-Late Devensian glaciation (P104112)

(Plate 13) Glacial sculpturing of Old Red Sandstone bedrock, Hill of Findon, south of Gardenstown (D6142).

(Plate 14a) Glacial drainage channels. a Blackhills subglacial drainage channel cut into Dalradian psammitic rocks of the Glen Lethnot Grit Formation, north of Stonehaven (P104113)

(Plate 14b) Glacial drainage channels. b Ice-marginal channels formed at the retreating margin of the Moray Firth ice stream of the Main Late Devensian ice sheet in the valley of the Burn of Fishrie [NJ 775 579], looking southwards. (P104114)

(Plate 15) The Kippet Hills Esker looking northwards from Broom Hill [NK 0332 3058], near Collieston. Meikle Loch occupies a large kettle hole beside the esker, which is capped by red clayey diamicton typical of the Logie-Buchan Drift Group. The esker merges northwards into an outwash fan at Knapsleask (D2791).

(Plate 16a) Glacial indicator erratics. a Rhomb porphyry erratic from the Oslo Graben, Norway, identified in 1979 by Dr J A Dons, Mineralogisk-Geologisk Museum, Oslo. The erratic was found in the Sandford Bay Till Member of the Hythie Till Formation at Sandford Bay [NK 1245 4348], south of Peterhead (P015377)

(Plate 16b) Glacial indicator erratics. b Tabular erratic of sillimanite-bearing pelitic hornfels from an erratic train to the west of Monymusk (P104115)

(Plate 17a) Glacial rafts. a Tectonic lamination in a raft of sand (possibly glaciomarine) disrupted by lowangle shears at Ardglassie, near Fraserburgh (P104123). Spade is 0.9 m long

(Plate 17b) Glacial rafts. b Attenuated, fractured and sheared lenses of sand within the Whitehills Glacigenic Formation at the Boyne Limestone Quarry (P104124).

(Plate 18) Glaciofluvial terrace gravel infilling an ice-wedge cast at Inverurie [NJ 7678 2255]. The structure has developed in glaciolacustrine silty sands and sandy diamicton, both derived mostly from local decomposed mica schist (P104116).

(Plate 19) Earth-pillar in the valley of the Allt Dearg, near Fochabers, formed in weathered Devonian conglomerate (Spey Conglomerate Formation) at the Craigs of Cuildell (D2144).

(Plate 20) Postglacial raised beach deposits bordering Spey Bay, looking westwards towards Binn Hill. The long, parallel ridges of beach shingle reach 7 m above OD (MNS 2113).

(Plate 21) Raised wave-cut rock platform at Whyntie Head [NJ 626 658], near Portsoy, looking westwards. The platform was cut prior to the deposition of shelly sands of the Whitehills Glacigenic Formation, which overlie it locally (P104117).

(Plate 22) Southern face of Kirkhill Quarry [NK 011 528] in 1979 (P528224). The sands and gravels resting on bedrock belong to the Pitscow Sand and Gravel Formation. The conspicuous pale band at their surface is the podzolic Kirkhill Palaeosol Bed. Periglacial sediments of the Swineden Sand Bed and the Camphill Gelifluctate Bed succeed this. The base of the Rottenhill Till Formation is at shoulder level with the figure. It is weathered towards the top (Fernieslack Palaeosol Bed) and is overlain by the pale coloured Corsend Gelifuctate Bed. The overlying Hythie Till Formation has a sharp erosional base. Note, the deformation structure affecting sediments beneath the till (above and to the left of the figure). Approximately 8 m of sediments are exposed.

(Plate 23) Partially glacitectonised, podzolic Teindland Palaeosol Bed below sandy diamicton of the Woodside Diamicton Formation at Teindland quarry [NJ 297 570]. Luminescence dates of 79 ± 6 ka and 67 ± 5 ka have recently been obtained from the sands overlying the palaeosol bed to the left of the scale card. (P104118). Scale bar showing 1 cm intervals.

(Plate 24) Attenuated rafts of sand and brecciated dark grey pebbly clay within the Whitehills Glacigenic Formation at Oldmill pit [NK 0243 4389]. Greyish brown diamicton of the Banffshire Coast Drift Group caps the section above a metre-thick unit of grey, sheared sandy diamicton (a penetrative glacitectonite). The ice that overrode the sequence flowed towards the left (east-south-east) (P104119).

(Plate 25) Brittle star (Ophiocten sericium) in glaciomarine clay of the Spynie Clay Formation. Collected from the former brick pit at Spynie [NJ 232 672], near Elgin (Royal Museum of Scotland specimen 1913.22.1 GN 035).

(Plate 26) Glacitectonised sequence at Balnakettle, north-east of Fettercairn. Till and gravel of the Mearns Drift Group are tectonically intercalated with older (pre- Late Devensian) head gravel composed of angular clasts derived from local Dalradian pelitic and semipelitic bedrock. The top of the tectonised sequence is truncated by a till of the East Grampians Drift Group, formed during a local re-advance (P100707).

(Plate 27) Sheared lens of the Burn of Benholm Peat Bed within olive grey shelly clay and diamicton (Benholm Clay Formation) in trial pit BBP4, at the Burn of Benholm site north of Johnshaven (Appendix A, 26). The shelly deposits are overlain by red-brown diamicton of the Mill of Forest Till Formation (D4879).

(Front cover) Cover photograph: Crovie Village (D6116). Sandstones and conglomerates of the Lower Devonian Crovie Group crop out on the foreshore and are down-faulted against Dalradian rocks of the Macduff Slate Formation that form the headland at the far end of the village. The landslip scar beyond the red telephone box exposes fine-grained deposits of the Kirk Burn Silt Formation of the Banffshire Coast Drift Group.

(Rear cover)

Tables

(Table 1) Summary of events preserved in the marine and onshore records in and around north-east Scotland (after Holmes, 1977).

(Table 2) Division of the Cainozoic.

(Table 3) Synonyms used on BGS drift maps of the district.

(Table 4) Brick clay deposits in north-east Scotland.

(Table 5) Potential peat resources between Aberdeen and Elgin (based on Fraser, 1948). Buchan drift groups.

(Table 6) Commonly occurring types of till and related sediments.

(Table 7) Correlation of lithostratigraphical units in north-east Scotland.

(Table 8) Radiocarbon dates from Late-glacial sites in the district.

(Table 9) Geological surveyors.

(Table 10) Survey history of geological maps.

(Table 11) Number of boreholes on each geological sheet.

(Table A1.1) Sites described in Appendix 1.

(Table A1.2) Lithostratigraphy of the Teindland area.

(Table A1.3) The lithostratigraphy at Castle Hill, Gardenstown.

(Table A1.4) Mollusca from shelly deposits of the Whitehills Glacigenic Formation.

(Table A1.5) Lithostratigraphy at the Howe of Byth.

(Table A1.6) Lithostratigraphical units at Kirkhill and Leys Quarries.

(Table A1.7) Radiocarbon dates from sites in the Philorth valley (after Smit

(Table A1.8) Lithostratigraphy at Camp Fauld.

(Table A1.9) Lithostratigraphy in the vicinity of Ellon, Leask and Slains.

(Table A1.10) Lithofacies associations at Mill of Dyce Pit.

(Table A1.11) Sedimentary facies at Strabathie (after Thomas,1984).

(Table A1.12) Pollen count from the Nether Daugh Pits.

(Table A1.13) Coleoptera from Nether Daugh (Pit 2).

(Table A1.14) Lithostratigraphy at Nigg Bay and in the vicinity of Aberdeen.

(Table A1.15) Stratigraphy of the Knockhill Wood site.

(Table A1.16) Outline pollen count from Knockhill Wood.

(Table A1.17) Lithostratigraphy in the vicinity of the Burn of Benholm.

(Table A2.1) Sand and gravel resource assessment surveys in north-east Scotland: primary data sources.

(Table A2.2) Summary results of mechanical and physical testing of naturally occurring gravels from north-east Scotland.

Maps

(Map 1) Glacial and glaciofluvial features and the distribution of tills in the Elgin district (after Peacock et al., 1968).

(Map 2) Glacial and glaciofluvial features and the distribution of glacigenic deposits on Sheet 96W Portsoy.

(Map 3) Glacial and glaciofluvial features and the distribution of glacigenic deposits on Sheet 96E Banff.

(Map 4) Glacial and glaciofluvial features and the distribution of glacigenic deposits on Sheet 97 Fraserburgh.

(Map 5) Glacial and glaciofluvial features and the distribution of glacigenic deposits on Sheet 86E Turriff.

(Map 6) Glacial and glaciofluvial features and the distribution of glacigenic deposits on Sheet 87W Ellon.

(Map 7) Glacial and glaciofluvial features and the distribution of glacigenic deposits on Sheet 87E Peterhead.

(Map 8) Glacial and glaciofluvial features and the distribution of glacigenic deposits on Sheet 76E Inverurie.

(Map 9) Glacial and glaciofluvial features and the distribution of glacigenic deposits on Sheet 77 Aberdeen.

(Map 10) Glacial and glaciofluvial features and the distribution of glacigenic deposits on Sheet 66E Banchory.

(Map 11) Glacial and glaciofluvial features and the distribution of glacigenic deposits on Sheet 67 Stonehaven. Frontispiece This winter image of north-east Scotland was acquired by the Landsat Enhanced Thematic Mapper sensor from an altitude of about 700 km, in December 2001. The image is a false colour composite, with bands 7, 4 and 1 displayed in red, green and blue. Bands 7 and 4 record infrared solar radiation reflected by the land surface, and highlight rocks/soil and vegetation, respectively, while band 1 records reflected visible blue light. Pink areas represent areas of snow and bright greens represent active vegetation. Coniferous forests are displayed as dark greens and moorland appears brown. Bare fields, rock and cultural features are seen as tones of red, while lowland areas appear in blue hues. Water is black due to absorption of solar radiation by the spectral ranges displayed. The Black Burn and Blackhills glacial drainage channels are conspicuous in the vicinity of Elgin, as are the channels associated with the Tore of Troup, east of Banff. The former inselberg of Mormond Hill dominates the low ground between Fraserburgh and Peterhead and the Insch and Alford basins show up particularly well.

Tables

(Table 2) Division of the Cainozoic

(Table 3) Synonyms used on BGS drift maps of the district

(Table 4) Brick clay deposits in north-east Scotland

(Table 5) Potential peat resources between Aberdeen and Elgin (based on Fraser, 1948).

(Table 6) Commonly occuring types of till and related sediments

(Table 8) Radiocarbon dates from Late-glacial sites in the district

(Table 9) Geological surveyors

(Table 10)  Survey history of geological maps

(Table 11) number of boreholes on each sheet

(Table A1.1) Sites described in Appendix 1

(Table A1.2) Lithostratigraphy of the Teindland area.

(Table A1.3) The lithostratigraphy at Castle Hill, Gardenstown.

(Table A1.4) Mollusca from shelly deposits of the Whitehills Glacigenic Formation.

(Table A1.5) Lithostratigraphy at the Howe of Byth.

(Table A1.6) Lithostratigraphical units at Kirkhill and Leys Quarries.

(Table A1.7) Radiocarbon dates from sites in the Philorth Valley (after Smith et al., 1982)

(Table A1.8) Lithostratigraphy at Camp Fauld

(Table A1.9) Lithostratigraphy in the vicinity of Ellon, Leask and Slains.

(Table A1.10) Lithofacies associations at Mill of Dyce Pit.

(Table A1.11) Sedimentary facies at Strabathie (after Thomas,1984).

(Table A1.12) Pollen count from the Nether Daugh Pits.

(Table A1.13) Coleoptera from Nether Daugh (Pit 2).

(Table A1.14) Lithostratigraphy at Nigg Bay and in the vicinity of Aberdeen.

(Table A1.15) Stratigraphy of the Knockhill Wood site.

(Table A1.16) Outline pollen count from Knockhill Wood.

(Table A1.17) Lithostratigraphy in the vicinity of the Burn of Benholm.

(Table A2.1) Sand and gravel resource assessment surveys in north-east Scotland: primary data sources.

(Table A2.2) Summary results of mechanical and physical testing of naturally occurring gravels from north-east Scotland.