Abstract This publication on glaciofluvial and glacio-lacustrine sedimentation is the outgrowth of a symposium on “Glacial Sedimentology” that was sponsored by the Eastern Section of S.E.P.M. and held at the joint meeting of the Northeastern Section of the Geological Society of America and the Eastern Section of S.E.P.M. at Buffalo, New York, March 10, 1972. Credit for suggesting a symposium on this topic must go to Frank W. Beales and Roger G. Walker, then respectively the President and Secretary of the Eastern Section of S.E.P.M. The symposium was chaired jointly by Alan V. Jopling and Barrie C. McDonald. The papers by Banerjee and McDonald, Saun-derson, Rust and Romanelli, Boothroyd and Ashley, Rust, and Gustavson, Ashley, and Boothroyd were presented in somewhat their present form at the symposium. The rest of the papers in the present collection were solicited from their respective authors. In an age when scientific publications are proliferating at a seemingly explosive rate, the question could arise whether a special publication should be devoted to glaciofluvial and gla-ciolacustrine sedimentation. It is apparent that there are a number of valid reasons why this particular topic should be treated as a separate entity and so become the focal point of a special volume. From a historical point of view, as noted in the review paper by Jopling (this vol.), the great interest generated in glacial phenomena during the nineteenth century had important implications and repercussions for the infant field of sedimentology. Without a doubt it provided its fair share of the

Abstract The purpose of this paper is to document the historical development of ideas on the interpretation of stratified (washed) drift of Quaternary age. The paper also purports to analyse the extent to which the roots of modern (twentieth century) sedimentology were influenced by glaciofluvial and glaciolacustrine studies of the nineteenth century. In the eighteenth century and early nineteenth century the occurrence of stratified drift was commonly ascribed to the action of natural catastrophies, and especially to the Biblical Deluge (the diluvial theory). Those who subsequently accepted Lyell's drift theory in the 1830's ascribed stratified drift to the washing action of glacial currents during a great submergence. When the drift theory was supplanted by the glacial (land-ice) theory ca. 1840-1850, the modern explanation for the origin of stratified drift was forthcoming. In all three theories, however, the stratified nature of the drift was attributed to water action; the specific agent, alias flood, marine submergence, or land ice, was the crux of the dispute. During the nineteenth and early twentierch centuries the sedimentological study of stratified drift was peripheral to the central theme of regional glaciation. Interpretative efforts were directed towards the spatial and chronologic aspects of glacial action and glacial causation; the approach was essentially morpho-stratigraphic, and to a much lesser extent morpho-sedimentologic. indeed, sedimentological studies in a “modern” context probably date from the 1850's or 1860's. Even towards the close of the nineteenth century it sufficed to know—for most investigators—that stratification in the drift was the product of water action, as opposed to ice action. Because of its rather complex and confusing bedding features, stratified drift was relegated to a low priority in the overall scheme of glacial investigation. Notwithstanding these intrinsic handicaps, the nineteenth century witnessed steady, if unspectacular, progress in the development of sedimentological techniques and expertise. Rapid gains were made in the last quarter of the century. The literature in the second half of the century includes a description of common bedding structures (“false” or cross-bedding, compound cross-bedding, ripple marks), particle textures (mainly size, shape, roundness, and composition), and imbrication in gravels. The directional properties of cross-bedding were used as an aid in the regional synthesis of glacial events, and heavy minerals from glacial sands were examined to elucidate provenance and regional drainage directions. Observations were also made on the downstream decrease of particle size in outwash areas, and calculations were made for the rate of transport of detritus in glacial streams, and for the rate of infill of glacial lakes. As a corollary, it was also possible to calculate the rate of denudation of drainage basins in modern Alpine areas. Climate, relief, and tectonism—all essential ingredients of modern sedimentology—were topical subjects of discussion in the nineteenth century. Some of the sedimentological studies of 80 to 100 years ago have a contemporary ring. Indeed, the pioneer workers of the nineteenth century laid a thorough groundwork for twentieth century glacial sedimentology. Specifically, their contributions to glacial stratigraphy and morphology—the channel through which glac al sedimentology evolved—were truly outstanding.

Abstract This paper reviews the hydrology and hydraulics of high energy, particularly proglacial, riverine and deltaic environments, and discusses some of the consequences for resultant patterns of sediment movement and deposition. The hydrology of proglacial rivers is under strong thermal influence and exhibits a singular pattern of flow, both seasonally and diurnally. Moderate flood flows are common. Sediment is frequently entrained and deposited, so that rapid evolution of fluvi u sedimentary features occurs on outwash plains. The possibility exists for extraordinary jokullhlaup floods to occur in front of many glaciers. The hydraulic behavior of proglacial rivers features frequent upper regime flow and rapid adjustment of channel resistance to accommodate the wide variations in discharge and sediment transport. Sediment entrainment is reviewed in sime detail, and theconceptsof “overloose” and “underloose” boundary are introduced. Sediment transport theory is reviewed and recommendations made for assessing total sediment yield. The sediment transport in proglacial rivers is anomalously high by comparison with that in nonglacial environments, because of the large volumes of drift delivered to the glacier margin. Flood deposits on outwash plains and surficial patterns of sediment texture are described. The character of the surface is conditioned by selective deposition of sediment in a simple, aggradational context. Depositional bedforms are classified as small forms (scale controlled by flow depth or lesser flow dimensions) and large forms (scale controlled by channel width). The former reflect purely local flow conditions, whereas the latter are influenced by the total flow pattern of the river. The persistence of bedforms as sedimentary structures in the stratigraphical record is considered. Gravels commonly exhibit only rudimentary plane bedding and imbrication, whereas fine materials commonly feature a wide variety of sedimentary structures. This is a consequence of the vertical distance available for deposition as compared to particle size, and the energy status of the depositional environment. River channels in coarse, noncohesive materials are wide and shallow, so that boundary resistance to flow is high. When large volumes of sediment are being transported in flood, total resistance may become too high to permit passage of the water plus sediment load; deposition and selective scour then produce narrower, deeper channels that are hydraulically more efficient. The morphological result is the occurrence of channel braiding, which is a frequent characteristic on outwash. The long profiles of proglacial rivers and outwash surfaces are concave upward as a consequence of persistent aggradation. Where proglacial rivers enter standing water bodies, classical, high-angle deltas develop. Sediment transported as bed load is deposited on the delta surface, or is deposited on the foreset wedge by avalanching over the delta lip. Turbidity flows (underflows) and slumps move coarse material farther into the water body. Finegrained material is carried in suspension into the standing water and settles to the bottom to form varves. Examples illustrating the application of principles and characteristic conditions are drawn from the literature on glaciofluvial and glaciolacustrine environments.

Abstract Glacierized areas in Norway are being considered for future hydro-electric power production due to favorable hydrologic conditions; these include high specific water yield at high altitudes and at relatively short distances from tide water. However, technical problems can be expected in utilization of sedimentladen water, particularly in reservoirs and in the turbines. A study program of sediment transport and deposition was therefore initiated, and some of the major results are given in this paper. Suspended-sediment transport was determined at five selected glaciers by frequent water sampling, at least five samples being taken daily. Sediment concentration ranged from tens of milligrams per liter to several grams per liter, and rapid variations were experienced, particularly in periods of increasing water discharge. No simple correlation was found between water discharge and suspended sediment load but, in general, years of low total water discharge gave less sediment transport than years of high total discharge. Bed load was determined at one glacier (Nigardsbreen outlet glacier from the Jostedalsbreen Ice Cap) by trapping all coarse material in a large, strong fence built across the river. The bed load accounted for 30 to 50 percent of the total transport of solid matter, as measured close to the glacier terminus. The rate of sedimentation in a natural lake close to a glacier was established both as a difference between sediment input and output and by a study of older deposits on the lake bottom. Annual layers (varves) were easily recognized and studies were made of their grain size and mineral composition. “Winter" layers had finer grains, more mica and less quartz than "summer” layers. A great part (70 percent or more) of the suspended-sediment input into the lake settled on its bottom, and the remainder that left the lake consisted mainly of the finest grain size fractions.

Abstract Faults in glaciofluvial sediments are interpreted in the light of the experimental work of Sanford (1959). Internal displacements in the sedimentary sequence resulted from the melting of associated ice. Field examples are used to illustrate that (a) melting of a discrete buried ice mass will produce a downthrown block of sediment bounded by high-angle reverse faults that are convex upwards; (b) melting the ice walls that supported an esker of subglacial origin will result in high-angle reverse faults, concentrated in the flank of the esker, that strike parallel to the esker ridge and dip steeply in toward the esker axis; such walls probably were not vertical but, rather, dipped inward toward the esker axis; and (c) large normal faults may be related to deposition over a saucer-shaped ice base. Minor normal faults may complement high-angle reverse fault systems in (a) and (b). The scale, type, and distribution of faults, and the orientation of faults with respect to the sediment body and to paleocurrent directions can contribute to a reconstruction of the sedimentation environment. These data provide useful criteria for identification of fluvial sediments as glaciofluvial.

Abstract Broad questions of esker sedimentation are reviewed in this paper. Two main environmental factors, nature of the conduit through which the esker stream flowed, and site of deposition, control esker sedimentation and commonly can be determined from the sedimentary succession. Interaction of these two factors permits definition of three different models of esker sedimentation: open-channel, tunnel and deltaic. Morphology of the esker ridge, sedimentary structures, facies relationships and paleocurrent variability are important parameters of proposed sedimentation models. The models are discussed on the basis of field data from eskers at Peterborough, Ontario and at Windsor, Quebec. Sediments of the Peterborough esker were deposited largely in an open channel bordered laterally by ice walls. Backset beds related to antidunes are preserved at places. A common environment was deltaic , where dunes and ripples delivered sediment to avalanche faces; progressively downstream from the large foresets were regressive, sinusoidal, and progressive ripples, respectively. These in turn pass into graded beds and then into lacustrine rhythmites. Tunnel sedimentation is illustrated by sediments in single steep-sided ridges in the Windsor esker. Sheetlike cross-bedded and parallel-bedded gravel and sand units persist downstream without facies change and are arranged in vertically stacked cycles that may be annual. Flow depth in the tunnel was 1 to 4 m and accumulation of sediment was accommodated by a melting upward of the ice roof. Deltaic sedimentation is illustrated by beads in the Windsor esker that were deposited annually as subaqueous fans in the water body at the mouth of the subglacial tunnel. Cobble and pebble gravel at the proximal end of the bead intertongues over a few meters in a downstream direction with ripple-laminated fine sand, units of “structureless'' fine and medium sand, and graded beds.

Abstract The internal grain size and structural properties of the Brampton esker and its associated deposits were grouped into facies and used to test several theories of eskerine sedimentation. Each time-stratigraphic unit consisted of the following proximal to distal sequence of facies: (a) cross-bedded sand and gravel of topset origin, deposited by braided streams, or otherwise during sheetflooding; (b) delta-front sands that are poorly sorted and characterized by massive structure, graded bedding, cut-and-fill structures, irregular lamination, and parallel lamination deposited in the upper flow regime; (c) cross-laminated cosets of fine sands which show stoss-side erosion, stoss-side preservation, and sinusoidal lamination formed by the movement of climbing ripples when the suspended load contributed increasingly to sedimentation; and (d) prodeltaic rhythmites of sand and silt-clay mixtures deposited almost exclusively from suspended load. The delta-front and prodeltaic sediments represent the transition between eskerine gravels and distal rhythmites, both of which are end members of a spectrum of cogenetic deposits. The facies associations of the Brampton area were separated into: (a) those of the esker proper, and (b) those of the associated deposits adjacent to the esker. The De Geer theory of eskerine sedimentation adequately explains the paleocurrent pattern and facies associations comprising the second group of sediments, but the kind of environment for the esker proper was probably open channel rather than closed conduit. A closed conduit flowing partly full, however, would function in the same way as an open channel. The sedimentary environment consisted of a braided, topset network of streams that prograded into a glaciolacustrine environment between the walls of a re-entrant within the Ontario ice lobe during late Wis-consinan time.

Abstract Stratified sands and gravels lying above Wisconsinan till and below fossiliferous marine deposits near Ottawa have markedly fluvial characteristics. Textures, paleocurrents, and sedimentary structures including channels, cross-beds, and ripple cross-lamination all indicate deposition by unidirectional flows of water. Many deformation structures can be ascribed to movement of water or water-saturated sediment, but some are regarded as collapse features related to the melting of buried ice. The succession in which the stratified sediments occur indicates that they were deposited by glacial melt-water issuing below nearly a hundred meters of standing water. They are therefore termed subaqueous out-wash and are made up of proximal boulder gravel with a sharp change to distal sand. Features which are thought to be characteristic are an association with flow till, ice-contact deformation, and (for the sand only) graded units of ripple-dri ft cross-lamination, and large scale channels with apparently massive fill. Conditions suitable for the formation of subaqueous outwash were probably widespread during Pleistocene deglaciation because large-scale isostatic depression of the land tended to pond marine or fresh water against ice fronts.

Abstract The Scott and Yana outwash fans on the northeastern Gulf of Alaska exhibit a succession of facies from glacier terminus to tidewater that are each characterized by differences of gradient, clast size, bar morphology, and sedimentary structures. The upper fan has steep gradients (as much as 17.6 m/km), large maximum clast size (>10 cm), and longitudinal bars. The midfan has gentler gradients (2 to 6 m/km), clast size ranging from less than 10 cm to sand, and predominantly longitudinal bars. The lower fan, a sand area, has gradients less than 2m/km, longitudinal and linguoid bars in braided reaches, and point and lateral bars in meandering reaches. The longitudinal bars of the upper fan consist mainly of well imbricated, poorly sorted gravel that have clast long-axes oriented transverse to the flow direction. Bars are often covered with transverse ribs, here interpreted as an upper flow regime bedform, and perhaps as relict antidunes. The midfan area is characterized by a decrease in gravel, with a corresponding increase in sand. Sand is deposited as flat upper regime beds interbedded with gravel and as lower regime megaripples in channels forming trough cross-beds. The longitudinal bars of the lower fan show planar cross-beds formed by migration of the bar slipface topped by flat beds on the bar-surface and low-angle ripple-drift cross-lamination. Linguoid bars exhibit large-scale planar to tangential cross-beds topped by ripple-drift cross-lamination. Point and lateral bars are characterized by large-scale planar to trough cross-beds caused by migration of the bar surface and bar slipface. Overbank deposits of silty sand ripple-drift cross-lamination and draped lamination increase in importance downfan. Stream regimen is governed by early summer floodng. Measurements taken during a rising stage and a declining stage indicate gravel movement in channels on the upper fan (bars were mostly emergent), but little gravel movement on the midfan. Megaripple migration in midfan channels and linguoid-bar migration on the lower fan continue at low flow stages.

Abstract Late Wisconsinan ground moraine in the Rocky Mountain Trench, southeastern British Columbia, is dissected by meltwater channels which formed during the final retreat of the Cordilleran glacier. Outwash underlying the channels is coarse, poorly sorted, shows large-scale cross-bedding, and was deposited in chan-nel-bar complexes of high-energy proglacial rivers. Length of transport of outwash gravel from till source to channel depositional site is relatively short. The coarse fraction of till is also of local derivation, because lithologies are similar to nearby bedrock lithologies. Peak discharges calculated from channel morphometry and maximum particle size were 10,000 to 20,000 cubic meters per second, larger than the estimated maximum discharges of several thousand cubic meters per second attributable to summer runoff. Many channels transmitted peak discharges during jökullhlaups from glacial lakes in tributary valleys. An empirical relationship between total volume discharged during documented jökullhlaups and corresponding maximum instantaneous discharges is applied here to Glacial Lake Elk to show that discharges equal to, or larger than, those calculated from channel morphometry were attained during jökullhlaups. Mean discharge (excluding jökullhlaups) for most of the active channels was estimated to be less than 1000 cubic meters per second; the ratio of maximum instantaneous to mean discharges was about 10:1, but may have exceeded 20:1 in some channels. At low and moderate stages rivers had braided patterns within individual meltwater channels, but during peak flows channel bars were submerged.

Abstract Paleocurrents were determined from gravel fabrics measured on the surface and in vertical sections of proximal reaches of a valley train in the Yukon, Canada, and an outwash plain in Iceland. They agree closely with the mean orientation of surface channels and with the long axes of braid bars where data are grouped for a sufficiently large area. Orientation was measured as the direction and angle of dip of AB, the maximum projection plane of discoidal clasts. Samples of 30 to 40 clasts gave consistently significant individual vector means, whereas 20-clast samples appear to be sufficient if data from several samples are grouped. Coarser gravels tend to have a higher degree of preferred orientation than finer gravels. The dip of AB averages 26 to 33 degrees upstream, giving a fabric which is consistent with deposition on subhorizontal surfaces. It is also consistent with an internal bar structure dominated by poorly defined horizontal bedding or massive gravel, with rare cross-stratification. Gravel braid bars are regarded as primary bedforms that are stable under flood conditions when all the bed material is in motion. Horizontal bedding is typical of most glaciofluvial gravels, because even at flood stage the flow is spread over a wide tract and the water is too shallow for the bars to develop slip faces. This contrasts with some Pleistocene gravels deposited by floods of exceptional depth and discharge, which formed giant braid bars with abundant large-scale cross-bedding (Bretz, Smith and Neff, 1956; Malde, 1968).

Abstract Malaspina Lake is a large proglacial lake which lies along the southeastern margin of the Malaspina Glacier, Alaska. It is density stratified with respect to suspended sediment content, which ranges from 0.1 grams per liter at the surface to 0.7 grams per liter at a depth of 45 m. The lake is not thermally stratified since water near the temperature of its maximum density (3.94°C) occurs close to the lake surface and water as cold as 0.3°C occurs at the lake bottom. Two large surface streams, Russell Stream and Tarr Stream, flow into the lake, and their combined discharge was measured at 140 cubic meters per second. Discharge from the lake ranged from 490 cubic meters per second to 600 cubic meters per second, and thus subglacial and englacial streams were apparently discharging as much as 460 cubic meters per second into the lake. Tarr Stream, which loses suspended sediment in a series of small lakes that act as settling basins, enters the lake as an overflow. Russell Stream and Tunnel Stream, which is an englacial stream, are both highly charged with suspended sediment and enter the lake as continuous turbidity currents or underflows. Two underflows and two interflows were recorded along the ice-contact margin of the lake. These underflows and interflows apparently originate from englacial or subglacial streams discharging into Malaspina Lake. Bottom topography of the lake is quite irregular except where the selective infilling of basins by turbidity currents has produced flat to gently sloping topography. Cores taken from these flat areas contain varved sediments. Varves, deposited at depths of SO m or more, contain numerous normal and reverse graded beds. The current-bedded portion of the varve was deposited from an underflow or turbidity current. The varve was completed by the deposition of a clay unit from suspension when continuous underflows, or turbidity currents, ceased during the winter months.

Abstract Sedimentary structures in distal outwash deposits, in glaciolacustrine deltas, and in lake sediments proximal to glaciolacustrine deltas are similar where the grain-size distributions of the sediment are similar, and where depositon occurs under smilar flow conditions. Draped lamination, a common structure in distal outwash deposit, consists of parallel laminae of sand, silt, and clay deposited from suspension and draped over an underlying bedform. Thickness of the laminae remains essentially unchanged across the underlying bedform, and neither silt nor mica flakes are concentrated anywhere along draped laminae. Overbank deposits in outwash decrease in grain size along the length of the stream, but contain similar sequences of sedimentary structures. Commonly, type A ripple-drift cross-lamination is overlain by low-angle type B, which in turn is overlain by draped lamination. This sequence of sedimentary structures suggests deposition under decreasing flow strength. Ripple-drift sequences in glaciolacustrine deltas are separated in many cases by winter clay layers. The basic sequence begins as a thin unit of draped lamination deposited on a subjacent winter clay. A thin unit of type B ripple-drift cross-lamination follows and is in turn overlain by a relatively thick unit of type A. The type A grades upward into a second unit of type B. As the angle of climb of the type B ripple-drift cross-lamination increases to the vertical, type B grades upward into draped lamination. A superjacent winter clay layer completes the sequence. The basic sequence of sedimentary structures reflects a relatively rapid decrease in the ratio of rates of deposition from suspension to bed-load transport, followed by a gradual increase. Flow strength appears to increase quickly and then decrease. The thin graded beds and rare ripple- drift cross-lamination or isolated ripples observed within lacustrine varved clays are distal equivalents to the sequences observed within glaciolacustrine deltas. Glaciolacustrine deltas are the products of rapid sedimentation into low-energy lake environments and are seldom seriously modified by wave action or currents. Deltas are built into lakes as overlapping lobes of sediment that were deposited from density currents issuing from meltwater distributaries. As distributaries migrate across the delta subaerial plain, lobes move laterally and overlap each other, forming an arcuate delta front.

Abstract Physiographic and stratigraphic associations of relatively coarse-grained (mainly sand) deposits, in the Edmonton and Okanagan Valley areas, Canada, illustrate deposition in glaciolacustrine deltas. Sedimentary successions are described in terms of seven facies states: gravel, cross-bedded sand, flat-bedded sand, crosslaminated sand, alternating beds, parallel-lamination, and diamicton. Proportional thickness, preferred vertical position, and transition probabilities together with grain-size and paleocurrent analysis form the basis of detailed analysis of the deltaic successions. A simple succession is recognized in which flat-bedded sands dominate the lower part and cross-stratified sands dominate the upper part. In these successions, paleocurrent directions are consistent, and there is a significant fining-upward trend. The flat-bedded sands are interpreted as multi-storied lateral-accretion deposits of distributary channels. The cross-laminated sands are interpreted as distributary-mouth bar deposits. The succession of stratification and grain-size illustrates deposition in increasingly distal positions during ice-marginal recession. A second type of succession shows complex, broad cyclicity. Rhythmic beds, with a lower member of cross-laminated or alternating beds and an upper member of parallel-laminated beds, form a thick interval between flat-bedded distributary-channel deposits. The rhythmic beds are interpreted as levee and interdiatributary-bay sediments that were deposited under seasonally fluctuating discharge. The complexity of these successions results from a combination of channel shifting and from increasingly distal position during ice retreat. In the Okanagan section, a deep sedimentation basin and coarse deposits resulted in development of a prograding Gilbert-type delta with steep foresets.

Abstract Most of the fine-grained bottom sediments of Glacial Lake Hitchcock are rhythmites composed of silt- clay couplets that occur in three structural groups : Group I—clay thickness greater than silt thickness. Group II—clay thickness approximately equal to silt thickness Group III—clay thickness less than silt thickness . Thin sections of impregnated sediments show flat bedding with as many as 40 graded laminae in one 5-cm layer. Erosional contacts and ripple cross-lamination are common in Group III, but rare in Groups I and II. The contact between silt and the overlying clay layer in any one rhythmite is gradational in less than 25 percent of the samples. Other sedimentary features include two distinct types of trace fossils. Mean grain size of the silt layers (5.50 to 8.50) depends up the environment of deposition of silt within the lake. Mean grain size of the clay layersis much the same everywhere (averaging 10.5ø). Data from 34 localities suggest thatthe rhythmites are annual (that is, they arevarves), andthefol lowing depositional mechanism is proposed. Sediment was transported by streams from the glacier and from nearby deglaciated uplands. Gravel andsand were deposited on deltas. Silt and claywere carriedoutinto the lake, mainly by turbidity currents.Sediments coming into the lake contained asignificant amount of clay that settled out continuously, but clay became the dominant deposit only during the winter when coarse material was less available. Varves belonging to Groups I and II generally were formed in still water away from river mouths, where little sediment was received directly from density currents. Group III varves were formed relatively near delta fronts, in environments characterized by high sedimentation rates.

Abstract This publication is the outgrowth of a symposium on Glacial Sedimentology that was held in Buffalo, New York, March 1972. The great interest generated in glacial phenomena during the nineteenth century had important implications and repercussions for the infant field of sedimentology. It provided its fair share of the background stimulus necessary to establish sedimentology as a separate branch of the earth sciences in the twentieth century. The time for reciprocity is now at hand; feedback from the expertise gained in the burgeoning field of sedimentology can greatly help the Quaternary specialist solve particular field problems. The last decade has witnessed a growing interest in the sedimentology of the Quaternary, and it seems appropriate now to summarize progress in the study of stratified drift, to present results of some recent studies, and to focus attention on avenues of research that should be explored in the near future.