The suggestion that patches of basal ice may freeze to the bed of a glacier due to certain regelation effects has been tested in the laboratory by applying high hydrostatic pressures to ice samples at the pressure-melting point. During compression, ice temperatures follow the pressure-melting point closely, but after rapid decompression the ice temperature at first returns only half to three-quarters of the way to the pressure-melting point, after which it appears to warm by thermal conduction from outside the ice sample. If the moving ice at the base of a glacier behaves in the same way as it is exposed to changing pressure fields, frozen patches at bedrock are to be expected.

Records of strain variations in a tunnel beneath Glacier d’Argentière show two types of strain events. The first is a rapid jump or offset in the recorded strain, while the second are strain excursions, initiated by a change in strain over a period up to ten seconds, followed by a gradual recovery to the original strain over some minutes. It is suggested that the offset events are due to nearby stress release due to fracturing of frozen patches of ice at the bedrock while the strain-excursion events show the more distant adjustment of the glacial bed to the former events due to the time lags associated with changes of water-film thickness and regelation heat flow.

Basal ice at the glacier Østerdalsisen commonly deforms under stress within cavities at the glacier bed. Some ice subject to rapid strain, however, fails in a brittle manner. The creep-rate of basal ice is influenced by sediment particles within it, and by ice grain-size and structural anisotropy; a flow law for such ice has not been derived experimentally. Near the glacier bed, particles enclosed in ice may migrate as a result of stress concentration, bubbles may be flattened, and anisotropic structures may develop. Recrystallization leads to grain-size changes. Both regelation spicules and layers of regelation ice, generally thin, form within subglacial cavities but their occurrence is limited. Regelation-ice formation has a significant influence on the distribution of rock debris at the ceilings of subglacial cavities.

A comparison of bedrock fractures with orientation of structures and fabric in the overlying late Pleistocene basal till indicates a sequence of common glacial origin. The deformations in bedrock and till, and the deposition of the till layer, were accomplished by two glacial flows that were part of a mountain ice sheet: the regional (moving south-south-east) and the local (moving south-south-west) along Peavine Valley.

The following sequence of events is postulated from this investigation: the initial glacial advance from the north-north-west that produced conjugate low-angle shear fractures in the bedrock was followed by competition of both the regional and the Peavine Valley glacial flows which also produced low-angle fracture systems. As the over-riding ice became thicker, high-angle conjugate fractures were formed by both glacial flows. Subsequently, south-east-dipping injection wedges were formed by the regional ice, which was concurrently eroding the down-glacier parts of the wedged blocks, depositing local till, and producing sigmoidal deformations in this till. The main deposition of the overlying lodgement till was accomplished by an interaction of both the regional and the Peavine Valley glacial flows, and the lodgement was supplemented by shearing of the till, producing fissility. Bedrock wedging, as a significant process of glacial erosion, is believed to be discussed here for the first time.

Glacial grooves are known on all the flat-lying limestone islands in western Lake Erie. The largest megagroove complexes are on Kelleys Island, each 2–6 m deep, 5–20 m wide, and 100–400 m long before quarrying. Each megagroove floor is rolling with second-order rounded ridges and “sine-shaped” grooves 10–90 cm deep, and 5–40 m long. Furthermore, these are gouged laterally by sharply curving and nesting gouges up to 10 cm deep. All grooves exhibit third-order striation up to 1–2 mm deep and 5–200 cm long made by individual tools of debris traveling mostly 240–260°.

Ice of the Erie lobe was under compressive flow as it rose from 9 m below present lake level to 12 m above. Some sort of ice vortices (fixed eddies in the basal ice stream) are postulated to make the scoop marks and ridge ends. Fast-moving ice rather than water or a slurry is favored because there are no percussion marks. Very likely, a till mat of interfering engraving points is required. A strong convergence (2× to 10×) of striae into the deepest grooves indicates squeezing together of the debris tools and increase in local ice velocity. Interglacial subareal streams may well have shaped the initial trough up which these ice streams concentrated, because dendritic tributary grooves intersect, and main groove sets curve as much as 10° or 20°. Furthermore, sharp 100–180° meander curves are preserved at the bottom of the deepest grooves. Hiram-age clay-till with rare erratics half fills the original grooves; it shows that this groove cutting was completed before 15 500 14C years ago.

In 1975 and 1977, 24 bore holes were drilled to the bed of South Cascade Glacier, Washington, U.S.A., using both electrothermal and hot-water drills. Only two holes connected directly with the basal water system, a significant decrease from the four to five such connections in 13 holes drilled in 1973 and 1974 (Hodge, 1976). Most of the bed, possibly as much as 90%, appears to be hydraulically inactive and isolated from a few active subglacial conduits. Bore holes which penetrate these inactive areas initially should connect eventually with the active basal water system due to bed pressurization by the water standing in the bore hole, provided there is a sufficient supply of water available to form and maintain the connection passageway. These sealed-off areas probably consist of the sub-sole drift and permeability barriers found recently at the bed of Blue Glacier by Engelhardt and others (1978); an increase in the area of bed covered by these features probably caused the decrease in chance of bore-hole connection. This apparently was not due to any external cause but rather was the result of a real internal change in the subglacial hydraulic system which occurred between 1974 and 1975.

If most of the area of a glacier bed is hydraulically isolated sub-sole drift, or something similar, such features may well control large-scale glacier sliding changes, since changes in the amount of water having access to the glacier bed will take considerable time to affect the interstitial water pressure in the more widespread sub-sole drift.

Water pressures in the active part of the basal water system of South Cascade Glacier are generally in the range of 50–75% of the ice overburden pressure. Water levels in a connected bore hole are probably representative over an area of the bed 100 m or more in extent. A correlation of bore-hole water levels with changes in surface motion supports the idea that the sliding of a temperate glacier is controlled largely by the basal water pressure.

Recent studies of subglacially precipitated carbonate deposits and associated solutional furrows have provided interesting new insight on subglacial water films, as well as on chemical exchange at the glacier bed. Considerable information on the film thickness and its temporal and spatial variability has been gained by analyzing several properties of subglacial carbonate deposits including: (1) the morphology of surface features aligned parallel to ice flow, (2) the laminated structure, and (3) the size distribution of fine rock fragments presumably transported in the film prior to their incorporation in the deposits. Chemical analyses of water from pro-glacial streams, together with calculations of CaCO3 solubility and mass balance, show that the channelized water is chemically distinct from the film water in which CaCO3 precipitates, and that subglacial precipitation is not possible where there is a considerable water flux through the film in excess of that associated with regelation sliding. The principal implication of these studies is that a temperate cirque glacier is characteristically separated from its bed by a thin water film, probably micrometers in thickness; however, the film appears to occasionally thicken, at least locally by as much as a hundred fold in exceptional cases. Furthermore, the water flux and/or solute concentration in the basal film undergoes periodic, probably seasonal, variations possibly related to variations in the amount of water reaching and flowing through the basal film.

The spatial pattern of the formerly active processes of water flow, cavitation, abrasion, dissolution, and precipitation at the base of a small cirque glacier has been reconstructed by detailed mapping of surficial features on recently deglaciated limestone bedrock near the glacier terminus. Our interpretation of these features, which reflect basal conditions averaged over a period of several or several tens of years, leads us to the following conclusions:

1. 1. A nearly continuous, non-arborescent network of cavities and incised channels existed and probably acted as the primary drainage of melt waters. This network evolved through time as many channels were filled, perhaps intermittently, by basal ice.

2. 2. At least 20% of the glacier sole was separated from the bed by water-filled cavities. The rest of the glacier–rock interface characteristically comprises a very thin water film.

3. 3. Abrasion was locally intensified, relative to chemical alteration, in 5–10 m wide zones paralleling the ice-flow direction, perhaps as a result of locally enhanced sheet flow of subglacial water.

Two components of discharge through the internal hydrological systems of Alpine glaciers were separated on the basis of chemical composition of water. Some surface melt waters retain low solute contents after flowing without delay through conduits in which no chemical enrichment occurs, whereas those flowing slowly at the glacier bed have increased ionic concentrations. A simple mixing model was used to investigate temporal variations in the quantities of water routed through each of the two sub-systems. Electrical conductivity was taken as an indicator of melt-water composition and was monitored for periods during the summer ablation season of 1975 at Gornergletscher and of 1977 at Findelengletscher. At both glaciers, conductivity of melt waters varied diurnally inversely with discharge fluctuations, depending on the proportion of total discharge routed through the two sub-systems. Total discharge and the flow component routed rapidly through conduits within the glacier, a large proportion (50–80%) of total discharge, exhibited in-phase rhythmic diurnal hydrographs at the two glaciers. Distinctive subglacial hydrological regimes are contrasted. At Findelengletscher, the hydrographs of total discharge and of subglacial chemically enriched flow were in phase. At Gornergletscher, the subglacial hydrograph occurred with reverse asymmetry and out of phase. A possible interpretation is that water was temporarily stored in basal cavities during high total discharge. During the night, stored water was released, contributing much of the total discharge at times of low flow.

For the construction of a hydro-electric power station in western Norway, a diversion tunnel was made to collect subglacial melt water under an outlet glacier from the Folgefonni ice cap. Many investigations were carried out by glaciologists and engineers before the project could begin, and several unexpected problems arose during the completion of the project. This paper deals with some of the problems and how they were solved.

To avoid coarse glacier-carried material from being flushed into the water-collecting tunnel system, a large sedimentation chamber was constructed in the bedrock under the glacier. The dimensions of this huge chamber were decided from sediment-transport studies in the glacier stream and from studies of old bottom deposits in a lake close to the glacier front.

Ice-velocity measurements were made on the glacier surface and similar studies were attempted in sub-glacial ice caves made by spraying hot water near the glacier bed, where the ice is 170 m thick.

The subglacial water-drainage system was studied from a horizontal tunnel constructed in the bedrock under the glacier. Some preliminary conclusions are drawn from these studies.

In future, it will still be possible to undertake subglacial studies because inspection tunnels have been left in the bedrock, and the accessibility is relatively good.

Some of the problems with earlier theories for erosion and transport by ice sheets are discussed, and it is noted that those theories cannot simply account for the often-reported finding that most till is derived from bedrock only a few tens of kilometers up-glacier. Considerations of the mass balance of debris in transport lead to the conclusion that ice sheets are capable of transporting most debris only a short distance.

The theory that the break-up of bedrock is mostly a preglacial process is developed. The advancing ice sheet collects the debris and then deposits it after a short travel. As the ice sheet first advances over the regolith, debris is frozen onto the base and is carried until basal melting due to geothermal and frictional heat causes lodgment till deposition. Most debris is deposited during the advance of the ice sheet and is carried only a short distance. A generally small amount of debris is carried at higher levels and is deposited during ice standstill and retreat as melt-out and ablation tills.

The present theory makes many predictions, among them, that most till units are not traceable over long distances, that thick till sequences represent unstable glacier margins and not necessarily long periods of glacier occupation, and that lodgment tills are to be interpreted in terms of ice advances and ablation tills in terms of ice retreats.

This paper is published in full in Journal of Geology, Vol. 86, No. 4, 1978, p. 516–24.

The 1976/77 Norwegian Antarctic Research Expedition carried out studies of the sea bed by side-scan sonar. The equipment was operated from the expedition vessel down to about 350 m depth by personnel from the Continental Shelf Institute, Trondheim. Various types of plough marks mostly ranging from 10 to 100 m in width were observed. These included several generations of crossing plough marks as well as plough marks with abrupt changes in trend reflecting changing iceberg motion. The investigations will be expanded during the 1978/79 expedition to include towing at greater depths, and mapping of sea-bed morphology by mosaic towing patterns.

Understanding the relationship between the morphology of former ice-sheet beds and glaciological processes is handicapped by the difficulty of establishing which stage of a cycle of ice-sheet growth and decay is responsible for most erosion. Discussions at this conference and in the literature display a variety of opinions, some favouring periods of ice-sheet build up, others periods of fluctuations, and still others steady-state maximum conditions. Here it is suggested that there is geomorphological evidence which points to the dominance of maximum conditions.

Along the eastern margins of the Laurentide and Greenland ice sheets there is a sharp discontinuity between Alpine relief which stood above the ice-sheet surface at the maximum and plateau scenery which was covered by the ice sheet. Often the two types of relief are adjacent and yet separated by an altitudinal difference of only 100–200 m. The existence of an abrupt rather than gradual transition from one relief type to the other suggests that most glacial sculpture must have taken place while the ice sheet was at its maximum extent. In other geomorphological situations where high mountains were submerged by ice sheets, the major erosional landforms are frequently found to relate to ice sheets rather than to local mountain glaciers, again suggesting the dominance of erosion during full ice-sheet conditions. Finally, the identification of patterns of glacial erosion on an ice-sheet scale in North America and Greenland points to erosion when the ice sheets were fully expanded, rather than to the variable flow conditions associated with growth or decay.

If ice-sheet erosion is accepted as being a result of maximum conditions, then it places certain constraints on glacial theory, for example the need to develop theories of glacial erosion which apply beneath ice thicknesses of several thousand metres. It also suggests that the use of steady-state models of ice sheets is likely to be a profitable way of relating glaciological processes to the morphology of former ice-sheet beds.

If glacier ice is assumed to be a uniaxial birefringent material then two parameters are needed for describing the orientation behaviour of the polarization ellipse of a radio signal returned from the glacier bed. These are δ, the relative phase between the ordinary and extraordinary wave, and r, the ratio of their attenuation coefficients (Woodruff and Doake, 1979). If it is assumed that the attenuation in ice is isotropic, then a value of r other than unity will be due to anisotropic reflection coefficients at the glacier bed. A simple way of measuring r separately from δ could then be used to characterize the reflecting surface.

Two fundamental boundary conditions for basal ice may be recognized by radio echo-sounding: ice–rock and ice–water interfaces. The latter is identified on the basis of horizontal echoes displaying slow fading, high persistent returned power, and with a reflection coefficient lying between –1 and –3 dB. Such water, located beneath an ice sheet, provides a known limiting condition to the vertical temperature profile and allows more precise calculation of the basal temperature gradient and estimation of the heat flux through the sole.

Variegated Glacier is a surge-type glacier in the St Elias mountain range in Alaska. The interval between surges is about 20 years; the last one occurred in 1964 to 1965. This glacier has been studied extensively since 1973 (Bindschadler and others, 1977). Thus far, measurements of ice velocities have been restricted to the surface. They have been analyzed using geophysically measured ice depths, in order to estimate ice velocities in the ice mass and at the base (Bindschadler and others, 1978). From 1973 to 1977 the distribution of annual ice velocities along most of the length of the glacier can be explained primarily by internal deformation without major contribution from sliding at the base. However, the variation of surface velocity with time gives definite indication that sliding occurs in summer and that the average summer rate is increasing progressively from summer to summer and that in a zone 5 to 7 km below the head of the glacier the summer-to-summer increase in inferred sliding rate is especially rapid. This is a notably distinguishing feature, which is probably indicative of a build-up toward the next surge. In order to obtain direct information about sliding-rates and water pressures at the base in this zone, a bore hole was drilled to the bottom of the glacier about 6 km below the glacier head. Observations in the hole started in June 1978 and were continued until 31 July 1978. The hole connected to an englacial water system at a depth of 204 m whereupon the water level dropped gradually to about 100 m below the surface. The last 6 m above-the base at 356 m could be drilled only by means of a cable tool because of the presence of debris-rich ice. Upon reaching the bottom, the water level increased rapidly to the firn water table at about 8 m below surface. Large variations in water level of about 200 m occurred during the following period of observation of 35 d. Major events such as audible icequakes, heavy rainfalls, and a period of unusually high ablation were associated with abrupt increases of water level up to the firn water table. High water pressure at the bottom drove a flow of muddy and sandy water upward in the hole. Consequently high freezing rates in the lower 150 m of the hole produced a very rough bore-hole wall covered with ledges, coral-reef-like features, grooves, and pockets filled with sand. Near the bottom, embedded rocks stuck out of the bore-hole wall. These features were recognized by bore-hole television. The bore-hole bottom consisted of sand which continuously proliferated and washed into the hole. Attempts to remove this sand by means of a sand pump failed, the bailed-out sand being replaced immediately. From bore-hole inclinometry an internal deformation of the ice mass of 0.22 m d−1 was obtained. Together with average surface velocity of 0.47 m d−1 we get a sliding velocity of 0.25 m d−1, averaged over the time of observation. This result confirms the sliding velocities inferred from surface velocity measurements. It also lies on the exponential trend line of increasing summer-to-summer velocities showing a doubling of sliding velocities about every two years (Bindschadler and others, unpublished). This strongly indicates that the next surge is likely to occur in the early eighties. Input of water from the surface probably will play a role in triggering the surge.

Analysis of field data from Variegated Glacier supports the conclusion of Meier (1968) that no simple relationship between basal shear stress and sliding velocity can be found. On the other hand, an index of bed separation is defined and evaluated that correlates very well with the longitudinal variation of summer sliding velocity inferred for Variegated Glacier. This bed separation parameter is defined as where τ is the basal shear stress and is proportional to the drop in normal stress on the down-glacier side of bedrock bumps and Neff is the effective normal stress equal to the overburden stress minus the subglacial water pressure. The water-pressure distribution is calculated assuming water flow to be confined in subglacial Röthlisberger conduits. The excellent agreement between the longitudinal profiles of I and sliding velocity suggests that calculations of the variation of bed separation can be used to deduce the variation of sliding velocity in both space and time. Further, it is possible that a functional relationship can be developed that adequately represents the geometric controls on basal sliding to permit accurate predictions of sliding velocities.

Ice velocity, acceleration, and strain-rate are being measured at intervals of about 45 d using automated processing of aerial photographs. Preliminary results for the lowest 6 km of Columbia Glacier for the period July 1976 to November 1977 show the following:

• (1) High rates of flow (2–6 m/d) and longitudinal extension (occasionally exceeding 1/a).

• (2) A smooth, almost sinusoidal change in velocity 6 km above the terminus from about 2 m/d in August 1976 and 1977 to about 3 m/d in February–March 1977.

• (3) Very rapid increases in velocity near the terminus in October 1976 and September–October 1977 which follow embayment formation (increased iceberg calving).

• (4) A suggestion that the rapid velocity increases propagate up-glacier for 3–4 km before becoming unrecognizable.

• (5) Seasonal changes in velocity (which presumably relate to basal sliding) ranging from about 1 m/d 6 km above the terminus to about 4 m/d near the terminus.

Almost continuous recordings of glacier seismicity were obtained on Unteraar-gletscher (Berner Oberland, Switzerland) over a period of two weeks at the end of May 1977. The seismograph array consisted of micro-earthquake recorders, some recording directly onto smoked paper (three Sprengnether model MEQ-800) and others recording onto magnetic tape (one MARS-72 apparatus and two seismic-event recorders). The frequency range covered by the instruments lies between 1 and 85 Hz. The array spanned a relatively crevasse-free area of about 500 m by 700 m.

At the beginning of the period of observation only a few events were recorded per hour. Probably as a consequence of increased run-off of melt and rain, which acted as a lubricant or as an agent of hydraulic fracturing within or beneath the glacier, the activity rose in two days to a maximum of 120 events in one hour, and then with minor fluctuations it gradually decreased again to its initial level. The variations in activity showed signs of correlating with the velocity measurements, which however were only started after the main maximum, but did not seem to be influenced by the uplift of the glacier surface or by the diurnal meteorological fluctuations.

Most of the recorded signals could be classified according to their frequency content into three groups: 40, 20, and 5 Hz. The 40 Hz events, whose duration was less than 1 s, are located near the glacier surface and probably correspond to those attributed by Neave and Savage (1970) to the formation of crevasses. The majority of the events were of the 20 Hz type, each of about 1 s duration. Due to inadequacies in the seismometer array, only three of them could be located with sufficient accuracy: they originated in different areas of the glacier at a depth of about 150 m, which is strong evidence that icequakes are not, as previously thought, limited to the upper 40 m. The signals of the 5 Hz type, lasting several seconds, are thought to be 20 Hz events whose higher frequencies have been attenuated by a longer travel path. The lower-frequency signals from the glacier bed, reported by Bindschadler (unpublished) and by Weaver and Malone (1979), were not observed. This could indicate that a variety of mechanisms are operative at different times of year. In particular, the subglacial conduits were only poorly developed in Unteraargletscher at the time of these recordings.

Flat-surfaced subglacial lakes, such as those discovered by Oswald (1975), can form where basal sliding velocities are low. As the glacier moves onto a subglacial lake, its underside retains the shape of the vertical relief at the lake edge. Differential heat transfer from the lake into high and low points in the interface flattens the surface by extra melting of the downward projections or freezing in the high areas. A time of the order of 103 years is required to flatten the fluting, and the down-glacier part of a subglacial lake can be flat-surfaced if the glacier takes longer than some 103 years to traverse the lake. Such a slow traverse time is expected near ice-drainage divides, and it is there that flat-surfaced lakes were discovered by radar sounding.