An analysis is made of the steady-state width of an ice sheet whose base is below sea-level, whose basal temperatures are such that appreciable melting occurs at the base, and which is fringed by fastmoving ice streams that drain most of the outward ice flux. The fast ice velocities of the ice streams are considered to be a consequence of substantial subglacial water flow underneath the ice streams. The source of this water is the water melted from the base of the ice sheet which is diverted to flow beneath the ice streams. If the depth of the sea at the edge of the ice sheet is not a function of the width of the ice sheet, then an ice sheet with a steady-state width is in a situation of unstable equilibrium. Only if the sea-level depth at the edge of the ice sheet increases as a function of ice-sheet width at a rate greater than the 2/3rd power of the width can a stable, steady-state ice sheet exist. This condition (taking into account elastic rebound) is not satisfied for the West Antarctic ice sheet along an ice-flow path from the ice divide above Byrd station out to the Ross Sea. An increase of the mean precipitation, such as might occur under a C02-induced climatic warming, would cause growth of both stable or unstable steady-state ice sheets.

The preparation of a US Geological Survey Professional Paper, “Satellite image atlas of glaciers”, has produced a 1:5 000 000 scale “Landsat index map of Antarctica”, in which each of the 2 470 Landsat nominal scene centers is represented by a symbol showing the suitability of available Landsat images for the preparation of planimetric image maps and for glaciological studies. Landsat has the potential for imaging about 79% of the area of Antarctica, and 70% of the Landsat imaging area, or about 55% of the continent, was found to have excellent or good (less than 10% cloud cover) coverage.

Australia, Japan, New Zealand, the United Kingdom, and the United States of America have published Landsat image maps, either as single Landsat scenes or as mosaics of two or more images. The Federal Republic of Germany and the Republic of South Africa also plan to publish Landsat image maps in the near future.

Available Landsat images could be used, in combination with Doppler satellite technology for geodetic control, to triple the area of Antarctica presently mapped at scales of 1:250 000 or larger. Landsat-3 RBV images can also be used to prepare 1:100 000 scale image maps.In addition to eventually using Landsat images to compile an accurate coastline of Antarctica, Landsat images have been successfully used for glaciological studies. Recent measurements of successive images of Pine Island Glacier, Walgreen Coast, West Antarctica, showed an average speed of flow of the terminus of 6 m d−1 over 750 d.

Radiation budget measurements were made at Mizuho station (70°42'S, 44"20'E, 2 230 m a. s.1.), East Antarctica, in 1979, within the framework of the Japanese POLEX-South programme. Global, and reflected short-wave and downward and upward long-wave radiat i on fluxes were measured at the snow surface and at the top of a 30 m tower. Direct solar radiation was also measured at the snow surface.

Seasonal variations of net radiation and net short-wave and net long-wave radiation are presented. Daily variation of net radiation is also presented with the daily value of meteorological elements. The monthly amounts of net radiation in winter months had very large negative values of about -80 MJ m−2 month−1. (-2 kly month−1). Daily totals of net radiation for clear skies were negative even i n summer, and were always smaller than those for cloudy skies. Monthly amounts of net radiation in summer months (about -1 MJ m−2 month−1 in December) were the smallest among the several Antarctic stations compared, and whether the balance was negative or positive depended on the ratio of clear and cloudy days. Comparison of seasonal variations of radiation components was made and the dominant cause of the radiation balance was discussed.

The distribution of mean net accumulation rate over a large sector (90°E to 150°E) of the IAGP (International Antarctic Glaciological Project) area in Antarctica is presented in map form. The basic data have been compiled from direct measurements of accumulation at stake networks, and by dating horizons in the snow pack by gross β-activity measurements, on four long traverse routes. The spatial variability can be high, over 40%, for a single year, and typically about 20% for multiyear averages. Local surface topography in the form of undulations of several kilometres wavelength can account for about 35% of the variance. Averaging over Intervals of 50 km or more will give the smooth large-scale distribution pattern. But the resultant values can still deviate from the long-term average accumulation by 25% or more due to temporal variability over time scales of years to decades.

A very high correlation (r = 0.97) was found for the relationship a = 1414 exp (0.060 θ10) between accumulation, a kg m−2 a−1, and 10 m depth snow temperature, θ10°C, when the data set was limited to a single large drainage basin in Wilkes Land.

According to our present picture, the Ross Ice Shelf is subject to relatively rapid changes, perhaps constantly out of steady state, but not undergoing a long-term secular change. Recent supporting evidence comes from a flow band of ice extending from the edges of Beardmore Glacier as far as Nimrod Glacier. The boundaries of that band and of ice stemming from several individual glaciers within it have been traced on airborne radar records. Using measurements made as part of the Ross Ice Shelf Geophysical and Glaciological Survey (RIGGS) program, mass-flux variations along the bands have been calculated. The band from Nimrod Glacier, a major outlet glacier from the East Antarctic inland ice sheet, shows no significant deviations from zero for the sum of thence thickness change rate ∂H/∂t and bottom melt rate. ḃH. We interpret this to mean that ∂H/∂t and ḃH are separately small. Significant flux variations in the entire flow band are then attributed to relatively large variations in input flux from the alpine glaciers of the Transantarctic Mountains, and from zones between the glaciers. Although flux variations are not coherent between individual glacier bands, the average strengths of internal reflections (from bottom crevasses and/or included moraine), exhibit a semi-coherent variation with a period of 400 a that correlates with 180 variations in ice cores from Dome C and Byrd station.

An extensive program of seismic refraction and reflection shooting was undertaken in the vicinity of Dome C station, East Antarctica (latitude 74°39'S, longitude 124°10'E). Refraction experiments (1978–79, 1979–80) were performed at 37 different shot-point-to-center-of-spread distances along four lines of different azimuth using receiver spacings of 2 to 30 m. The principal result is an extremely well-determined velocity-depth relation for the upper few hundred meters of the ice sheet. Availability of good bore-hole control at Dome C allows us to demonstrate the validity of Kohne's (1972) density-velocity formula, which was developed for warmer ice. Also, after a careful search for azimuthal dependence of variations in the time-distance relationship along different refraction lines, deviations from transverse isotropy in the firn and upper ice are not indicated. Our reflection work (1979–80), conducted along three mutually-centered lines (10.2, 10.2, and 10.7 km in length) oriented at 60° angles to one another, used a series of 16 bore holes, each 30 m in depth, and 18 separate 24-channel recording stations. This extensive coverage combined with common reflection-point shooting geometries (to minimize topographic error) and small charge sizes (to increase resolution) yielded the highest quality set of reflection data yet obtained in East Antarctica. Significant crystalline anisotropy in the lower portion of the ice sheet is indicated. A model with 75% of the ice thickness below the firn (approximately 2 400 m) having c-axes distributed throughout a vertical cone with a semi-apex angle of 20° gives good agreement with the field observations.

Gravity data from cruises 32, 51, and 52 of the USNS Eltanin (Hayes and Davey 1975), approximately adjusted to the new gravity datura (IGSN 71) and reference system (GRS 67), have been used to extend the Ross Ice Shelf gravity map to the edge of the Ross Sea continental shelf. Various regional gravity anomaly fields obtained by applying low-pass filters to these data are strongly negative over the entire Ross embayment; values increase gradually from approximately -300 gu (-30 mgal) near the Siple Coast to -150 gu (-15 mgal) near the edge of the Ross Sea continental shelf.

A preliminary numerical model of ice-stream grounding-line retreat and thinning as a function of eustatic sea-level rise has been developed. Grounding-line retreat is computed using the methods of Thomas and Bentley (1978). The change in thickness along the ice-stream profile is found by solving the equation of continuity in two dimensions during the time intervals corresponding to increments of grounding-line retreat. The model was applied to ice stream E, West Antarctica, assuming that the ice stream was initially grounded to the edge of the Ross Sea continental shelf.

Extensive and detailed radar surveys at Dome C were conducted during the 1978–79 and 1979–80 austral field seasons by groups from the University of Wisconsin. Measurements were conducted within a 10 × 10 km grid centered approximately on the Dome C camp (some additional studies were carried out as far as 20 km from the camp) and involved profiling of internal layers and bottom topography. In addition, a new digital recording system for the radar was used during the first season and has yielded interesting results on internal layers near the French bore hole.

Analysis of these data has produced a local map of bottom topography showing generally rough terrain. In particular, there is a rapid deepening of the bottom topography (about 500 m over 2 km) just grid south-west of the camp. Internal layers were found to be discontinuous on a scale of tens of meters. The deepest internal layers were detected at depths of about 2 500 m although a set of remarkable layer-like returns were observed about 50 to 100 m above the interpreted base of the ice. (It is not yet clear whether these returns represent reflections from layers internal to the ice or whether they are reflections and diffractions associated with the rough terrain.) In addition the processed digital records show an abrupt decrease in the reflection strength of internal layers at about 1 700 m. Because the digitally recorded data were collected at only one site, we reserve concluding that this observation is characteristic of the entire Dome C area until more of the photographically recorded data can be reduced.

Surface and airborne radar sounding data were used to identify and map fields of bottom crevasses on the Ross Ice Shelf. Two major concentrations of crevasses were found, one along the grid-eastern grounding line and another, made up of eight smaller sites, grid-west of Crary Ice Rise.

Based upon an analysis of bottom crevasse heights and locations, and of the strength of radar waves diffracted from the apex and bottom corners of the gridcrevasses, we conclude that the crevasses are formed at discrete locations on the ice shelf. By comparing the locations of crevasse formation with ice thickness and bottom topography, we conclude that most of the crevasse sites are associated with grounding. Hence we have postulated that six grounded areas, in addition to Crary Ice Rise and Roosevelt Island, exist in the grid-western sector of the ice shelf. These pinning points may be important for interpreting the dynamics of the West Antarctic ice sheet.

Deep geoelectric soundings using a Schlumberger array were carried out at Dome C, Antarctica. To penetrate the 3 500 m thick ice, electrode half spacings up to 8 km (the largest yet made on polar ice) were used. A constant-current dc transmitter with voltages up to 10 kV, resulting in constant currents of up to 10 mA, was employed; potentials were measured with an electrometer and recorded continuously on an x-y recorder.

A computer program has been developed to calculate apparent resistivities on an ice sheet in which the density, temperature, and impurities, and thus the actual resistivity, vary continuously with depth. Preliminary models show that the actual resistivity at a depth of about 500 m increases downward by a factor of about three, and that elevated resistivities must persist to much greater depth. This change in resistivity appears to be linked to transition from ice-age to Holocene ice.

The cause of this resistivity increase is not well understood, and may be attributed to any of several physical and chemical changes that also occur at about the same depth.

A common-reflection point profiling experiment to obtain electromagnetic wave velocities in the ice at Dome C at 35 MHz was carried out to a maximum antenna separation of 2 km. Four different recording systems were used for this experiment. The echoes from numerous internal reflecting horizons within the ice and the bedrock were recorded in four different ways: in A-display form on film using an oscilloscope, in intensity-modulated form using the Honeywell Visi-corder and thermal intensifier on heat-sensitive silver paper, and in both raw and signal-averaged form on magnetic tape.

Travel times of oblique reflections from nearly 160 internal layers down to a depth of 2 600 m and reflections from the ice bottom were measured at each station along the profile. The average wave velocities from the surface of each internal layer were measured to obtain a continuous mean velocity vs depth profile.

Velocity-density models derived from the dielectric mixture equations of Looyenga, Böttcher, Lichtenrecker, Hanai-Bruggeman and Wiener were compared against the measurements. In addition Robin's empirical relationship was also used in this study. The preliminary results show that the observed velocity depth profile for the ice column is compatible with Wiener's equation (with Formzahl = 0) and a newly derived empirical relationship (from this study) of where γi is ice density, γf firn density, ε dielectric constant of mixture, and εi dielectric constant of ice; γf>0.55 kg m−3. It appears that the electromagnetic wave velocity in the firn is 20 m MS−1 or more, which is higher than previously assumed.

The above models were also compared with velocity measurements in the Devon Island bore holes (Robin 1975) and at several sites on the Ross Ice Shelf where the ice is relatively thin compared to Dome C so that the effect of firn on mean velocities is more pronounced. The results for these sites suggest again that Wiener's equation and empirical relationship from this study produce velocities compatible with the observed values.

The heat regime and dynamics of the Antarctic ice sheet are studied using numerical modelling for two flow lines, one of which passes Vostok station and the other Byrd station. A two-dimensional non-steady heat-transfer equation with an energy dissipation term was used. The study consists of two parts. The first is a study of velocity and temperature distributions within the glacier under steady-state conditions. The second study was performed assuming surface temperature changes intended to model palaeoclimatic changes for the last 100 ka and also to model future climate changes due to a possible "greenhouse" effect.

Computer numerical modelling shows that the Antarctic ice sheet retains a record of the climatic temperature minimum 18 ka BP. Numerical modelling of the greenhouse effect assumes a temperature increasing by 10 deg within the next 100 a; its influence increases after this even if the surface temperature then remains the same for the next 20 ka. It is shown that for the next 1 ka the temperature wave will penetrate only a thin surface layer of the ice. Even in 20 ka the bottom temperature of the ice sheet will still be unchanged. Small increases of ice velocity can produce ice-sheet thinning of the order of 10 mm a−1.

Various methods have been used to estimate mean multi-year values of moisture, radiation, and heat exchange in the Antarctic ice sheet/atmosphere system. The major components of the balance have been determined as absolute and relative values. The net advection of moisture is taken as 100%, of which 83% is deposited as accumulation on the ice sheet, and the residue in the atmosphere is 15%; loss from the icesheet surface is 2%. In the radiation balance, input at the top of the atmosphere is 57%, absorption in the atmosphere is 43%, loss due to reflected shortwave radiation is 35%, and long-wave radiation from the atmosphere is 78%, while net outgoing long-wave radiation from the surface is 9%. The heat-budget components are:

The Antarctic ice sheet is a vast heat sink. Constant negative surface-radiation balance and low temperature of the ice sheet suggests that it will survive with even small amounts of precipitation. Thus the contemporary glaciation of Antarctica is rather stable.

The ice core from the 1 415 m Vostok bore hole has been studied. It was found that the ice-grain size increases with depth in the upper 700 m, a sharp gradient change occurring in the 300 to 400 m range. The grain cross-section area at depths of 100, 200, 300, 400, 500, 600, and 700 m was 1.1, 2.0, 1.5, 1.9, 2.6, and 3.3 mm2 respectively. Since grain size is a function of age, and is determined by initial size and growth rate, the latter being exponentially related to ice temperature, an attempt was made to interpret the obtained data in terms of palaeoclimatology.

Calculations show that the upper part of the ice sheet (down to 300 m depth) formed during the past 12 ka, and grew under temperatures higher than those at which the lower part of the ice formed, that is ice at 300 to 700 m depth. This conclusion was confirmed by the results of oxygen isotope analysis.

The air content of ice at depths 100 to 650, 650 to 850, 850 to 1 100, and 1 100 to 1 400 m reduced to normal conditions was 65, 70, 75, and 70 mm3 g−1 respectively. Calculations suggest that 3 to 30 ka BP the ice-sheet elevation at Vostok station was close to the present one, while 30 to 40, 40 to 55, and 55 to 75 ka BP it was 500, 1 000, and 500 m lower than at present, respectively.

Palaeoglaciological studies, including glaciogeomorphological observations and comprehensive studies of the composition of glacial deposits, undertaken by scientists of a number of countries, enable the major stages in the evolution of glaciation of some regions of East Antarctica to be outlined.

In this report, palaeoglaciological reconstructions for certain key territories: Queen Maud Land, Mac. Robertson Land, and Victoria Land are considered. The completeness and reliability of such reconstructions are also discussed.

The region of Prince Charles Mountains (Mac. Robertson Land) turned out to be one of the most significant for palaeoglaciology. In this region, the author has discovered and studied glacial deposits of at least six age stages, their formation having taken place during approximately 20 Ma.

An attempt is made to compare the results of regional studies and to present the evolution of the development of the whole East Antarctic ice sheet in space and time.

Different examples of palaeoglaciological reconstructions of the ice sheet of East Antarctica are presented, the possibilities of different approaches are evaluated practically, and schematic maps of the change in glaciation of East Antarctic regions at different evolutional stages, compiled by the author, are presented for discussion.

George VI Sound lies between Alexander Island and the Antarctic Peninsula and is over 20 km wide and 500 km long. At present an ice shelf fills the sound and is nourished largely by ice from the Antarctic Peninsula which flows across the sound to ground against the coast of Alexander Island. Ice-free areas, comprising small nunataks and larger massifs, fringe both sides of the sound and contain evidence of the former glacial history of the area. This paper describes the field evidence in detail and uses geomorphological and sedimentary analyses to put forward a relative glacial chronology, constrained by two absolute dates.

The chronology distinguishes:

• (1) a maximum state during which all ice-free areas were submerged by ice flowing into George VI Sound from both the Antarctic Peninsula and Alexander Island and thence along the sound as an ice stream. This occurred in the late Wisconsin and followed an interstadial or interglacial when George VI Sound was free of an ice shelf.

• (2) a valley-based stadial during overall deglaciation represented by pronounced marginal moraines on Alexander Island.

• (3) deglaciation to a stage where there was less landbased ice on Alexander Island than today. At this stage isostatic recovery was incomplete, relative sealevel was higher, and George VI Ice Shelf penetrated further into embayments on Alexander Island than at present.

• (4) probable disappearance of George VI Ice Shelf by 6.5 14C ka BP.

• (5) neoglacial readvance of local glaciers on Alexander Island to form three closely spaced terminal moraines and the growth of a new George VI Ice Shelf which was again more extensive than at present.

• (6) subsequent oscillations of both smaller Alexander Island glaciers and George VI Ice Shelf probably during the Little Ice Age. These fluctuations are similar to those in other sub-Antarctic Islands in the Scotia Sea and also in southern Chile.

Criteria for the determination of geocryological zones are: (1) distribution of permafrost, (2) the mean annual temperature of permafrost, (3) thickness of the active layer, (4) types of ground ice and ice content of permafrost, (5) regularities of cryogenic structure, (6) cryogenic phenomena, and (7) types of cryogenic relief.

Interpretation of the available data allows us to delineate the following geocryological zones in the Antarctic: (1) sub-Antarctic islands, (2) coastalcontinental, and (3) intercontinental. The brief general geocryological characteristic of these zones is given in the paper.

The sub-Antarctic island zone features discontinuous and continuous permafrost. The mean annual temperature of the permafrost is near 0°C and the maximum thickness is several tens of metres. The seasonal thawed zone has a thickness of 0.4 to 1.2 m. There are visible ice inclusions and very small ice forms in the permafrost and seasonal frost. Frost action causes frost sorting, hill - slope forms, and thermokarst. Ice-wedges are not found in the sub-Antarctic zone.

The coastal-continental zone has continuous permafrost. The mean annual temperature of the permafrost therejs -1°C in the Antarctic Peninsula and up to -5 to -7°C in East Antarctica. Maximum thickness of the permafrost is over 100 m and the active layer is 0.3 to 1 m. The ground ice is mainly pore ice but in some places there is buried glacier ice in moraine. Frost action, frost sorting, creep, nivation, and pre-glacial slope forms are developed in the coastalcontinental zone.

The intercontinental zone consists of continuous permafrost only. The mean annual temperature of the permafrost there is -10°C or lower. Thickness of this permafrost is several hundreds of metres and of the active layer is 0.1 to 0.5 m. The ice content is low with only small ice forms. The cryogenic structure of the permafrost is the same as the coastal-continental zone. Because alternate freezing and thawing is infrequent in the intercontinental zone all cryogenic processes are less rapid than in the other zones. Thermokarst is not found there and 1n general the intercontinental zone is not as well studied as the two other zones.

Trilateration and single line surveys have been made to about 900 km inland of Casey, Wilkes Land, to measure surface elevation, ice thickness, horizontal velocity, and other parameters. On the large scale the velocity U increases smoothly from 8 m a−1, 800 km inland, to 280 m a−1 inland of the fast outlet streams. This increase in velocity is associated with a corresponding increase in the large-scale smoothed (over about 30 ice thicknesses) basal shear stress τb from 0.4 to 1.5 bar. The mean shear strain-rate through the ice sheet U/Z = kτb4 , where Z is the ice thickness (range 4 500 to 1 700 m).

At scales of one to several ice thicknesses large variations occur in surface slope and ice thickness without proportionally large velocity variations, because of the effect of the longitudinal stress. Detailed measurements made over a 30 km section indicated that the surface longitudinal strain-rate gradient varied from -1.7 to +1.3×l0−6 a−1 m−1 along with variations in surface slope of from -3.5 to +1.5%.

A multilayer model, based on the solution of the biharmonic equation for the stream function, was used in a study of the ice flow associated with these surface undulations. Given the bedrock topography and large-scale flow parameters, the model closely predicted the measured surface profile when the variation of the surface accumulation rate over an undulation was also considered.

During the Norwegian Antarctic Research Expeditions 1967–77 and 1978–79 a land party worked on Riiser-Larsenisen, an ice shelf 120 km wide on the coast of Dronning Maud Land (15° to 20°W). In February 1977 three patterns of six stakes each were laid out over areas of about 4 km2 as part of a combined investigation into absolute movements and strain-rates. Eight stakes for absolute movement were also laid out. The stakes were also used for determination of mean snow accumulation together with a stake line across the ice shelf. The positions of the stakes were determined by theodolite observations from the stake points and from trigonometric stations located on land and on an ice rise near the ice front.

In February 1979 all these measurements were repeated, and absolute movement, deformation, and mean accumulation were calculated. The absolute velocity of the ice shelf varied from 30 m a−1 near the grounding line to 130 m a−1 near the ice front. The mean annual accumulation was 510 mm water equivalent on the outer part of the ice shelf and 580 mm at the grounding line. Based on these measurements, together with snow-density measurements down to 16 m and measurements of the height of the ice shelf, the mass balance of the ice shelf was studied.

The stake pattern across the grounding line showed considerable expansion. This is interpreted as a result of water freezing in bottom crevasses made by the tides.

In 1979 snow temperatures were measured in eight bore holes down to 10 m depth. The snow temperatures at 10 m depth were used as a measure of the annual mean air temperature. The annual mean air temperature ranged from -16.8°C near the ice front to -19.2°C at the grounding line. Snow temperatures in bore holes on the slope of the ice rise and inland indicate a mean atmospheric temperature inversion of 0.28°C 100 m−1 for a 440 m layer near the grounding line, and 0.30°C 100 m−1 for a 160 m layer near the ice front. Here a mean inversion of 2.8°C 100 m−1 was found for the lowest 40 m layer of the atmosphere.