An accurate knowledge of the thermal properties of firn and ice within a glacier is essential for any reliable mathematical model of heat transfer. This paper considers the problem of determining the thermal properties of firn at Dome C, Antarctica, for use in such a model.

First, the difficulties in accurately determining thermal properties are discussed. Then a physical experiment which can be performed under field conditions, but which will yield a well-posed mathematical problem for determining the unknown properties, is presented. Next, two different numerical techniques for solving the mathematical problem are discussed. Finally, some numerical approximations and error estimates are presented for the results of applying our numerical procedure to data from Dome C. Although insufficient data were obtained to test our methods fully, we have established a measurement procedure and a method of analysis which appear to be promising.

The Antarctic ice sheet has almost no net annual ablation on its surface, so most mass losses are by iceberg calving along its perimeter, which may be either grounded in shallow water or floating in deep water. An ice cliff forms along the perimeter in both cases. Wave action undercuts ice margins in the tide-water zone along beaches, and causes coastal calving if the rate of undercutting compares with the forward ice velocity. If the ice velocity is sufficiently greater, the ice sheet advances into deeper water and becomes a float at depths of 200 to 300 m (Robin 1979). A floating ice shelf then forms and icebergs calve along the ice front. Iceberg calving along this ice front may be due to several causes (Holdsworth 1977,Robin 1979). Since iceberg calving, either from ice shelves or in the tidewater zone of beaches between ice shelves, is the principal ablation mechanism of the Antarctic ice sheet, it is important to understand calving dynamics quantitatively. This paper presents the results of a finite-element examination of calving along floating margins of the ice sheet.

Variation of the microparticle concentration in an ice core from Mizuho station, East Antarctica, does not show the annual cycle that has been demonstrated for ice cores from Antarctica and Greenland by other authors. Possible reasons for the lack of a annual cycle are considered and two causes are suggested. (1) Semi-annual variation of microparticle concentration as observed in drift-snow. The low particle concentration in March and in August to October is not due to minor particle transport, but to the dilution of microparticles transported mainly from arid regions in the southern hemisphere through the troposphere by falling snow. (2) Discontinuous surface-layer formation. A seasonal or an annual sequence of the variation in microparticle concentration in surface snow layers may be interrupted by the absence of surface snow-layer formation.

Observations during February and March 1980 of structures in 66 separate floes in Weddell Sea pack ice show widespread occurrence of frazil ice in amounts not previously reported in sea ice of comparable age and thickness in the Arctic. It is estimated that as much as 50% of the total ice production in the Weddell Sea is generated as frazil. Average floe salinities also appear higher than those of their Arctic counterparts. Comparative studies of fast ice at 28 locations in McMurdo Sound show this ice to be composed almost entirely of congelation ice that exhibits crystalline textures and orientations that are similar to those observed in Arctic fast ice. However, average fast-ice salinities in McMurdo Sound are higher than those reported for Arctic fast ice of comparable age and thickness.

A comparison of the crystalline texture and fabric of the two deep cores to bedrock from Camp Century, Greenland, and Byrd station, Antarctica, reveals striking similarities. Each core exhibits a fabric profile which progresses from depositional fabrics at shallow depths through multi-maxima transitional stages into single maximum distributions. The major difference between the two cores occurs near the base where the Camp Century fabrics maintain a high degree of orientation while the Byrd station fabrics break up into a diamond pattern. The most remarkable similarity between the two profiles occurs at the glacial-interglacial transition where crystal sizes decrease and crystal orientations strengthen significantly over a very short interval. Similar changes occur in other deep ice cores, thus indicating the possibility that the late Wisconsin ice has a unique crystallographic signature.

Some results from a dynamic-thermodynamic simulation of the seasonal cycle of the Weddell Sea pack ice are described. The model used for the study is similar to that developed by Hibler (1979) for a numerical investigation of the Arctic ice cover. It employs a plastic ice rheology coupled to a two-level ice-thickness distribution. The thickness characteristics evolve in response to ice dynamics, and to ice growth and decay rates dictated by surface heat-budget calculations and by heat storage in a fixed depth oceanic boundary layer. Observed time-varying wind, temperature, and humidity fields based on the 1979 Australian analysis are used together with empirical radiation fields and fixed ocean currents to drive the model. Employing these fields, the model is integrated over two seasonal cycles. The second-year results are compared to observed ice-drift data obtained by Ackley (1981[b]) and to ice-edge characteristics determined from satellite imagery. In analyzing the results, particular attention is paid to ascertaining the relative roles of ice dynamics and thermodynamics in determining the ice-edge characteristics. A sensitivity analysis of the effect of a different parameterization of the atmospheric temperature and humidity fields is also carried out.

Overall, the results suggest that (1) ice dynamics are essential in describing the seasonal cycle, and (2) a feedback between the atmospheric temperature and the presence of ice may be a major cause of the rapid decay of the Antarctic ice cover during the spring-summer period.

Recent measurements conducted on Erebus Glacier Tongue (EGT) have enabled improvements to be made to previous analyses of earlier data (Holdsworth 1974). The use of LANDSAT imagery and controlled aerial photography up to 1978 have enabled glacier flow rates and hence strain-rates to be determined more accurately than before. Re-analysis of the stress and strain-rate data leads to flow law constants which are now more compatible with those of Thomas (1973). The thinner ice appears to be “softer” than the thicker ice. Bending of EGT appears to be the result of current pressure exerted on the south side of the glacier, and computed bending strains seem to account for most of the observed curvature of EGT.

Basal ice melting predominates over any possible refreezing; melt rates along the south edge exceed basal melt rates by a factor of 5-10. Short-period, pseudo-elastic deformations occurring in EGT as a result of ocean wave-induced motion may be responsible for continually activating crevasses and producing the apparent “softening” effect in the creep of the thinner ice. Such oscillations may also lead to the calving of EGT although recent observations suggest that a hinge-line calving due to horizontal bending of EGT may be just as likely.

The mostly terrestrial East Antarctic ice sheet is ten times larger and probably more stable than the mostly marine West Antarctic ice sheet. It is natural to suppose that the former appeared first, and that perhaps the latter was partially formed from an outflow of East Antarctic ice onto the West Antarctic continental shelf. Alternatively, the largely marine West Antarctic ice sheet may have appeared first, provided that post-Eocene continental drift allowed heavy snow precipitation over the West Antarctic island archipelago, but prevented warm South Pacific ocean currents from entering. Sea ice could then thicken and ground from surface accumulation and basal freezing, creating broad marine ice domes in shallow West Antarctic embayments. These marine ice domes could then merge with ice caps on the islands to create a marine ice sheet that could expand into East Antarctica and merge with highland ice caps to form the terrestrial East Antarctic ice sheet. A Cenozoic glacial history of Antarctica is outlined, based on this marine ice transgression hypothesis.

It is proposed that an ice shelf disintegrates when its calving front retreats faster than its grounding line. This paper examines the role of ice thinning in grounding-line retreat. Thinning occurs as a result of creep spreading and ice melting. Thinning by creep is examined for the general regime of bending converging flow in an ice shelf lying in a confined embayment, and at the grounding lines of ice streams that supply the ice shelf and ice rises where the ice shelf is grounded on bedrock. Thinning by melting is examined at these grounding lines for tidal pumping and for descent of surface melt water into strandline crevasses, where concentrated melting is focused at the supposed weak links that connect the ice shelf to its embayment, its ice streams, and its ice rises. Applications are made to the Ross Ice Shelf.

Total gas contents of ice cores together with temperature estimates derived from 180/160 and 2D/1H values have been used to separate topographic and climatic changes in the deposition temperature history of the ice (Raynaud 1977, Jenssen 1978). The most recent analysis (Jenssen in press) made use of two linear relationships (one purely empirical, the other established empirically but subsequently justified theoretically) to derive an algebraic expression for the change of surface temperature with ice-sheet elevation. A physical line of reasoning is presented which instead infers the climatic history from changes in the surface topography of the ice sheet. This suggests that a complete interpretation of core data must go hand in hand with ice-sheet modeling.

For the investigation of time-dependent vertical deformation of upper layers of ice shelves loaded by accumulated snow and/or structures a large-scale field test was started in 1980-81 at Georg von Neumayer station (70°37´S, 8°22´W), Antarctica. Specially developed base plugs were fixed to the bottoms of 5.0 and 10.0 m deep bore holes by melting and refreezing the surrounding firn. The deformations of the snow firn layers will be measured between these base plugs and metal plates placed on the snow surface during installation. The bore holes are kept open by tubes of different diameters in order to allow for load-induced vertical deformations of snow and firn. Some measurements are directly affected by the presence of the station. The design of the measuring device, the procedure of installation, and the measurement of deformation are described. The results of calculations of deformation and measurements of first settlement are presented.

In the 1979–80 field season, a site survey was carried out in the southern Weddell Sea for the construction of a wintering station. The survey comprised glaciological investigations on the Filchner and Ronne ice shelves, such as ice movement, deformation, ice thickness, calving, accumulation, and snow temperature measurements which are relevant to mass-balance studies. Along with the investigations on the ice, oceanographical and bathymetrical observations were carried out along the ice front. The field data and observations lead to assumptions about the advance of the ice shelf as well as about the mass loss at the frontal zone due to calving and bottom melting. The ice-front velocity at 50°W is about 1 070 m a−1 and the bottom melting is estimated to be 3.2 m a−1 at 20 km south-west of the ice front.

Infiltration of brine into the McMurdo Ice Shelf is dominated by wave-like intrusions of sea-water triggered by periodic break-outs of the ice front. Observations of a brine step 4.4 m in height in the McMurdo Ice Shelf show that it has migrated about 1.2 kin in four years. The present brine wave is overriding an older brine-soaked layer. This migration is proof of the dynamic nature of the step, which is the leading edge of a brine wave that originated at the ice front after a major break-out of the McMurdo Ice Shelf. The inland boundary of brine penetration is characterized by a series of descending steps that are believed to represent terminal positions of separate intrusions of brine of similar origin. The inland boundary of brine percolation is probably controlled largely by the depth at which brine encounters the firn/ice transition (43 m). However, this boundary is not fixed by permeability considerations alone, since measurable movement of brine is still occurring at the inland boundary. Freeze-fraction-ation of the sea-water as it migrates through the ice shelf precipitates virtually all sodium sulfate, and preferentially concomitant removal of water by freezing in the pore spaces of the infiltrated firn produces residual brines approximately seven times more concentrated than the original sea-water.

Occurrences of tephra layers in four ice cores from Antarctica are reviewed. A new tephra, informally named the Vostok tephra, is described from a depth of 100.85 m in an ice core from Vostok station. Ice associated with four tephra layers at depths between 1390 and 1450 m in the Byrd station ice core shows increased levels of sulfate and nitrate which correlate well with peaks in particle concentrations. High levels of sulfate and nitrate are also associated with the Vostok tephra.

Tephra offer great potential as stratigraphic markers and should be useful in providing time planes as well as assisting in correlation between widely spaced ice cores.

Summer profiles of sea-water temperature, salinity and flow were obtained on George VI Ice Shelf near its northern ice front. At each depth, temperature salinity and density show little variation between sites. Their respective variation to 250 m depth confirms a linear temperature-salinity dependence. This is the first place in the world where observations confirm precisely the form of the T-S diagram predicted for fresh ice melting in sea-water. Both tidal and residual flow are small, except at the western margin of the ice front, where a strong outflow is concentrated immediately beneath the ice shelf. The observations lead to a simple circulation model for the ice-shelf regime. Warm Deep Water flows southwards into George VI Sound, replacing the colder water that spreads northwards in the surface outflow. Thermohaline exchanges beneath the ice shelf determine the salinity profile, which itself provides evidence of upwelling. Estimates can be made of the basal melt rate of the ice shelf. The rates vary from around 10 m a−1 at the ice front to an average value for the ice shelf of order 1 m a−1. The average value is consistent with earlier estimates from surveys of ice-shelf strain.

Antarctic ice-elevation maps are presented, based on 4 000 measurement points, analysed from traverses made by constant-density balloons. These balloons were launched during the Tropical Wind, Energy Conversion and Reference Level Experiment (TWERLE). The TWERLE data yielded daily maps of the elevation of the 150 mbar pressure surface in the southern hemisphere. These maps serve as a reference level from which the altimeter readings are subtracted to yield the ice elevation.

The accuracy of the ice-elevation data is estimated at better than 60 m. The accuracy of the contouring depends on the density of measurements in the particular area. Two contouring techniques were used. One is based on generating elevation data at fixed grid-points from arbitrarily distributed measurements, using Cressman's method. The second is a surface spline interpolation technique by Duchon.

The elevation data are of special value on the inland plateau areas. There, the typical distance between balloon measurements is sufficiently small, relative to the typical distance between contour lines at 100 m intervals. Furthermore, the inland areas are the least accessible to other mapping techniques.

Finite-element simulations of the Ross Ice Shelf flow yield good qualitative agreement with observation, but velocity magnitudes tend to be too low up-stream of ice rises and too high near the ice front. These errors will be significantly reduced by better model convergence to the flow law, more accurate values of the flow-law parameters, better resolution of the ice streams, and inclusion of subgrid scale stress variability produced by crevassing and ice-fabric orientation. A model run featuring a hypothetical ice rise down-stream of Crary Ice Rise indicates that the distinctive strain-rate patterns produced by an isolated ice rise can be cancelled by neighboring ice rises. This suggests that an isolated ice rise might migrate up-stream along sea-bed ridges unless there are neighboring ice rises acting to stabilize the surrounding ice-thickness tendencies.

Static electrical conductivity (SEC) and dielectric measurements were made on Antarctic ice cores at -15°C. The observed behaviour is markedly different from that of temperate glacier ice. The influence of impurities on the electrical properties of polar ice is discussed. Variations of SEC do not appear to be correlated with variations in the relaxation frequency when the loss factor is maximal. Different growth conditions coupled to ageing effects in the first few metres of the firn may explain variations in the electrical properties along an ice core. Moreover, it is suggested that the relaxation frequency increases as the specific surface area increases. Defects are produced in the firn near the surface and may be retained in the subsequent ice. Discontinuous changes in the high-frequency electrical conductivity between adjacent layers can induce internal reflections of radio waves within ice sheets.

It has been suggested (Hammer 1977, Gudmandsen and Overgaard 1978, Millar 1981) that layers of acidic ice, formed after large volcanic eruptions, are a cause of radio echo layering. The power reflection coefficient (PRO of these reflecting layers may be related to their acid content and thickness. The present paper makes estimates of ice acidity from radio echo-sounding (RES) and compares them with measured values along Greenland ice cores. Dated Antarctic PRC/depth profiles are also presented, and acidity estimates made which give a crude indication of past volcanic activity in the southern hemisphere to ~150 ka BP.

A detailed survey has been made of outlet glaciers between longitudes 48°E and 62°E. Surface and bedrock profiles were constructed from radio echo-sounding data and glacier grounding lines were located by changes in the characteristics of bedrock echoes. A preliminary investigation of the dynamics of these fast-moving (~1 km a−1) glaciers is presented. From the measurements, the ice-mass flux through each glacier can be calculated. Mass-budget studies inland (Morgan and Jacka 1981) are thus extended to the coast. In the inland region, an excess of accumulation over outflow of ~100% was estimated, while a near balance condition appears likely for the coastal region