Visible and near-infrared spectral reflectances of snow and superimposed ice were measured in Dronning Maud Land, Antarctica, during the 1992-93 austral summer. Spectral-reflectance curves of both snow and superimposed ice remain high ( > 80%) in the visible region. A pronounced decrease in reflectance appears in the near-infrared, especially for superimposed ice. Superimposed ice with a 1 cm thick surface layer of ice-bound snow crystals had a considerably higher reflectance than superimposed ice containing only a few snow crystals. Furthermore, these data prove that snow and superimposed ice reflect solar radiation specularly and suggest that the anisotropy strengthens with increasing wavelengths. Integrated in-situ reflectances corresponding to Landsat TM bands 1-4 show that TM band 1 is least affected, whereas TM band 4 is most affected by anisotropy. Furthermore, the anisotropy increases with increasing off-nadir viewing angles up to an angle corresponding to 90°-θs(θs= solar elevation). For a 15° off-nadir sensor-observation angle, the average snow reflectance for TM bands 1--4 is about 10% higher than at nadir. Similarly, the apparent reflectance can be more than 50% higher than the nadir reflectance for larger observation angles. Consequently, if satellite-derived reflectances are going to be considered as absolute values, a topographic-correction model is needed to correct for the effects of anisotropy.

Two uncontrolled mosaics were assembled at an ~1 : 1000 000 scale, covering for the first time almost the entire Larsen Ice Shelf area, using Landsat TM images of 1986-89 and the Kosmos KATE-200 photographic products of late 1975. By comparing them, it was possible to estimate the change along the 600 km north-south ice-shelfs seaward margin where substantial calving has occurred in the recent past. In overall extent, the ice shelf has decreased by ~9300km2 since 1975. The interpretation of the available satellite data also allowed estimates of the ice-shelf’s inland boundary and detection of new surface ice features, especially in the very disturbed region south of Jason Peninsula, which may help to explain the significant calving events occurring in this region. Estimates of surface velocities were derived by photographically co-registering sequential imagery in the chaotic rifted-crevassed area east of Kenyon Peninsula, where conspicuous ice features could be clearly detected even after a 10.4 year time interval. The Landsat TM mosaic, though uncontrolled, provides a basic data set for future studies of global change of the major ice shelf on the Antarctic Peninsula which is probably very sensitive to any climatic trend.

Previous studies have established that, together with the development of a preferred crystal-orientation fabric in ice undergoing creep deformation to high strains, there also develops a tertiary equilibrium crystal size, i.e. the crystal size, rather than affecting the creep rate, is a result of the deformation to large strains.

Equilibrium crystal size is considered here as a “balance” between crystal growth with time as a function of temperature and crystal change as a result of temperature dependent deformation. The temperature effects in these two processes (Arrhenius relation) are similar and consideration of the activation energies for the two processes indicates that it may be appropriate to cancel them, yielding a dependence of equilibrium crystal size on stress alone. The results from laboratory experiments of steady-state crystal size plotted as a function of stress support the above proposition.

The possibility of using the relationships between steady-state crystal size and deviatoric stress as a polar ice-mass piezometer is discussed.

Passive microwave brightness temperatures from the Special Sensor Microwave Imager (SSMI) are studied together with surface air temperatures from two Automatic Weather Stations (AWS) for the year 1989. One station is located on the East Antarctic plateau (Dome C) and the other on the Ross lee Shelf (Lettau).The satellite data for frequencies 19, 22 and 37 GHz with vertical polarization,centered on the two AWS stations, are studied. A simple thermodynamic model and asimple radiative-transfer model, that takes into account the snow temperature profile and assumes a constant annual emissivity, are proposed. The combination of these two models enables us to compute extinction coefficients, penetration depths and toretrieve the measured brightness temperature variations from the AWS surface temperatures. Afterwards, this model is reversed in order to retrieve the snow-surface temperatures from the satellite data. Results are promising but strong approximationsand a priori knowledge of the extinction coefficient are still needed at this point.

The mean rate-of-change of the surface elevation on lower Lambert Glacier is measured with satellite-radar altimetry from the Geosat Exact Repeat Mission (ERM) (1987 89) and Seasat (1978), using orbit cross-over analysis. The measurement region extends 190km in the along-flow direction from 72.1° to 70.4° S, an area that includes the grounding zone of Lambert Glacier. The Geosat ERM and Seasat altimetry are referenced to GEM T2 orbits. The random-noise levels in the data are reduced by low-pass filtering. The orbit bias between the two satellites is estimated from cross-over differences over sea ice seaward of the calving front of the Amery Ice Shelf. The results show a mean rate of increase of the surface height on lower Lambert Glacier of 31±10 mm year-1 during the decade from 1978 to 1987–89. The computation is also carried out independently using data not low-pass filtered but with orbit bias minimized by adjusting the Seasat and Geosat ERM orbits into a common ocean surface. The latter analysis shows a mean rate of increase of the surface height of 83 ± 9 mm year-1 during the same time period. Cross-over analyses carried out using altimetry not low-pass filtered and with orbits not adjusted into a common ocean surface yield intermediate results. Taken together, the cross-over analyses show that the surface height on lower Lambert Glacier increased at a mean rate within the range 20–90 mm year-1 during the decade 1978 to 1987–89.

The depth, density, load and isotopic (δ18O) composition of the snow cover on pack-ice floes were measured in late austral summer 1992 in the eastern Ross, Amundsen and western Bellingshausen Seas. Snow-density values commonly exceeded 350 kg m-3 and some were as high as 500 kg m-3. The densification of the snow occurs quickly and is attributed to a windy environment. The high density and sometimes considerable depth of the snow on the floes accounts for loads of as much as 700 kg m-2 and resultant sea-water flooding of the underlying sea ice. Lower mean δ18O values in the Ross/Amundsen Seas snow cover suggest that the region might have a cooler climate than the Bellingshausen Sea region. Snow depths on floes in the Bellingshausen Sea region were lower than those in the Ross/Amundsen Seas region, because the Bellingshausen Sea floes were first-year ice. Possible annual units in the isotope profiles of snow, as much as 2m deep, indicate that floes in the Ross/Amundsen Seas region were 2-3 years old.

Landsat Thematic Mapper (TM) images were used prior to the 1991–92 field season to infer the position of the grounding line at the mouths of Ice Streams D and E, West Antarctica. Our field plan for mass-balance studies was based on this determination, and thus the imagery played a central role in both the scientific and logistics planning. A radar profile along the flow direction was made across the inferred grounding line at one location, and ice-thickness measurements together with surface surveying enable us to compare the hydrostatic surface and the actual topography to determine the point at which the ice becomes grounded. The profile transits from floating to grounded ice at the same location as the grounding line inferred from the imagery. Changes in the radar-echo strength also occur at this location, giving further support to this interpretation. Tilt studies of the ice flexure caused by tidal variations at locations on either side of this grounding point give additional evidence that grounding is occurring close by. The combination of these three measurements therefore confirms the grounding-line location derived from the satellite imagery and supports the use of this technique to determine grounding.

Satellite radar-altimeter data from Seasat (1978) and the Geosat Exact Repeat Mission (1987–89) are evaluated to investigate the question of advance or retreat of Lambert Glacier, Amery Ice Shelf, East Antarctica. New maps based on a fine-scale 3 km grid arc calculated using ordinary kriging. The break in slope at the 100 m elevation contour, relative to the WGS 1984 ellipsoid, is taken as a proxy for the grounding line. Measurements indicate that the irregular grounding line, which includes shoals, advanced approximately 10km between 1978 and 1987‐89, corresponding to a mean advance rate of about 1000 m year-1.

The launch of ERS-1 provides coverage, by satellite altimetry, of 80% of the Antarctic ice sheet, allowing topographic mapping of areas which previously had a dearth of accurate elevation data. Four 35 d repeat cycles of fastdelivery altimeter data were used in this study, comprising a total of approximately 1000000 height estimates. About 40% of these were rejected during a careful filtering procedure designed to remove erroneous values caused by poor tracking or complete loss of the returned echo. The OSU-91A geopotential model was used to convert ellipsoidal elevations to geoidal values. Corrections for surface slope were applied and a Digital Elevation Model (DEM) was produced with a grid spacing of 20km.

The precision of the data was assessed from an analysis of crossing points of ascending and descending tracks. For 43864 cross-overs, the standard deviation was 6.8m. Regional biases associated with geoid, orbit and topography-induced errors reduce the accuracy of the height measurements. This was assessed by a comparison with ground-survey data. The DEM was compared with a 700km levelling survey, with an accuracy ranging from 1 to 5m, from the Lambert Glacier basin region (≈73° S, 55° E). The mean difference was found to be-1.6m with a standard deviation of 14m. A similar result was obtained for a 600km traverse line in Wilkes Land (75° S,≈1l0° E).

The DEM was then compared with a digitized version of the Scott Polar Research Institute (SPRI) Antarctic folio map. This map was derived from orthometric measurements of surface elevation, primarily from pressure altimetry. Differences in excess of 300 m were observed between the two data sets. Only 37% of the region covered showed agreement to better than 50m, and a significant proportion ofthis was composed of the Ross and Filchner-Ronne Ice Shelves. The largest discrepancies occurred in marginal areas where there is poor coverage by both satellite altimetry and terrestrial data. Inland, significant differences were also found.

Glacio-geological field work and radar ice-thickness sounding were carried out in the area between David and Mawson Glaciers. A subglacial topographic map has been compiled from radio-echo-sounding data. The northern part of this map shows that the trench of David Glacier reaches a depth of more than 1000 m below sea level. The area south of David Glacier comprises a landscape of nunatak clusters dissected by glaciated valleys with ice thicknesses as much as 800 m. Subglacial cirques occur at the outer margins of the nunatak clusters. A model for the regional glacial history is proposed. It starts with a major deglaciation in the Pliocene, which results in marine transgression in basins west of the Transantarctic Mountains. During the late Pliocene, the ice advanced towards the northeast, depositing a thin layer of (Sirius Group) till containing reworked mid-Pliocene marine diatoms. Due to accelerated mountain uplift, the ice cut iIlto the pre-Pliocene peneplain, eroding broad valleys. A period of ice-sheet retreat followed to expose a landscape of large nunataks separated by wide valleys. During this period, local cirque glaciation occurred. When the ice sheet advanced again, another phase of uplift forced the glaciers to cut deeper into the valleys. Probably since the Last Glacial Maximum the ice surface has lowered by about 100 m.

The small-scale variability of physical properties and structural characteristics of multiple pairs of fast-ice cores obtained during the austral summer of 1991-92 at two Antarctic sites, McMurdo Sound (MCM) and Pine Island Bay (PIB), are examined and discussed with respect to the growth and decay of the sea ice. The ice at the MCM site was thicker than that at the PIB site and was covered by a somewhat thinner snowpack. While mean salinity and temperature of the ice at the two sites were similar, small-scale variations in both salinity and temperature were greater at PIB than at MCM. The ice sheet at MCM was a two-layer medium consisting of congelation ice overlying platelet ice. The ice from the PIB site, however, was composed of mainly frazil ice and layers of congelation ice with occasional thin layers of snow-ice at the surface of the cores. Crystal sub-structure measurements, c-axis orientation and brine-layer spacing from the MCM cores revealed that the congelation ice had moderately aligned, horizontally oriented c axes, suggesting that east-west currents off the southwest tip of Hut Point Peninsula control crystal-growth orientation.d others: Variability of physical and structural characteristics of Antarctic fast ice

A numerical study has been carried out on the flow of ice in the transition zone between an ice sheet and an ice shelf. The study was motivated by the need for global ice-sheet-ice-shelf modelling to determine the characteristics of the transition zone. The problem is dealt with from an academic viewpoint, and the study especially focuses on two-dimensional vertical sharp transition zones. Stokes equations are solved using a finite-element method. Conclusions include: (1) in ice-sheet-ice-shelf modelling, each of the two components can be computed separately then linked by a jump-boundary condition [or the horizontal velocity; (2) as shown by studies on the response of an ice shelf to tidal forcing, the surface elevation/thickness ratio passes through the hydrostatic equilibrium value.

SEM observations of microparticles in ice-core samples retrieved by the Japanese Antarctic Research Expedition in east Dronning Maud Land have been carried out since 1987. Morphology and elemental composition by EDS of many microparticles taken from various depths of the 700 m Mizuho ice core were compared with each other and with those of stratospheric microparticles in NASA Cosmic Dust Catalogs and microparticles hitherto found in deep ice cores retrieved in Antarctica. Number concentrations of microparticles were measured on all samples throughout the 700 m Mizuho ice core. Remarkable fluctuations found in the depth profile of the concentration seem to coincide with cold climates indicated by δ18O of the same core. Compositional analysis of volcanic ash at a depth of 500.7 m in the Mizuho ice core, dated at approximately 6000 years BP, indicates that the ash originated from the South Sandwich Islands.

The causes and nature of ice-sheet radio-echo internal reflections at deep layers in polar ice sheets are discussed, based on the dielectric properties of ice that have been measured at microwave frequency and radio frequency. The reflection coefficients of electromagnetic waves in ice sheets due to two causes the change in permittivity induced by changes in crystal-orientation fabrics with depth, and changes in conductivity induced by changes in acidity with depth - were derived respectively as a function of the frequency used in radar sounding and the temperature of ice, and both were compared quantitatively. It is shown that at single-plane boundaries the reflection coefficients due to the former cause are independent of frequency and temperature and that they are large enough to produce dominant internal reflections. In contrast, reflection coefficients due to the latter cause strongly depend on frequency and temperature. Since they are inversely proportional to the frequency, the latter cause can be dominant only when frequencies below about 60 MHz are used. Examination of previous observational data has suggested that not only changes in acidity but also changes in crystal-orientation fabrics exist at depths corresponding to the dates of earlier volcanic eruptions.

Two empirical equations for firn densification have been obtained,considering firn porosity as a function of overburden pressure. In the first equation, thereduction ratio of porosity in firn is assumed to be proportional to the increasing ratioof overburden pressure and the mth power of the porosity. The porosity exponent m should be close to -2, so as to have a best-fit with 14 depth-density profiles fromGreenland and Antarctica. In the second equation, the reduction ratio of porosity wasassumed to increase proportionally to the increment of overburden pressure and thenth power of the porosity. The most satisfactory values of the exponent range from -1 to 1. It has been suggested that firn density, determined primarily by overburdenpressure and firn temperature, contribute to a lesser degree.

The air-bubble formation process has been studied experimentally by using five ice cores from the Greenland and Antarctic ice sheets. Bubble volumes in firn-ice samples were measured by a classical method based on Boyle Mariotte's law for an ideal gas. It was found that the bubble volume varies with depth as a function of bulk density in the firn-ice transition layer, which is represented by an exponential function of firn density. Air bubbles start to form rapidly at a bulk density of 0.763–0.797 Mg m-3. This density (ρib) seems to be correlated with the ice temperature in the ice sheets; ρib increases with a decrease in the ice temperature. Vb shows the maximum value in the density range 0.819–0.832 Mg m-3. The corresponding porosity of the density ranges between 0.110 and 0.097. This porosity does not seem to correlate with ice temperature or accumulation rate at the coring site. These characteristics of firn densities probably affect the amount of entrapped air in glacier ice (total air content) in polar ice sheets.

Velocity measurements were made for two time intervals on the Smith Glacier ice tongue (1973–88 and 1988–90) and three on the Dotson Ice Shelf (1972–88, 1973–88 and 1988–90). The Smith Glacier ice tongue velocities for the two intervals are similar near the grounding line but show a progressive increase toward the terminus in the late 1980s. The Dotson Ice Shelf velocities remained virtually constant during all three time intervals. The increased velocities of the Smith Glacier ice tongue may be attributed to a general loss of densely packed icebergs that buttressed the terminus during the 1970s but drifted out to sea during the late 1980s. The Smith Glacier ice tongue receded as much as 10 km between 1973 and 1988, the Dotson lee Shelf 5–7 km in the same time period. Similar observations of drifting and ca1ving were made for the adjacent Thwaites Glacier ice tongue. The cause of the loss of ice in the region is unknown but it may have been a change in winds or a warming of the air or water during the late 1980s.

A hot-water drill developed during the past 2 years at Alfred-Wegener- Institut was used to penetrate Ekströmisen several times near the German wintering-over base, Neumayer (70°39´S, 08° 15´W). The drilling operation was very successful and the initial large diameter (≥ 35 cm) allowed easy access to the ocean. One hole was used to install an ultrasonic echo-sounder which recorded the ablation at the ice-shelf bottom continuously. Another hole was destined for emplacement of a thermistor string throughout the 237 m thick ice shelf for ice-temperature measurements. Several CTD profiles in the 175 m deep water column and the analysis of water samples provided valuable data for the understanding of ice-shelf-ocean interactions.

A continuous ECM profile (strong acid concentration) has been measured along the 2191 m of ice core recovered at Byrd Station, Antarctica, in 1968. The ECM profile reveals continuous and systematic seasonal changes which are used for dating the ice core back to 50 000 bp. Hammer and others: ECM stratigraphic dating of Byrd Station ice core, Antarctica

The accumulation and distribution of the 2H content of near-surface layers in the eastern part of the Ronne Ice Shelf were determined from 16 firn cores drilled to about 10 m depth during the Filchner IIIa and IV campaigns in 1990 and 1992, respectively. The cores were dated stratigraphically by seasonal δ2H variations in the firn. In addition, 3H and high-resolution chemical profiles were used to assist in dating. Both the accumulation rate and the stable-isotope content decrease with increasing distance from the ice edge: the δ2H values range from about 195‰ at the ice edge to -25‰ at BAS sites 5 and 6, south of Henry Ice Rise, and the accumulation rates from about 210 to 90 kgm-2 a-1. The δ2H values of the near-surface firn and the 10 m firn temperatures (Θ) at individual sites are very well correlated: dδ2H/dΘ = (10.3 ± 0.6)‰K-1; r = 0.97.

The δ2H profiles of the two ice cores BI3 and BI5 drilled in 1990 and 1992 to 215 and 320 m depth, respectively, reflect the gradual depletion in 2H in the firn upstream of the drill sites. Comparison with the surface data indicates that the ice above 142 m in core BIS and above 137 m in core BI3 was deposited on the ice shelf, whereas the deeper ice, down to 152.8 m depth, most probably originated from the margin of the Antarctic ice sheet.