An old portable 60 MHz radar has been upgraded with a new digital data-processing and -acquisition system and a new antenna construction enabling a fast and low-cost installation on a Twin Otter aircraft. Augmented by a laser altimeter and kinematic global positioning system (GPS), the system has the capability of acquiring accurate data on location and ice-surface elevation, and adequate-quality data on ice thickness. The system has been applied successfully in mapping the Nioghalvfjerdsfjorden glacier, northeast Greenland, in spite of the difficult conditions with melting water on the glacier surface. The measurements from the floating part of the glacier have been evaluated by comparison of radar data with laser-altimeter and in situ measurements.

The spatial distribution of accumulation across Siple Dome, West Antarctica, is determined from analysis of the shapes of internal layers detected by radio-echo sounding (RES) measurements. A range of assumed accumulation patterns is used in an ice-flow model to calculate a set of internal layer patterns. Inverse techniques are used to determine which assumed accumulation pattern produces a calculated internal layer pattern that best matches the shape of internal layers from RES measurements. All of the observed internal layer shapes at Siple Dome can be matched using a spatially asymmetric accumulation pattern which has been steady over time. Relative to the divide, the best-fitting accumulation pattern predicts 40% less accumulation 30 km from the divide on the south flank of Siple Dome and 15–40% more accumulation 30 km from the divide on the north flank. The data also allow the possibility for a small time variation of the pattern north of the divide. The mismatch between the calculated and the observed layer shapes is slightly reduced when the accumulation rate north of the divide is higher in the past (> 5kyr BP) than at present. Sensitivity tests show that the predicted change in the spatial accumulation pattern required to cause the slight Siple Dome divide migration (inferred from previous studies) would be detectable in the internal layer pattern if it persisted for > 2 kyr. Our analysis reveals no evidence that such a change has occurred, and the possible change in accumulation distribution allowed by the data is in the opposite sense. Therefore, it is unlikely that the Siple Dome divide migration has been caused by a temporal change in the spatial pattern of accumulation. This conclusion suggests the migration may be caused by elevation changes in Ice Streams C and D at the boundaries of Siple Dome.

The characteristics and sensitivities of a cold-based glacier on the Tibetan Plateau, where the summer monsoon provides most of the mass input to glaciers, are discussed using an energy-balance model incorporating the process of water refreezing. The model accurately represents the observational results related to the mass balance of Xiao Dongkemadi glacier on the central plateau during 1992/93. Our data revealed that the mass balance of cold glaciers cannot simply be described by the surface mass/heat balances, because about 20% of infiltrated water is refrozen and thus does not run off from the glacier. Model calculations demonstrate that glaciers in an arid environment can maintain their mass since the monsoon provides precipitation during the melting season. Snowfall in summer keeps surface albedo high and largely restrains ablation. Nevertheless, the calculations also make clear that glaciers on the plateau are more vulnerable than those of other regions because of summer accumulation. In the monsoon climate, warming would cause not only a decrease in accumulation, but also a drastic increase in ablation in combination with surface-albedo lowering. Therefore, although glaciers on and around the plateau can be sustained by summer accumulation, they are more vulnerable to warming than winter-accumulation-type glaciers.

The partitioning between till deformation and sliding in the fast flow of ice streams with active basal melting is examined assuming no adhesion of the till to the ice base and incompressible viscous fluid behavior for the till. For deforming-till thickness of 10 m or less the predicted contribution to basal motion by sliding is larger than shearing in the till unless there is short-scale roughness with wavelengths less than order 0.1 m on the ice sole. At such short scales strain heating within the till and focused melting on the ice sole would quickly eliminate the roughness. Thus, fast flow over a till bed would be expected to be mostly by sliding over the subglacial till. More realistic continuum behavior of the till including non-linear and compressible deformation strengthens the conclusion. If sliding is not dominant, then there must be adhesion of the till to the ice base, some mechanism that continuously generates short-scale roughness on the ice–till interface, or very weak internal slip boundaries within the till.

A one-dimensional second-order closure model and in situ observations on a melting glacier surface are used to investigate the suitability of bulk and profile methods for determining turbulent fluxes in the presence of the katabatic wind-speed maximum associated with glacier winds. The results show that profile methods severely underestimate turbulent fluxes when a wind-speed maximum is present. The bulk method, on the other hand, only slightly overestimates the turbulent heat flux in the entire region below the wind-speed maximum and is thus much more appropriate for use on sloping glacier surfaces where katabatic winds dominate and wind-speed maxima are just a few meters above the surface.

During summer 1997, a 15 m firn core was recovered from Dasuopu glacier (28°23′ N, 85°44′ E; 7000 m a.s.l.) on the northwest margin of Xixabarngma Feng in the central Himalaya. Oxygen isotope values and concentrations of Ca2+, Mg2+, NH4+, SO42− and NO3− were measured over the 10 years of snow accumulation captured in the firn core. The seasonal variations of δ18O values and major-ion concentrations in the Dasuopu core indicate that summer monsoon and dust signals are clearly recorded in Dasuopu glacier. Annual variations in δ18O values are controlled by the amount effect, with more negative (i.e. lighter) δ18O values representing summer monsoon precipitation characteristic of tropical regions. Higher concentrations of Ca2+, Mg2+ and SO42− reflect the influx of mineral aerosols from the vast arid and semi-arid desert regions to the north and west during the spring dust-storm period. High spring concentrations of NH4+ and NO3− appear to reflect changes in regional biogenic-source strength.

An interferogram reflecting the motion of Black Rapids Glacier, Alaska, U.S.A., was obtained from two European Remote-sensing Satellite (ERS-1) syntheticaperture radar (SAR) images, acquired on 22 and 25 January 1992. We investigate whether the interferometric data are quantitatively consistent with terrestrial velocity measurements along three transverse profiles. These terrestrial data are from different years (1987, 1990, 1996) and cover different periods (6–28 April, 23 May–7 July and a whole year) than the SAR interferogram. Terrestrial ice velocity at the date of the SAR imagery is obtained via seasonal and annual corrections that are calculated from other terrestrial velocity measurements available at higher time resolution for selected sites on the glacier. Interferometric and terrestrial velocity are in excellent agreement if a (terrestrially measured) surface-normal velocity component (v⊥) is properly accounted for. This suggests that both the interferometric velocities and the conversions of terrestrial data to the winter period are reliable. The terrestrial velocity measurements show that ice flow in the upper ablation area (14–16 km sites) changes from longitudinal compression in mid-winter (v⊥ = + 0.82 cm d−1) to moderate longitudinal extension during summer (v⊥ ≈ –0.25 cm d−1). In the lower ablation area, the seasonal variations of the longitudinal strain rate are much smaller: +1.0 and about +0.85 cm d−1 for the respective mid-winter and summer values of v⊥ at the 20 km site.

Shearing of subglacial till has been invoked widely as a mechanism of glacier motion and sediment transport, but standard indicators for determining shear strain from the geologic record are not adequate for estimating the very high strains required of the bed-deformation model. Here we describe a laboratory study of mixing between shearing granular layers that allows an upper limit to be placed on bed shear strain in the vicinity of till contacts. Owing to random vertical motions of particles induced by shearing, mixing can be modeled as a linearly diffusive process, and so can be characterized with a single mixing coefficient, D. Ring-shear experiments with equigranular beads and lithologically distinct tills provide the value of D, although in experiments with till D decreases systematically with strain to a minimum value of 0.0045 mm2. Kinetic gas theory provides an estimate of the dimensionless mixing coefficient which is within an order of magnitude of laboratory values. Knowing the minimum value of D, the distribution of index lithologies measured across till contacts in the geologic record can be used to estimate the maximum shear strain that has occurred across till contacts. Application of this technique to the contact between the Des Moines and Superior Lobe tills in east-central Minnesota, U.S.A., indicates that shear strain did not exceed 15 000 at the depth of the contact.

General equations of ice dynamics are re-examined, using scale analysis, in order to derive a simplified thermomechanically coupled model for ice flow and heat transfer in ice caps filling volcanic craters. Relatively large aspect ratios between crater depths and diameters, low surface temperatures and intense volcanic heating are the principal characteristics of such craters. The conventional boundary-layer (shallow-ice) approximation is revised to account for these conditions and, in addition, the variable density of the snow, firn and bubbly ice. Large crater depths and intense bottom melting result in low longitudinal balance velocities, controlled by both shear and longitudinal stresses, and hence small surface slopes. In such situations ice can be assumed to be linearly viscous. A flowline model of the glacier dynamics is developed using this assumption. Explicit predictive formulas for ice-particle trajectories and age–depth relations, thus obtained, suggest that the age of ice at the bottom of glaciers in volcanic craters on Kamchatka Peninsula, Russia, may reach hundreds or thousands of years. Ice cores from these glaciers represent unique climatic and volcanic archives.

A new method for ice-fabric analysis which is easily automated is described. The technique relies on a set of digital images of an ice thin section viewed through crossed polarizers and rotated to various angles. From the acquired set of images, the c-axis orientation of all the grains in the image can be determined. The theoretical basis for the technique is described in detail, including corrections for refraction, as well as other corrections for off-axis grains and polarization/reflection effects. An experiment was performed comparing results of the new technique with those of the standard technique. Excellent agreement was obtained. It is expected that an automated system using this technique (currently almost complete) will have a large impact on the amount of information and physical knowledge which can be extracted from ice fabrics in the future.

Mass-balance and dynamic measurements carried out on glacier de Saint Sorlin since 1957 provide a good opportunity to study the dynamics of this glacier. Ice-flow analysis shows that dynamic changes have been important over the last 40 years and that these changes are not consistent with the concepts usually used in glacier modelling. Present velocities are larger than the 1960 velocities, although the thickness decreased everywhere (10–30 m in the ablation zone). A simple numerical ice-flow model which does not include longitudinal stress gradients has been used to investigate these phenomena. This model allows us to infer the sliding velocity from observed surface and calculated deformation velocities. We conclude that: (1) the sliding velocity cannot be described by Weertman analysis or empirical relations which link the sliding to the thickness and surface slope; (2) the inferred sliding velocity is uniform over at least half of the glacier; and (3) there is no clear link between the sliding process and the quantity of water coming from surface ablation. Furthermore, it may not be reasonable to calibrate model flow parameters from geometry changes because the surface geometry is relatively insensitive to velocity changes over some decades.

Analysis of the cross-flow transmission of force from the central parts of a well-lubricated ice stream to its margins shows that there is a corresponding shift in the lateral location of motion-induced heat generation. The rate of basal heat generation in the center can be substantially smaller than the local rate of potential energy loss given by driving stress times the speed of downslope motion. The basal heating is a maximum for an intermediate level of lubrication for which speed is about 40% of the speed over a friction-less bed and base stress is about 25% of the driving stress. Stable and unstable balances between meltwater production and drainage on the bed are identified. A stable steady state with a speed less (more) than that giving maximum heat generation is termed drainage-(production-) limited, since an increase in speed would lead to increased (decreased) basal melting and must (need not) be balanced by increased drainage. It is shown that gradual evolution of the basal water drainage system and the factors affecting basal melting can cause discontinuous jumps between fast- and slow-moving states. A simplified analysis applied to six cross-sections of West Antarctic Ice Streams B, D, E and Rutford Ice Stream shows them to be diverse in the level of support from the sides and corresponding shift of mechanical heating sideward from their central parts. The cross-sections of Ice Stream B near “Upstream B” may be production-limited, because of especially high lubrication and related support from the sides. Cross-sections in the upper part of Ice Stream D, Ice Stream E and Rutford Ice Stream are in a drainage-limited condition. Substantial reduction of basal heat generation by side drag (in most cases) and expected high heat flow into the basal ice associated with low thickness (in some cases) tends to favor basal freezing. Nevertheless, all of the examined cross-sections except one are expected to experience basal melting with a modest geothermal heat-flux density of 60 m W m−1 or less in some cases. The lower part of Ice Stream B is an exception, where the analysis indicates that geothermal flux density must exceed 80–100 m W−1 m to maintain melting. If this high geothermal flux is not present, then the base of the lower part of Ice Stream B may be freezing, which would suggest continued deceleration of this part of Ice Stream B.

A degree-day model is used to assess the sensitivity of the mass balance of five Swiss glaciers to temperature changes. The model uses temperature data extrapolated from nearby climate stations, and is tuned by varying precipitation to make the model fit the observed distribution of mass balance with altitude. Once the model is tuned, the effect of temperature change is simulated by recalculating the mass balance with the same parameters as before, but with a temperature increase of 1°C throughout the year. The largest mass-balance changes, involving increased ablation of > 1 m w.e. a−1 °C−1, occur at the snout, with a progressively smaller increase with altitude. The area-averaged sensitivities for the five glaciers are −0.7 to −0.9 m w.e. a−1 °C−1. If annual precipitation also increased by 20% it would partly offset the effect of the 1°C higher temperatures but could not compensate for it.

A new method of ice-core drilling uses an annulus of hot-water jets to melt out a cylindrical ice core. This lightweight device used in combination with a fast hot-water drill can quickly obtain ice cores from any depth.

The Lambert Glacier–Amery Ice Shelf system is a major component of the East Antarctic ice sheet. This paper presents two digital elevation models (DEMs) that have been generated for the Lambert–Amery system from validated European Remote-sensing Satellite (ERS-1) radar altimeter waveform data. The first DEM covers the Amery Ice Shelf only, and was produced using kriging on a 1 km grid. The second is a coarser (5 km) DEM of the entire Lambert–Amery system, generated via simple averaging procedures. The DEMs provide unprecedented surface elevation information for the Lambert–Amery system and allow new insight into the glaciology of the region.

Single crystals with a wide variety of orientations were cut from large pucks of laboratory-grown ice. Constant-strain-rate compression tests were performed on the crystals either at an axial strain rate of 1 × 10−5 s−1at –20°C or at axial strain rates from 1 × 10−6 s−1 to 1 × 10−4 s−1 at –10°C. In agreement with previous studies of ice flow, the compression tests showed a linearly rising stress with increasing strain, followed by a sharply declining stress after reaching a peak. With further strain, the sharp decline in stress slowed and the flow stress approached a plateau that was only weakly dependent on strain. For all crystallographic orientations, it was found that Schmid’s (critical resolved shear stress) law was obeyed by the peak stress. Slip lines clearly showed that basal slip was the deformation mode.

We have tested the ability of a 1.12–1.76 GHz bandwidth airborne Frequency Modulation–Continuous Wave (FM-CW) radar with an effective pulse duration of 3 ns to penetrate temperate ice of the ablation zone of Black Rapids Glacier, central Alaska, U.S.A. We used high-gain horn antennas to suppress clutter, and tested over cold and nearly ideal surface conditions. Englacial horizons dipping to at least 60 m depth were found along three sections of one axial profile. More narrow-band (1.21–1.29 GHz), low-resolution (24 ns pulse duration) profiles from a fourth section detected events at about 100–150 m depth. Comparative profiles recorded with a 100 MHz short-pulse-type radar reproduce the horizons of two of the sections, and verify the penetration in all cases. All horizons are composed of diffractions. We interpret voids from the phase of the 100 MHz diffractions within one of the horizons. The diffraction nature of the horizons, the void interpretation and the proximity to a nearby and up-glacier pothole field lead us to conclude that the horizons within two of the sections are meandering drainage channels. A more complex, branching structure with near-surface horizons profiled within the third section much farther down-glacier may also be a complex drainage system fed by near-surface melting. The FM-CW signal-to-clutter-noise ratios of some of the targets predict that they could be detected at 200 m depth in the 1–2 GHz range. Significant performance improvements at maximum vertical resolution could be achieved with higher-gain antennas.