West Antarctic ice streams show pronounced flow variability in their downstream reaches, with changes stranding formerly fast-flowing ice and redirecting discharge. A simple model, in which the temperature gradient in basal ice provides control of fast sliding in the downstream reach, can explain this behavior. Downstream thinning steepens the temperature gradient near the bed, increasing upward heat flow and the tendency toward basal freezing. The basal temperature gradient is steepest and the tendency toward basal freezing the strongest in ice that has experienced the most rapid downstream thinning, that is, the fastest-flowing ice. The most ‘successful’ rapid outflows are regions where basal water from elsewhere is likely to be consumed. Freezing here leads to episodic slow-downs and redirections of flow, the history of which appears in satellite imagery as ice rises, distorted streaklines, and margin jumps created when discharge migrates to areas with more favorable basal conditions. One compelling consequence of this process is that it makes catastrophic collapse less likely; if discharge currents are forced to slow when they become too fast (thin), then there may be an upper bound on the retreat rate and discharge flux of the West Antarctic ice sheet (WAIS) ice-stream system under the present climate.

Debris bands associated with supraglacial moraines and associated basal deposits have been logged and sampled for their ice and debris at three glaciers in northwest Spitsbergen, Svalbard. Physical properties, including sediment concentrations, sediment particle-size distributions, clast macro-fabrics, and oxygen isotope compositions, indicate that all transverse and some longitudinal debris bands originate from the basal zone of these glaciers. Transverse supraglacial bands are composed of extensive stratified-facies basal ice that is enriched in 18O and which contains polymodal debris with spatially consistent clast fabrics. These properties suggest initial formation as basal ice and subsequent elevation into an englacial position by thrusting rather than formation as crevasse fills. The formation of longitudinal debris bands results from laterally compressive folding in response to the convergence of multiple flow units into a narrow glacier tongue. In common with transverse debris bands, longitudinal bands appear to be composed of stratified basal ice. The bands exposed at the surface of austre Brøggerbreen comprise two subfacies, strongly suggesting that the glacier was at least partially warm-based in the past, when the basal ice formed.

To better understand apparent stratigraphic disturbances in ice cores such as Greenland Ice Sheet Project 2 (GISP2), we examine how ice-sheet flow can transform gentle open folds into order-disturbing recumbent folds. The initial disturbances in the stratigraphy have their roots in transient dynamic processes and local rheological inhomogeneities, but the kinematics of even a simple ice-flow model can deform these disturbances enough to alter paleoclimatic interpretation of an ice core. The local vorticity number suggests which structures can be passively overturned, but analyzing the finite strain along particle paths gives a more complete picture, especially when taken relative to a hypothetical core location. Core-relative isochrones, or “pre-cores”, predict which stratigraphic disturbances will appear as obviously overturned layers in a core. The deformation-gradient tensor along particle paths allows us to calculate the rotation of segments of various reference slopes. These calculations suggest that observed 20° dips in the GISP2 core are rotating on a time-scale of a few hundred years and could result from distortions with much smaller slopes produced upstream. The time during which they can be recognized to be overturning is short because the rotation rate is high. Once overturned they are flattened further and may be hard to recognize, especially in the small cross-section of a core.

Understanding the microstructure of ice underpins the interpretation of ice-core measurements and many ice-sheet properties. A detailed study of polar snow and ice using scanning electron microscope (SEM) and X-ray analysis revealed the micro-structural distribution of soluble impurities. Sublimation under vacuum (etching) concentrated impurity from both the bulk and grain boundaries on to the specimen surfaces in detectable quantities. Sublimation in the cold room before examination (pre-etching) collected previously unobservable quantities of impurity at triple junctions. A heterogeneous distribution of impurities was observed. Chloride was frequently found to originate from the lattice, but not usually at triple junctions. Other impurities, particularly sodium chloride, were detected at grain boundaries and bubble surfaces. Sulphate was often found at triple junctions in specimens containing a high bulk concentration of the acid, frequently in conjunction with cations. This suggests the possibility that veins were only filled with significant amounts of impurity when the surrounding grain boundaries were saturated. The model of impurity arrangement inferred from the results reconciles differences between previous SEM studies; additionally it is consistent with and explains recent electrical conduction observations. The disconnected arrangement of impurity-filled grain boundaries and veins limits opportunities for significant post-depositional solute movement.

A level-set method is proposed for modeling the evolution of a glacier surface subject to a prescribed mass balance. This leads to a simple and versatile approach for computing the evolution of glaciers: the description of vertical fronts and overriding phenomena presents no difficulties, topological changes are handled naturally and steady-state solutions can be calculated without integration over time. A numerical algorithm is put forth as a means of solving the proposed model of glacier surface evolution. It is evaluated by comparing different numerical solutions of the model with analytical and published numerical solutions. The level-set method appears to be a reliable approach for dealing with different glaciological problems.

The North Greenland Icecore Project (NorthGRIP) palaeoclimatic information back to about 120 kyr BP. The size distributions of ice crystals in the upper 880 m of the NorthGRIP ice core, which cover a time-span of approximately 5300 years, have been obtained previously. The distributions evolve towards a universal curve, indicating a common underlying physical process in the formation of crystals. We identify this process as an interplay between fragmentation of the crystals and diffusion of their grain boundaries. The process is described by a two-parameter differential equation to which we obtain the exact solution. The solution is in excellent agreement with the measured distributions.

A new snow—soil—vegetation—atmosphere transfer (Snow-SVAT) scheme, which simulates the accumulation and ablation of the snow cover beneath a forest canopy, is presented. The model was formulated by coupling a canopy optical and thermal radiation model to a physically based multi-layer snow model. This canopy radiation model is physically based yet requires few parameters, so can be used when extensive in situ field measurements are not available. Other forest effects such as the reduction of wind speed, interception of snow on the canopy and the deposition of litter were incorporated within this combined model, SNOWCAN, which was tested with data taken as part of the Boreal Ecosystem—Atmosphere Study (BOREAS) international collaborative experiment. Snow depths beneath four different canopy types and at an open site were simulated. Agreement between observed and simulated snow depths was generally good, with correlation coefficients ranging between r2 = 0.94 and r2 = 0.98 for all sites where automatic measurements were available. However, the simulated date of total snowpack ablation generally occurred later than the observed date. A comparison between simulated solar radiation and limited measurements of sub-canopy radiation at one site indicates that the model simulates the sub-canopy downwelling solar radiation early in the season to within measurement uncertainty.

Dewatering structures are a common feature used to identify melt-out till, and the lack of such structures in till could preclude deposition by melt-out. To assess the conditions under which melt-out till can be deposited without forming dewatering structures, I use geotechnical data and a quasi-two-dimensional model of geothermal melt-out. Critical discharge determined from geotechnical data suggests that low-hydraulic-conductivity till can transport up to 1.3 m3 water a–1m–2 without forming dewatering structures, which is two to three orders of magnitude greater than the volume of meltwater produced at the base of glaciers. The model indicates that debris content of the ice and the ability of the till to drain govern effective pressure during melt-out. If the drainage system is poorly developed or the till comes from debris-poor ice, effective pressure is below zero, the condition under which dewatering structures could form. However, till from relatively debris-rich ice (>40% debris) with a well-developed drainage system (channels every 10 m) can dewater without forming dewatering structures. This suggests that the lack of dewatering structures in till does not necessarily imply deposition by lodgement or deforming bed.

The discussion on global change has led to increased interest in glacier mass balance since glaciers can be used as climatic indicators. To meet the need for high-quality mass-balance data requires critical examination of traditional mass-balance methods and their possible errors. One issue regarding mass-balance measurements that has received little attention is internal accumulation. Our study shows that internal accumulation in the firn layer of Storglaciären, Sweden, significantly affects the mass balance of the glacier. This occurs because the winter cold wave penetrates below the previous year’s summer surface and into underlying firn. We estimated internal accumulation from measurements of temperature and water content in firn. The depth of the 0°C isotherm correlated with snow depth and air temperature, so that low snow depth and low air temperature separately cause a deeper 0°C isotherm. We determined irreducible gravimetric water content in firn to 2–3%, which corresponds to an irreducible water saturation of 6–8%. Our value for firn is relatively high compared with that for snow, probably due to trapped water in isolated firn pores. Refreezing of percolating meltwater in spring accounted for ~30% of annual internal accumulation. The remaining 70% was due to re-freezing of retained capillary water in firn pores during winter. Disregarding internal accumulation would lead to underestimation of annual net mass balance by 0.04–0.06 m w.e., corresponding to 3–5% of annual accumulation of the entire glacier in an average year. Hence, internal accumulation potentially becomes a source for systematic error if not accounted in mass-balance measurements.

Interferometric radar images collected by ERS-1, ERS-2 and RADAR- SAT-1 are used to observe the rupture tip of rifts that propagate along Hemmen Ice Rise on the Ronne Ice Shelf, Antarctica. Interferograms generated in 1992 and 1997 allow for the observation of ice deformation accumulated over 9 and 24 days respectively. These interferograms are combined, in order to separate the continuous process of creep deformation from the more cyclic motion caused by variations in ocean tide. An examination of local gradients in creep deformation reveals the pattern of ice deformation around and near the rupture tips and rifts with great precision (up to 10 cm a-1). We compare the observations with a deformation model for ice and obtain the following results: (1) The tidal oscillation of the Ronne Ice Shelf only yields small deformations along the rifts and near the rupture tips. (2) Along the ice front, the rifts and at the rupture tips, vertical bending is observed which is well explained by a model of viscous deformation of ice. Furthermore, the model indicates that the deformation pattern observed at the rupture tips is a sensitive indicator of the propagation state of the rifts (i.e. active vs inactive). (3) The viscous adjustment of ice is the dominant mode of deformation, masking the deformation pattern predicted by linear elastic fracture mechanics (LEFM). (4) Yet, at a spatial scale equivalent to the length of a rift, the propagation rate is well predicted by LEFM.

The floating terminal of Jakobshavn Isbræ, the fastest Greenland ice stream, has disintegrated since 2002, resulting in a doubling of ice velocity and rapidly lowering inland ice elevations. Conditions prior to disintegration were modeled using control theory in a plane-stress solution, and the Missoula model of ice-shelf flow. Both approaches pointed to a mechanism that inhibits ice flow and that is not captured by either approach. Jamming of flow, an inherent property of granular materials passing through a constriction (Jakobshavn Isfjord), is postulated as the mechanism. Rapid disintegration of heavily crevassed floating ice accompanies break-up of the ice jam.

A new model of isotopic diffusion in the upper few meters of firn tracks the isotopic composition of both the ice matrix and the pore-space vapor through time in two dimensions. Stable isotopes in the vapor phase move through the firn by diffusion along concentration gradients and by advection. Wind-driven ventilation carries atmospheric water vapor into the firn, where it mixes with existing pore-space vapor. Unlike previous models, our model allows disequilibrium between pore-space vapor and the surrounding snow grains. We also calculate the isotopic effects of ventilation-driven sublimation and condensation in the firn. Model predictions of isotopic diffusion in firn compare favorably with existing diffusion models. Model results quantify what other investigators have suggested: isotopic change in the upper few meters is more rapid than can be explained by the Whillans and Grootes (1985) model; isotopic equilibration with atmospheric vapor is an important component of post-depositional isotopic change; and ventilation enhances isotopic exchange by creating regions of relatively rapid sublimation and condensation in the firn.

An ice shove along the Alaskan Chukchi Sea coast occurred in June 2001, affecting the communities of Barrow and Wainwright, some 150 km apart. Aerial photography before and after the event allowed measurement of ice displacement vectors near Barrow where up to 395 m of ice motion was accommodated almost entirely in discrete ridges up to 5 m high. The forces required to build these ridges are estimated at 35−62 kN m-1, and driving forces of the whole event are investigated. Most ice deformation at or near the beach coincided with local onshore winds, but the event was preceded by the compaction of pack ice in the central Chukchi Sea and the closure of the coastal flaw lead, driven by the larger-scale wind field acting over several days beforehand. Whether this acted to impart pack-ice stress to the coast or simply to create a critical fetch of consolidated ice is uncertain. The near-melting near-isothermal state of the ice may have been a complicit factor and affected the behavior of the land-fast ice. Coastal morphology and bathymetry affected the location of deformation. This study highlights the range of scales at which processes act and culminate to have implications for Arctic communities.

A photogrammetric survey in 1998 formed the basis for compiling a map of the Tuyuksu glacier region in the northern Tien Shan, Central Asia. Comparison with a map from 1958 enables calculation of the change in glacier volume over 40 years. The results are compared with direct annual measurements carried out continuously since 1956. Central Tuyuksu glacier shows a mass balance of –12.6 mw.e. for the geodetic method and –16.8 m for glaciological measurements. In view of the high accuracy of the maps, the discrepancy can be explained mainly by deficiencies in the glaciological measurements. Application of the precipitation–runoff model HBV-ETH represents the hydrological method of mass-balance determination and delivers specific mass-balance losses almost twice those indicated by the geodetic observations. This discrepancy is suspected to be caused by erroneous runoff measurements. The volume balance derived by the geodetic method is used to calibrate both the glaciological and the hydrological method.

An essential problem in snow science is to predict the changing form of ice grains within a snow layer. Present theories are based on the idea that form changes are driven by mass diffusion induced by temperature gradients within the snow cover. This leads to the well-established theory of isothermal- and temperature-gradient metamorphism. Although diffusion theory treats mass transfer, it does not treat the influence of this mass transfer on the form — the curvature radius of the grains and bonds — directly. Empirical relations, based on observations, are additionally required to predict flat or rounded surfaces. In the following, we postulate that metamorphism, the change of ice surface curvature and size, is a process of thermodynamic optimization in which entropy production is minimized. That is, there exists an optimal surface curvature of the ice grains for a given thermodynamic state at which entropy production is stationary. This state is defined by differences in ice and air temperature and vapor pressure across the interfacial boundary layer. The optimal form corresponds to the state of least wasted work, the state of minimum entropy production. We show that temperature gradients produce a thermal non-equilibrium between the ice and air such that, depending on the temperature, flat surfaces are required to mimimize entropy production. When the temperatures of the ice and air are equal, larger curvature radii are found at low temperatures than at high temperatures. Thus, what is known as isothermal metamorphism corresponds to minimum entropy production at equilibrium temperatures, and so-called temperature-gradient metamorphism corresponds to minimum entropy production at none-quilibrium temperatures. The theory is in good agreement with general observations of crystal form development in dry seasonal alpine snow.

Cryoconite holes are water-filled holes in the surface of a glacier caused by enhanced ice melt around trapped sediment. Measurements on the ablation zones of four glaciers in Taylor Valley, Antarctica, show that cryoconite holes cover about 4–6% of the ice surface. They typically vary in diameter from 5 to 145 cm, with depths ranging from 4 to 56 cm. In some cases, huge holes form with 5 m depths and 30 m diameters. Unlike cryoconite holes elsewhere, these have ice lids up to 36 cm thick and melt from within each spring. About one-half of the holes are connected to the near-surface hydrologic system and the remainder are isolated. The duration of isolation, estimated from the chloride accumulation in hole waters, commonly shows ages of several years, with one hole of 10 years. The cryoconite holes in the McMurdo Dry Valleys create a near-surface hydrologic system tens of cm below the ice surface. The glacier surface itself is generally frozen and dry. Comparison of water levels between holes a few meters apart shows independent cycles of water storage and release. Most likely, local freeze–thaw effects control water passage and therefore temporary storage. Rough calculations indicate that the holes generate at least 13% of the observed runoff on the one glacier measured. This hydrologic system represents the transition between a melting ice cover with supraglacial streams and one entirely frozen and absent of water.

The two-dimensional surface velocity of Thwaites Glacier, West Antarctica, was mapped with 23 ascending- and 22 descending-orbit European Remote-sensing Satellite synthetic aperture radar (ERS SAR) interferograms (time range 1995–2000). The velocity map covers 175 500 km2 from the Amundsen Coast to the southern turning point of the satellite orbit and comprises >80% of the Thwaites catchment. Relative velocity errors are <10% except for rare regions (about 5% of the total area) of unfavorable look geometry. Six individual tributaries were identified; their center-line velocities increase from 0 at the catchment boundary to ~0.3 km a–1 when they join the main glacier trunk. On the main trunk, velocity increases to ~1.8 km a–1 at the grounding line and 3.6 km a–1 on the floating tongue. As at neighboring Pine Island Glacier, no strong longitudinal velocity gradients are found except near the grounding line. Within expected error bounds, the flow pattern appears temporally stationary, i.e. flowlines agree with the delineation of flow suggested by the pattern of velocity magnitude. A potential temporal shift of tributary boundaries must consequently be <4.4 m a–1.

As part of a project to investigate the flow of ice at low effective stress, two independent strain-gauge systems were used to measure vertical strain rate as a function of depth and time at Siple Dome, Antarctica. The measurements were made from January 1998 until January 2002 at the ice divide and a site 7km to the northeast on the flank. The strain-rate profiles place constraints on the rheology of ice at low stress, show the expected differences between divide and flank flow (with some structure due to firn compaction and probably ice stratigraphy), and suggest that the flow of the ice sheet has not changed much in the last 8.6 kyr. The strain rates show an unexpected time dependence on scales ranging from several months to hours, including discrete summer events at the divide. Time dependence in strain rate, water pressure, seismicity, velocity and possibly basal motion has been seen previously on the Siple Coast ice streams, but it is especially surprising on Siple Dome because the bed is cold.

Short-term variations in horizontal and vertical surface motion were studied with high temporal resolution during the ablation season in Lauteraargletscher, Bernese Alps, Switzerland. Horizontal surface flow speed oscillated diurnally, showing a correlation with the water level in a borehole. Flow speed increased as a function of the water level, with an asymptote at the ice overburden level. This observation implied that the flow variations were principally controlled by the local water pressure which enhanced basal motions. Detailed examination of the diurnal variations, however, showed that the speed was larger when the pressure was increasing than when it was decreasing. Greater speed with increasing pressure was interpreted by subglacial water-cavity opening and/or longitudinal stress coupling with the upper reaches of the glacier. Upward surface movements were observed when the glacier flow speed increased. Simultaneous measurement of internal vertical strain in a borehole showed that the uplift had two different sources: vertical straining of ice and volume increase of subglacial water cavities. The vertical surface movement was largely affected by the vertical strain, and the uplift events could not be simply attributed to cavity opening.

We investigate the mechanics of ice-stream shear margins based on the assumption that the underlying bed behaves plastically. Sliding is assumed to occur if a prescribed, locally defined yield stress is attained, while no sliding is assumed possible if basal shear stress is lower than the yield stress. Mathematically, the ice-flow problem takes the form of a contact problem, in which the zones of sliding are part of the solution and cannot be prescribed arbitrarily. Simplistic assumptions about the location of till failure, or about mechanical conditions at the bed, predict stress singularities at the margins which lead to corresponding singularities in the basal melt rate. The ice-flow problem is solved using a complex variable method, and an associated quasi-static thermal problem is also solved using a Green’s function. High stress concentrations, which coincide with high rates of strain heating, are found on the ice-stream side of the margins, where basal melting is also greatest. Our results further indicate that a temperate zone may form over time above the bed in the margins. These findings differ from earlier studies based on different sliding laws, suggesting a high sensitivity of margin behaviour to basal conditions.