Search

The thermal regimes of alpine streams remain understudied and have important implications for cold-water fish habitat which is expected to decline due to climatic warming. Previous research has focused on the effects of distributed energy fluxes and meltwater from snowpacks and glaciers on the temperature of mountain streams. This study presents the effects of the groundwater spring discharge from an inactive rock glacier containing little ground ice on the temperature of an alpine stream. Rock glaciers are coarse blocky landforms that are ubiquitous in alpine environments and typically exhibit low groundwater discharge temperatures and resilience to climatic warming. Water temperature data indicate that the rock glacier spring cools the stream by an average of 3°C during July and August and reduces maximum daily temperatures by an average of 5°C during the peak temperature period of the first two weeks in August, producing a cold-water refuge downstream of the spring. The distributed stream surface and streambed energy fluxes are calculated for the reach along the toe of the rock glacier, and solar radiation dominates the distributed stream energy budget. The lateral advective heat flux generated by the rock glacier spring is compared to the distributed energy fluxes over the study reach, and the spring advective heat flux is the dominant control on stream temperature at the reach scale. This study highlights the potential for coarse blocky landforms to generate climatically-resilient cold-water refuges in alpine streams.

Ice streams are integral components of an ice sheet's mass balance and directly impact on sea level. Their flow is governed by processes at the ice-bed interface which create landforms that, in turn, modulate ice stream dynamics through their influence on bed topography and basal shear stresses. Thus, ice stream geomorphology is critical to understanding and modelling ice streams and ice sheet dynamics. This paper reviews developments in our understanding of ice stream geomorphology from an historical perspective, with a focus on the extent to which studies of modern and palaeo-ice streams have converged to take us from a position of near-complete ignorance to a detailed understanding of their bed morphology. During the 1970s and 1980s, our knowledge was limited and largely gleaned from geophysical investigations of modern ice stream beds in Antarctica. Very few palaeo-ice streams had been identified with any confidence. During the 1990s, however, glacial geomorphologists began to recognise their distinctive geomorphology, which included distinct patterns of highly elongated mega-scale glacial lineations, ice stream shear margin moraines, and major sedimentary depocentres. However, studying relict features could say little about the time-scales over which this geomorphology evolved and under what glaciological conditions. This began to be addressed in the early 2000s, through continued efforts to scrutinise modern ice stream beds at higher resolution, but our current understanding of how landforms relate to processes remains subject to large uncertainties, particularly in relation to the mechanisms and time-scales of sediment erosion, transport and deposition, and how these lead to the growth and decay of subglacial bedforms. This represents the next key challenge and will require even closer cooperation between glaciology, glacial geomorphology, sedimentology, and numerical modelling, together with more sophisticated methods to quantify and analyse the anticipated growth of geomorphological data from beneath active ice streams.

The classic model of fjord renewal is complicated by tidewater glacier fjords, where submarine melt and subglacial discharge provide substantial buoyancy forcing at depth. Here we use a suite of idealized, high-resolution numerical ocean simulations to investigate how fjord circulation driven by subglacial plumes, tides, and wind stress depends on fjord width, grounding line depth, and sill height. We find that the depth of the grounding line compared to the sill is a primary control on plume-driven renewal of basin waters. In wide fjords the plume exhibits strong lateral recirculation, increasing the dilution and residence time of glacially-modified waters. Rapid drawdown of basin waters by the subglacial plume in narrow fjords allows for shelf waters to cascade deep into the basin; wide fjords result in a thin, boundary current of shelf waters that flow toward the terminus slightly below sill depth. Wind forcing amplifies the plume-driven exchange flow; however, wind-induced vertical mixing is limited to near-surface waters. Tidal mixing over the sill increases in-fjord transport of deep shelf waters and erodes basin stratification above the sill depth. These results underscore the first-order importances of fjord-glacier geometry in controlling circulation in tidewater glacier fjords and, thus, ocean heat transport to the ice.

Ice-stream dynamics are strongly controlled by processes taking place at the ice/bed interface where subglacial water both lubricates the base and saturates any existing, underlying sediment. Large parts of the former Eurasian ice sheet were underlain by thick sequences of soft, marine sediments and many areas are imprinted with geomorphological features indicative of fast flow and wet basal conditions. Here, we study the effect of subglacial water on past Eurasian ice-sheet dynamics by incorporating a thin-film model of basal water flow into the ice-sheet model SICOPOLIS and use it to better represent flow in temperate areas. The adjunction of subglacial hydrology results in a smaller ice-sheet building up over time and generally faster ice velocities, which consequently reduces the total area fraction of temperate basal ice and ice streaming areas. Minima in the hydraulic pressure potential, governing water flow, are used as indicators for potential locations of past subglacial lakes and a probability distribution of lake existence is presented based on estimated lake depth and longevity.

We identify two previously unknown subglacial lakes beneath the stagnated trunk of the Kamb Ice Stream (KIS). Rapid fill-drain hydrologic events over several months are inferred from surface height changes measured by CryoSat-2 altimetry and indicate that the lakes are probably connected by a subglacial drainage network, whose structure is inferred from the regional hydraulic potential and probably links the lakes. The sequential fill-drain behavior of the subglacial lakes and concurrent rapid thinning in a channel-like topographic feature near the grounding line implies that the subglacial water repeatedly flows from the region above the trunk to the KIS grounding line and out beneath the Ross Ice Shelf. Ice shelf elevation near the hypothesized outlet is observed to decrease slowly during the study period. Our finding supports a previously published conceptual model of the KIS shutdown stemming from a transition from distributed flow to well-drained channelized flow of subglacial water. However, a water-piracy hypothesis in which the KIS subglacial water system is being starved by drainage in adjacent ice streams is also supported by the fact that the degree of KIS trunk subglacial lake activity is relatively weaker than those of the upstream lakes.

The hydrologic system beneath the Antarctic Ice Sheet is thought to influence both the dynamics and distribution of fast flowing ice streams, which discharge most of the ice lost by the ice sheet. Despite considerable interest in understanding this subglacial network and its affect on ice flow, in situ observations from the ice sheet bed are exceedingly rare. Here we describe the first sediment cores recovered from an active subglacial lake. The lake, known as Subglacial Lake Whillans, is part of a broader, dynamic hydrologic network beneath the Whillans Ice Stream in West Antarctica. Even though “floods” pass through the lake, the lake floor shows no evidence of erosion or deposition by flowing water. By inference, these floods must have insufficient energy to erode or transport significant volumes of sediment coarser than silt. Consequently, water flow beneath the region is probably incapable of incising continuous channels into the bed and instead follows preexisting subglacial topography and surface slope. Sediment on the lake floor consists of till deposited during intermittent grounding of the ice stream following flood events. The fabrics within the till are weaker than those thought to develop in thick deforming beds suggesting subglacial sediment fluxes across the ice plain are currently low and unlikely to have a large stabilizing effect on the ice stream's grounding zone.

Water that pressurizes the base of glaciers and ice sheets enhances glacier velocities and modulates glacial erosion. Predicting ice flow and erosion therefore requires knowledge of subglacial channel evolution, which remains observationally limited. Here we demonstrate that detailed analysis of seismic ground motion caused by subglacial water flow at Mendenhall Glacier (Alaska) allows for continuous measurement of daily to subseasonal changes in basal water pressure gradient, channel size, and sediment transport. We observe intermittent subglacial water pressure gradient changes during the melt season, at odds with common assumptions of slowly varying, low-pressure channels. These observations indicate that changes in channel size do not keep pace with changes in discharge. This behavior strongly affects glacier dynamics and subglacial channel erosion at Mendenhall Glacier, where episodic periods of high water pressure gradients enhance glacier surface velocity and channel sediment transport by up to 30% and 50%, respectively. We expect the application of this framework to future seismic observations acquired at glaciers worldwide to improve our understanding of subglacial processes.

Localized subglacial zones of high traction help to regulate ice sheet flow. Geophysical data from a palaeo-ice-stream suggest that methane gas accumulation and hydrate formation beneath ice sheets can produce such high-traction sticky spots.

Ice streams on the Ross Sea side of the West Antarctic Ice Sheet are known to experience flow variability on hourly, annual, and multicentury time scales. We report here on observations of flow variability at the decade scale on the Bindschadler and MacAyeal Ice Streams (BIS and MacIS). Our analysis makes use of archived ice velocity data and new mappings from composited Landsat 7 and Landsat 8 imagery that together span the interval from 1985 to 2014. Both ice streams speedup and slowdown in a range of about ±5 m a−2 over our various comparison intervals. The rates of change are variable in both time and space, and there is no evidence of external forcing at work across the two streams. Widespread changes are most likely linked to instability in the subglacial till and/or subglacial water flow. Sticky spots near the confluence of the two ice streams are loci for speed changes. These relatively young and slow-flowing features appear to be forcing shifts in margin position near the outlets of both streams. The margin jumps reduce the effective outlet widths of the streams by 20% and 30% on BIS and MacIS, respectively. Those magnitudes are similar to the outlet narrowing experienced by Kamb Ice Stream prior to its stagnation.

Satellite laser altimetry reveals short timescale changes in Antarctic ice sheet surface elevation that are suggested to be driven by subglacial water transport and storage. Here details of the interaction between the dynamics of ice stream flow, subglacial water system, and bed elevation relief are examined in the context of idealized, heterogeneous bed geometries. Using a two-way coupled model of ice and subglacial water flow, we show that basal topography controls the temporal and spatial variability of the sub–ice stream hydraulic system. The orientation and characteristic dimensions of the topographic undulations determine the morphology (connected subglacial ponds or channel-like subglacial water features) and timescales of the sub–ice stream drainage system. The short-term (several years to decades) variability of the simulated coupled ice stream/subglacial water system suggests that the short-term surface variations detected in remote sensing observations may be indicative of a rapidly evolving subglacial water system. Our simulations also show that interaction between ice flow and the highly dynamic subglacial water system has a strong effect on effective stress in the ice. Large effective stress magnitudes arise over areas where the basal traction is characterized by strong spatial gradients, that is, transitions from high to low basal traction or vise versa. These transitions migrate on multiyear timescales and thus cause large effective stress variability on the same temporal scales.

Idealized laboratory experiments investigate the glacier–ocean boundary dynamics near a vertical glacier in a two-layer stratified fluid. Discharge of meltwater runoff at the base of the glacier (subglacial discharge) enhances submarine melting. In the laboratory, the effect of multiple sources of subglacial discharge is simulated by introducing freshwater at freezing temperature from two point sources at the base of an ice block representing the glacier. The buoyant plumes of cold meltwater and subglacial discharge water entrain warm ambient water, rise vertically, and interact within a layer of depth H2 if the distance between the sources x0 is smaller than H2α/0.35, where α is the entrainment constant. The plume water detaches from the glacier face at the interface between the two layers and/or at the free surface, as confirmed by previous numerical studies and field observations. A plume model is used to explain the observed nonmonotonic dependence of submarine melting on the sources’ separation. The distance between the two sources influences the entrainment of warm water in the plumes and consequently the amount of submarine melting and the final location of the meltwater within the water column. Two interacting plumes located very close together are observed to melt approximately half as much as two independent plumes. The inclusion, or parameterization, of the dynamics regulating multiple plumes’ interaction is therefore necessary for a correct estimate of submarine melting. Hence, the distribution and number of sources of subglacial discharge may play an important role in glacial melt rates and fjord stratification and circulation.

We reconstruct englacial and subglacial drainage at Skalafellsjokull, Iceland, using ground penetrating radar (GPR) common offset surveys, borehole studies and Glacsweb probe data. We find that englacial water is not stored within the glacier (water content ~0-0.3%). Instead, the glacier is mostly impermeable and meltwater is able to pass quickly through the main body of the glacier via crevasses and moulins. Once at the glacier bed, water is stored within a thin (1 m) layer of debris-rich basal ice (2% water content) and the till. The hydraulic potential mapped across the survey area indicates that when water pressures are high (most of the year), water flows parallel to the margin, and emerges 3 km down glacier at an outlet tongue. GPR data indicates that these flow pathways may have formed a series of braided channels. We show that this glacier has a very low water-storage capacity, but an efficient englacial drainage network for transferring water to the glacier bed and, therefore, it has the potential to respond rapidly to changes in melt-water inputs.

Subglacial discharge influences glacier basal motion and erodes and redeposits sediment. At tidewater glacier termini, discharge drives submarine terminus melting, affects fjord circulation, and is a central component of proglacial marine ecosystems. However, our present inability to track subglacial discharge and its variability significantly hinders our understanding of these processes. Here we report observations of hourly to seasonal variations in 1.5-10 Hz seismic tremor that strongly correlate with subglacial discharge but not with basal motion, weather, or discrete icequakes. Our data demonstrate that vigorous discharge occurs from tidewater glaciers during summer, in spite of fast basal motion that could limit the formation of subglacial conduits, and then abates during winter. Furthermore, tremor observations and a melt model demonstrate that drainage efficiency of tidewater glaciers evolves seasonally. Glaciohydraulic tremor provides a means by which to quantify subglacial discharge variations and offers a promising window into otherwise obscured glacierized environments.

Fast-flowing ice streams in West Antarctica are separated from the nearly stagnant ice in the adjacent ridge by zones of highly localized deformation known as shear margins. It is presently uncertain what mechanisms control the location of shear margins and possibly allow them to migrate. In this paper we show how subglacial hydrological processes can select the shear margin location, leading to a smooth transition from a slipping to a locked bed at the base of an ice stream. Our study uses a two-dimensional thermomechanical model in a cross section perpendicular to the direction of flow. We confirm that the intense straining at the shear margins can generate large temperate regions within the deforming ice. Assuming that the melt generated in the temperate ice collects in a drainage channel at the base of the margin, we show that a channel locally decreases the pore pressure in the subglacial till. Therefore, the basal shear strength just outside the channel, assuming a Coulomb-plastic rheology, can be substantially higher than that inferred under the majority of the stream. Results show that the additional basal resistance produced by the channel lowers the stress concentrated on the locked portion of the bed. Matching the model to surface velocity data, we find that shear margins are stable when the slipping-to-locked bed transition occurs less than 500 m away from a channel operating at an effective pressure of 200 kPa and for a hydraulic transmissivity equivalent to a basal water film of order 0.2 mm thickness.

Deciphering the complex interplays between climate, uplift and erosion is not straightforward and estimating present-day erosion rates can provide useful insights. Glaciers are thought to be powerful erosional agents, but most published ‘glacial’ erosion rates combine periglacial, subglacial and proglacial erosion processes. Within a glaciated catchment, sediments found in subglacial streams originate either from glacial erosion of substratum or from the rock walls above the glacier that contribute to the supraglacial load. Terrestrial cosmogenic nuclides (TCN) are produced by interactions between cosmic ray particles and element targets at the surface of the Earth, but their concentration becomes negligible under 15 m of ice. Measuring TCN concentrations in quartz sand sampled in subglacial streams and in supraglacial channels is statistically compliant with stochastic processes (e.g. rockfalls) and may be used to discriminate subglacial and periglacial erosion. Results for two subglacial streams of the Bossons glacier (Mont Blanc massif, France) show that the proportion of sediments originating from glacially eroded bedrock is not constant: it varies from 50% to 90% (n = 6). The difference between the two streams is probably linked to the presence or absence of supraglacial channels and sinkholes, which are common features of alpine glaciers. Therefore, most of the published mean catchment glacial erosion rates should not be directly interpreted as subglacial erosion rates. In the case of catchments with efficient periglacial erosion and particularly rockfalls, the proportion of sediments in the subglacial stream originating from the supraglacial load could be considerable and the subglacial erosion rate overestimated. Here, we estimate warm-based subglacial and periglacial erosion rates to be of the same order of magnitude: 0.39 ± 0.33 and 0.29 ± 0.17 mm a−1, respectively.

In this study, we present the first measurements of iron (Fe) stable isotopic composition (δ 56 Fe) of subglacial streams draining the Greenland Ice Sheet (GIS). We measure the δ 56 Fe values [(δ 56 Fe, ‰ = ( 56 Fe/ 54 Fe) sample /( 56 Fe/ 54 Fe) standard − 1) × 10 3 ] of both dissolved and suspended sediment Fe in subglacial outflows from five distinct land-terminating glaciers. Suspended sediments have δ 56 Fe values that lie within the crustal array (δ 56 Fe ∼ 0‰). In contrast, the δ 56 Fe values of dissolved Fe in subglacial outflows are consistently less than 0‰, reaching a minimum of −2.1‰ in the outflow from the Russell Glacier. The δ 56 Fe values of dissolved Fe vary geographically and on daily time scales. Major element chemistry and mineral saturation state modeling suggest that incongruent silicate weathering and sulfide oxidation are the likely drivers of subglacial stream Fe chemistry, and that the extent of chemical weathering influences the δ 56 Fe of dissolved Fe. The largest difference in δ 56 Fe between dissolved and suspended load is −2.1‰, and occurs in the subglacial system from the Russell glacier (southwest GIS). Major element chemistry indicates this outflow to be the least chemically weathered, while more mature subglacial systems (i.e., that exhibit greater extents of subglacial weathering) have dissolved loads with δ 56 Fe that are indistinguishable from suspended sediments (Δ 56 Fe suspended-dissolved ∼ 0‰). Ultimately, the dissolved Fe generated in some subglacial systems from the GIS is a previously unrecognized source of isotopically light Fe into the hydrosphere. The data illustrate that the dissolved Fe supplied by subglacial weathering can have variable δ 56 Fe values depending on the degree of chemical weathering. Thus, Fe isotopes have potential as a proxy for subglacial chemical weathering intensity or mode. Finally, based on our regional Fe concentration measurements from each glacial outflow, we estimate a flux weighted continental scale dissolved iron export of 2.1 Gg Fe yr −1 to the coastal ocean, which is within the range of previous estimates.

We present conventional and swath altimetry data from CryoSat-2, revealing a system of subglacial lakes that drained between June 2013 and January 2014 under the central part of Thwaites Glacier, West Antarctica (TWG). Much of the drainage happened in less than 6 months, with an apparent connection between three lakes spanning more than 130 km. Hydro-potential analysis of the glacier bed shows a large number of small closed basins that should trap water produced by subglacial melt, although the observed large-scale motion of water suggests that water can sometimes locally move against the apparent potential gradient, at least during lake-drainage events. This shows that there are important limitations in the ability of hydro-potential maps to predict subglacial water flow. An interpretation based on a map of the melt rate suggests that lake drainages of this type should take place every 20–80 years, depending on the connectivity of the water flow at the bed. Although we observed an acceleration in the downstream part of TWG immediately before the start of the lake drainage, there is no clear connection between the drainage and any speed change of the glacier.

Once an ice sheet grows beyond a critical thickness, the basal thermal regime favors melting and development of subglacial water networks. Subglacial water is necessary for bedrock erosion, but the exact mechanisms that lead to preservation of subglacial topography are unclear. Here we resolve the freezing mechanisms that lead to long-term, high-altitude preservation across the Gamburtsev Subglacial Mountains in East Antarctica. Analyses of a comprehensive geophysical data set reveal a large-scale water network along valley floors. The ice sheet often drives subglacial water up steep topography where it freezes along high ridges beneath thinner ice. Statistical tests of hypsometry show the Gamburtsevs resemble younger midlatitude mountains, indicating exceptional preservation. We conclude that the Gamburtsevs have been shielded from erosion since the latest Eocene (∼34 Ma). These freezing mechanisms likely account for the spatial and temporal patterns of erosion and preservation seen in other glaciated mountain ranges.

Simultaneous observations of moulins and boreholes in western Greenland show that water delivery to the base of the ice sheet via moulins affects short-term ice velocity fluctuations, but not late-season ice velocity decelerations. Increased meltwater delivery to the base of the Greenland Ice Sheet will increase the ice sheet velocity, accelerating its inevitable rush towards the ocean and subsequent sea level rise. Or will it? Debate on this topic is at the forefront of cryospheric research, but has been hampered by the lack of simultaneous observations of hydraulic head in moulins, vertical shafts that deliver water to the base of the ice sheet, and boreholes, which monitor basal water pressure. Lauren Andrews and colleagues now provide these observations from a small region in western Greenland and reveal that water delivery by moulins into channelized basal flow is indeed linked to short-term fluctuations in ice velocity. Late season decelerations in ice velocity, however, seem to be controlled by changes in unchannelized flow, rather than any shifts in the moulin system. Seasonal acceleration of the Greenland Ice Sheet is influenced by the dynamic response of the subglacial hydrologic system to variability in meltwater delivery to the bed1,2 via crevasses and moulins (vertical conduits connecting supraglacial water to the bed of the ice sheet). As the melt season progresses, the subglacial hydrologic system drains supraglacial meltwater more efficiently1,2,3,4, decreasing basal water pressure4 and moderating the ice velocity response to surface melting1,2. However, limited direct observations of subglacial water pressure4,5,6,7 mean that the spatiotemporal evolution of the subglacial hydrologic system remains poorly understood. Here we show that ice velocity is well correlated with moulin hydraulic head but is out of phase with that of nearby (0.3–2 kilometres away) boreholes, indicating that moulins connect to an efficient, channelized component of the subglacial hydrologic system, which exerts the primary control on diurnal and multi-day changes in ice velocity. Our simultaneous measurements of moulin and borehole hydraulic head and ice velocity in the Paakitsoq region of western Greenland show that decreasing trends in ice velocity during the latter part of the melt season cannot be explained by changes in the ability of moulin-connected channels to convey supraglacial melt. Instead, these observations suggest that decreasing late-season ice velocity may be caused by changes in connectivity in unchannelized regions of the subglacial hydrologic system. Understanding this spatiotemporal variability in subglacial pressures is increasingly important because melt-season dynamics affect ice velocity beyond the conclusion of the melt season8,9,10.

The geological nature of glacier beds plays a key role in ice sheet dynamics. Whereas little is known about Greenland’s subglacial geology, the presence of basal sediments is a necessary condition for fast-flowing Antarctic ice streams. Such sediments sustain subglacial till layers, which if water saturated and under high pore-water pressure provide little resistance to ice flow. Using receiver function modeling of teleseismic P-waves, we report a thick (at least tens of meters) sediment layer beneath a site in Greenland’s ablation zone, ∼15 km away from the western ice sheet margin. Although we do not discuss the origin or detailed nature of these subglacial sediments, we suggest that they are capable of sustaining a till layer. Due to the prevalence of an inefficient, pressurized subglacial drainage system, this till layer would typically be under high pore pressures and fail at the shear stresses transmitted from the overlaying ice. Ice flow resistance is thus focused on regions where till is consolidated or absent, or where form drag over obstacles takes place. In contrast to Antarctica, Greenland’s surface melt now affects practically all latitudes, and efficient hydraulic connections between ice sheet surface and bed are pervasive throughout the ablation zone. This implies that rapid, melt-induced changes in subglacial water pressure are possible. The time required for basal till strength to react to these changes depends on the till’s properties. We estimate that formation and destruction of flow resistance can occur on time scales of less than a few years. This could lead to changes in ice flow that are currently difficult to predict.

On most glaciers and ice sheet outlets the majority of motion is due to basal slip, a combination of basal sliding and bed deformation. The importance of basal water in controlling sliding is well established, with increased sliding generally related to high basal water pressure, but the details of the interactions between the ice and water systems has not received much study when there is coupling between the systems. Here we use coupled subglacial hydrology and ice dynamics models within the Community Ice Sheet Model to investigate feedbacks between the ice and water systems. The dominant feedback we find is negative: sliding over bedrock bumps opens additional cavity space, which lowers water pressure and, in turn, sliding. We also find two small positive feedbacks: basal melt increases through frictional heat during sliding, which raises water pressure, and strain softening of basal ice during localized speedup causes cavities to close more quickly and maintain higher water pressures. Our coupled modeling demonstrates that a sustained input of surface water to a distributed drainage system can lead to a speedup event that decays even in the absence of channelization, due to increased capacity of the system through opening of cavities, which is enhanced through the sliding-opening feedback. We find that the negative feedback resulting from sliding-opening is robust across a wide range of parameter values. However, our modeling also argues that subglacial channelization is required to terminate speedup events over timescales that are commensurate with observations of late summer slowdown on mountain glaciers.

The links between structural glaciology, glacial debris entrainment and transport have been established in a number of different glacier settings. Here we document the structural evolution of a temperate Alpine valley glacier (Vadrec del Forno, Switzerland) and demonstrate that individual flow units within the glacier have very different structural and debris characteristics. The glacier consists of a broad accumulation area with multiple basins feeding a relatively narrow tongue and is formed from six distinct flow units. Each flow unit has its own characteristic structural assemblage. Flow units that narrow rapidly down-glacier are dominated by primary stratification that has evolved into longitudinal foliation. In contrast, wider flow units preferentially develop an axial planar foliation. Glacier structure plays a limited role in the entrainment of debris, which is more strongly influenced by ice-marginal rockfall and avalanche inputs onto the glacier surface. However, once entrained, glacier structure controls the reorientation and redistribution of debris within the ice mass. By taking a whole-glacier approach to describing glacier structure and debris transport, we conclude that individual flow units are unique with regard to structure and debris transfer. Copyright © 2013 John Wiley & Sons, Ltd.

[1] We show that subglacial freshwater discharge is the principal process driving high rates of submarine melting at tidewater glaciers. This buoyant discharge draws in warm seawater, entraining it in a turbulent upwelling flow along the submarine face that melts glacier ice. To capture the effects of subglacial discharge on submarine melting, we conducted 4 days of hydrographic transects during late summer 2012 at LeConte Glacier, Alaska. A major rainstorm allowed us to document the influence of large changes in subglacial discharge. We found strong submarine melt fluxes that increased from 9.1 ± 1.0 to 16.8 ± 1.3 m d−1 (ice face equivalent frontal ablation) as a result of the rainstorm. With projected continued global warming and increased glacial runoff, our results highlight the direct impact that increases in subglacial discharge will have on tidewater outlet systems. These effects must be considered when modeling glacier response to future warming and increased runoff.

Interactions between subglacial hydrology and the ocean make the existence of estuaries at the grounding zones of ice sheets likely. Here we present geophysical observations of an estuary at the downstream end of the hydrologic system that links the active subglacial lakes beneath Whillans Ice Stream to the ocean beneath the Ross Ice Shelf, Antarctica. This subglacial estuary consists of a hydropotential low upstream of the grounding zone, which is linked to the ocean by a hydropotential trough and a large subglacial channel. This subglacial channel, which is imaged using active source seismic methods, has an apparent width of 1 km and a maximum depth of 7 m. The hydropotential trough continues upstream of the grounding zone and results from an along-flow depression in surface elevations. Pressure differences along the trough axis are within a range that can be overcome by tidally induced processes, making the interaction of subglacial and ocean water likely.

[1] Subglacial water in continental Antarctica forms by melting of basal ice due to geothermal or frictional heating. Subglacial networks transport the water from melting areas and can facilitate sliding by the ice sheet over its bed. Subglacial water flow is driven mainly by gradients in overburden pressure and bed elevation. We identify small (median 850 m) water bodies within the Gamburtsev Subglacial Mountains in East Antarctica organized into long (20–103 km) coherent drainage networks using a dense (5 km) grid of airborne radar data. The individual water bodies are smaller on average than the water bodies contained in existing inventories of Antarctic subglacial water and most are smaller than the mean ice thickness of 2.5 km, reflecting a focusing of basal water by rugged topography. The water system in the Gamburtsev Subglacial Mountains reoccupies a system of alpine overdeepenings created by valley glaciers in the early growth phase of the East Antarctic Ice Sheet. The networks follow valley floors either uphill or downhill depending on the gradient of the ice sheet surface. In cases where the networks follow valley floors uphill they terminate in or near plumes of freeze-on ice, indicating source to sink transport within the basal hydrologic system. Because the ice surface determines drainage direction within the bed-constrained network, the system is bed-routed but surface-directed. Along-flow variability in the structure of the freeze-on plumes suggests variability in the networks on long (10s of ka) timescales, possibly indicating changes in the basal thermal state.

We use a physical model to investigate how changes in subglacial hydrology affect ice motion of Antarctic ice streams. Ice streams are modelled using silicone polymer placed over a thin water layer to mimic ice flow dominated by basal sliding. The model ice-stream force balance is calculated and compared directly to the observed force balance of Whillans Ice Stream (WIS). Dynamic similarity between the model and WIS is achieved when their force balances are equivalent. The WIS force balance has evolved over time owing to increased basal resistance. We test two hypotheses: (1) the subglacial water distribution influences the ice-flow speed and thus the force balance; (2) shear margins are locations where transitions in water layer thickness occur. We find that the velocity and force balance are sensitive to pulsed water discharge events and changes in lubrication that result in sticky spots, and that model shear margins tend to overlie water lubrication boundaries. We conclude that local changes in basal lubrication near margins (possibly as a result of the presence of sticky spots or subglacial lakes) influence the stability of ice-stream margin position and may be responsible for large and rapid shifts in margin location.

Basal sliding is a main control on glacier flow primarily driven by water pressure at the glacier base. The ongoing increase in surface melting of the Greenland Ice Sheet warrants an examination of its impact on basal water pressure and in turn on basal sliding. Here we examine the case of Russell Glacier, in West Greenland, where an extensive set of observations has been collected. These observations suggest that the recent increase in melt has had an equivocal impact on the annual velocity, with stable flow on the lower part of the drainage basin but accelerated flow above the Equilibrium Line Altitude (ELA). These distinct behaviors have been attributed to different evolutions of the subglacial draining system during and after the melt season. Here we use a high-resolution subglacial hydrological model forced by reconstructed surface runoff for the period 2008 to 2012 to investigate the cause of these distinct behaviors. We find that the increase in meltwater production at low elevation yields a more efficient drainage system compatible with the observed stagnation of the mean annual flow below the ELA. At higher elevation, the model indicates that the drainage system is mostly inefficient and is therefore strongly sensitive to an increase in meltwater availability, which is consistent with the observed increase in ice velocity.

Bakaninbreen in Svalbard and Trapridge Glacier in Yukon Territory, Canada, are two prominent examples of surging glaciers which are thought to be controlled by their thermal regime. Both glaciers have developed large bulges which have propagated forward as travelling wave fronts, and which are thought to divide relatively stagnant downstream cold-based ice from faster-moving warm-based upstream ice. Additionally, both glaciers are underlain by a wet, metres thick layer of deforming till. We develop a simple model for the cyclic surging behaviour of these glaciers, which interrelates the motion of the ice and till through a description of the subglacial hydrology. We find that oscillations (surges) can occur if the subglacial hydrological transmissivity is sufficiently low and the till layer is sufficiently thin, and we suggest that these oscillations are associated with the development and propagation of a travelling wave front down the glacier. We therefore interpret the travelling wave fronts on both Trapridge Glacier and Bakaninbreen as manifestations of surges. In addition, we find that the violence of the surge in the model is associated with the resistance to ice flow offered by undulations in the bed, and the efficiency with which occasional hydrological events can release water accumulated at the glacier sole.

Over the past decade, satellite observations of ice surface height have revealed that active subglacial lake systems are widespread under the Antarctic ice sheet, including the ice streams. For some of these systems, additional observations of ice stream motion have shown that lake activity can affect ice-stream dynamics. Despite all this new information, we still have insufficient understanding of the lake-drainage process to incorporate it into ice sheet models. Process models for drainage of ice-dammed lakes based on conventional "R-channels" incised into the base of the ice through melting are unable to reproduce the timing and magnitude of drainage from Antarctic subglacial lakes estimated from satellite altimetry given the low hydraulic gradients along which such lakes drain. We have developed an alternative process model, in which channels are mechanically eroded into the underlying deformable subglacial sediment. When applied to the known active lakes of the Whillans/Mercer ice stream system, the model successfully reproduced both the inferred magnitudes and recurrence intervals of lake volume changes, derived from Ice, Cloud and land Elevation Satellite (ICESat) laser altimeter data for the period 2003–2009. Water pressures in our model changed as the flood evolved: during drainage, water pressures initially increased as water flowed out of the lake primarily via a distributed system, then decreased as the channelized system grew, establishing a pressure gradient that drew water away from the distributed system. This evolution of the drainage system can result in the observed internal variability of ice flow over time. If we are correct that active subglacial lakes drain through canals in the sediment, this mechanism also implies that active lakes are typically located in regions underlain by thick subglacial sediment, which may explain why they are not readily observed using radio-echo sounding techniques.

We have identified two new subglacial lakes beneath the stagnated trunk of Kamb Ice Stream (KIS). Rapid fill-drain hydrologic events are inferred from Cryosat-2 altimetry, indicating that the lakes are connected by a drainage network. The orientation of the drainage network is inferred from the regional hydraulic potential, and clearly links the lake areas. The behavior of the subglacial lakes and strong thinning observed at the outlet near the grounding line implies that the subglacial water persistently flows from the region above the trunk to grounding line of KIS. In addition, the ice sheet in the KIS trunk estuary is thinning rapidly, and the thinning is accelerated by the activity of subglacial lake. We suggest that a transition from sheet flow of sub-glacial water to well-drained channelized flow may explain the shutdown of Kamb Ice Stream.

The Greenland Ice Sheet is losing mass at an accelerating rate due to increased surface melt and flow acceleration in outlet glaciers. Quantifying future dynamic contributions to sea level requires accurate portrayal of outlet glaciers in ice sheet simulations, but to date poor knowledge of subglacial topography and limited model resolution have prevented reproduction of complex spatial patterns of outlet flow. Here we combine a high-resolution ice-sheet model coupled to uniformly applied models of subglacial hydrology and basal sliding, and a new subglacial topography data set to simulate the flow of the Greenland Ice Sheet. Flow patterns of many outlet glaciers are well captured, illustrating fundamental commonalities in outlet glacier flow and highlighting the importance of efforts to map subglacial topography. Success in reproducing present day flow patterns shows the potential for prognostic modelling of ice sheets without the need for spatially varying parameters with uncertain time evolution., Quantifying Greenland's future contribution to sea level requires accurate portrayal of its outlet glaciers in ice sheet simulations. Here, the authors show that outlet glacier flow can be captured if ice thickness is well constrained and vertical shearing as well as membrane stresses are included in the model.

Subglacial lakes have long been considered hydraulically isolated water bodies underneath ice sheets. This view changed radically with the advent of repeat-pass satellite altimetry and the discovery of multiple lake discharges and water infill, associated with water transfer over distances of more than 200 km. The presence of subglacial lakes also influences ice dynamics, leading to glacier acceleration. Furthermore, subglacial melting under the Antarctic ice sheet is more widespread than previously thought, and subglacial melt rates may explain the availability for water storage in subglacial lakes and water transport. Modelling of subglacial water discharge in subglacial lakes essentially follows hydraulics of subglacial channels on a hard bed, where ice sheet surface slope is a major control on triggering subglacial lake discharge. Recent evidence also points to the development of channels in deformable sediment in West Antarctica, with significant water exchanges between till and ice. Most active lakes drain over short time scales and respond rapidly to upstream variations. Several Antarctic subglacial lakes exhibit complex interactions with the ice sheet due to water circulation. Subglacial lakes can therefore—from a modelling point of view—be seen as confined small oceans underneath an imbedded ice shelf.

Penetration of surface meltwater to the bed of the Greenland Ice Sheet each summer causes an initial increase in ice speed due to elevated basal water pressure, followed by slowdown in late summer that continues into fall and winter. While this seasonal pattern is commonly explained by an evolution of the subglacial drainage system from an inefficient distributed to efficient channelized configuration, mounting evidence indicates that subglacial channels are unable to explain important aspects of hydrodynamic coupling in late summer and fall. Here we use numerical models of subglacial drainage and ice flow to show that limited, gradual leakage of water and lowering of water pressure in weakly connected regions of the bed can explain the dominant features in late and post melt season ice dynamics. These results suggest that a third weakly connected drainage component should be included in the conceptual model of subglacial hydrology.

The growth and drainage of active subglacial lakes in Antarctica has previously been inferred from analysis of ice surface altimetry data. We use a subglacial hydrology model applied to a synthetic Antarctic ice stream to examine internal controls on the filling and drainage of subglacial lakes. Our model outputs suggest that the highly constricted subglacial environment of our idealized ice stream, combined with relatively high rates of water flow funneled from a large catchment, can combine to create a system exhibiting slow-moving pressure waves. Over a period of years, the accumulation of water in the ice stream onset region results in a buildup of pressure creating temporary channels, which then evacuate the excess water. This increased flux of water beneath the ice stream drives lake growth. As the water body builds up, it steepens the hydraulic gradient out of the overdeepened lake basin and allows greater flux. Eventually this flux is large enough to melt channels that cause the lake to drain. Lake drainage also depends on the internal hydrological development in the wider system and therefore does not directly correspond to a particular water volume or depth. This creates a highly temporally and spatially variable system, which is of interest for assessing the importance of subglacial lakes in ice stream hydrology and dynamics., The Cryosphere, 10 (4), ISSN:1994-0416, ISSN:1994-0424

Closed topographic basins are found beneath contemporary ice masses and within the footprint of former ice masses in all glaciated regions. We present the first integrated review of subglacial basin occurrence and formation and the implications of such basins for glaciological processes and the evolution of landscape. Our purpose is to motivate research in areas where understanding of basin origin and process significance is weak. Basins on the order of 10–102 m deep and 102–103 m long are produced by glacial erosion of subglacial rock and/or sediment and are known as ‘overdeepenings’. Outlet and valley glaciers can ‘overdeepen’ their beds far below sea level or local fluviatile base level. Larger basins, typically in ice sheet contexts, may have a pre-glacial (usually tectonic) origin. Subglacial basins are important glaciologically because they require ice, water and sediment to ascend an adverse subglacial slope in order to exit the glacial system, the efficiency of which is dependent upon the gradient of the adverse slope and that of the ice surface. Basins thus influence subglacial drainage system morphology and transmissivity, the thickness and distribution of basal ice and sediment layers, and the mechanisms and dynamics of ice flow. Adverse gradients that exceed 11 times that of the ice surface may even permit the formation of subglacial lakes. We speculate that, in comparison to ice masses with few or no subglacial basins, those with numerous or very large basins may respond to climatic changes with unexpected vigour. In addition, erosion rates and transport pathways of water and sediment through the glacial system, and the expression of these processes in the sediment and landform record, may be unexpectedly complex. Further, our review shows that, in a warming climate, ice masses resting on adverse slopes will be vulnerable to rapid and potentially catastrophic retreat; new lakes in subglacial basins exposed by mountain glacier retreat will present an increasing hazard; and subglacial lakes may drain catastrophically. On even longer time scales, we speculate that the glacial excavation and post-glacial filling of basins in mountainous regions should contribute importantly to climate-related changes in isostasy and relief. Although the controls on overdeepening and their influence on other glacial and landscape processes remain uncertain, we hypothesise that overdeepened glacial systems reflect an equilibrium ice–bed geometry that maximises the efficiency of ice discharge. Improved understanding of overdeepening processes, especially overdeepened-bed hydrology, is therefore necessary to understand fully the dynamic behaviour of valley and outlet glaciers, and thus the fate of Earth's largest ice masses.

Digital elevation models of the surface and bed of Midtdalsbreen, Norway are used to calculate subglacial hydraulic potential and infer drainage system structure for a series of subglacial water pressure assumptions ranging from atmospheric to ice overburden. A distributed degree-day model is used to calculate the spatial distribution of melt on the glacier surface throughout a typical summer, which is accumulated along the various drainage system structures to calculate water fluxes beneath the glacier and exiting the portals for the different water pressure assumptions. In addition, 78 dye-tracing tests were performed from 33 injection sites and numerous measurements of water discharge were made on the main proglacial streams over several summer melt seasons. Comparison of the calculated drainage system structures and water fluxes with dye tracing results and measured proglacial stream discharges suggests that the temporally and spatially averaged steady-state water pressures beneath the glacier are ~70% of ice overburden. Analysis of the dye return curves, together with the calculated subglacial water fluxes shows that the main drainage network on the eastern half of the glacier consists of a hydraulically efficient system of broad, low channels (average width/height ratio ≈ 75). The smaller drainage network on the west consists of a hydraulically inefficient distributed system, dominated by channels that are exceptionally broad and very low (average width/height ratio ≈ 350). The even smaller central drainage network also consists of a hydraulically inefficient distributed system, dominated by channels that are very broad and exceptionally low (average width/height ratio ≈ 450). The channels beneath the western and central glacier must be so broad and low that they can essentially be thought of as a linked cavity system.

The largest dischargers of ice in Greenland are glaciers that terminate in the ocean and melt in contact with sea water. Studies of ice-sheet/ocean interactions have mostly focused on melting beneath near-horizontal floating ice shelves. For tidewater glaciers, melting instead takes place along the vertical face of the calving front. Here we modify the Massachusetts Institute of Technology general circulation model (MITgcm) to include ice melting from a calving face with the freshwater outflow at the glacier grounding line. We use the model to predict melt rates and their sensitivity to ocean thermal forcing and to subglacial discharge. We find that melt rates increase with approximately the one-third power of the subglacial water flux, and increase linearly with ocean thermal forcing. Our simulations indicate that, consistent with limited field data, melting ceases when subglacial discharge is shut off, and reaches several meters per day when subglacial discharge is high in the summer. These results are a first step toward a more realistic representation of subglacial discharge and of ocean thermal forcing on the subaqueous melting of tidewater glaciers in a numerical ocean model. Our results illustrate that the ice-front melting process is both complex and strongly time-dependent.

Recent research has raised concerns about the potential influence of rapid climate change on the stability of major ice sheets. The behaviour of glaciers is determined largely by the processes and conditions operating at their base. Technological advances have allowed these factors to be examined and their contribution to ice flow constrained. This study investigated the rapid disintegration of an aquatic based Norwegian glacier, through the study of boreholes, video, ground-penetrating radar, differential global positioning system, bathymetry and Glacsweb wireless probes. Briksdalsbreen retreated dramatically between 2000 and 2007, with c . 56 × 10 5 m 3 of ice lost from the glacier tongue, equivalent to a rate of 70 m a −1 . This was due to the combined effect of higher summer temperatures, decreased precipitation (resulting from a negative phase of the North Atlantic Oscillation) and increased fracturing of the glacier tongue. The enlargement of a proglacial lake played a key role in Brikdalsbreen9s rapid retreat, allowing calving events and promoting crevassing and fluctuating water contents at the glacier margin. We suggest that hydro-fracturing was the dominant mechanism responsible for generating more crevasses each year, which facilitated the development of an efficient englacial drainage system. This fed increasing quantities of water to the bed, where it was stored in subglacial cavities and transferred through a distributed (‘slow9) drainage system. However, despite this increase in subglacial water content, ice velocities remained constant during the break-up. Comparisons are made between the processes observed at Briksdalsbreen and those associated with the acceleration and rapid retreat of Greenland9s tidewater glaciers.

Proglacial icings accumulate in front of many High Arctic glaciers during the winter months, as water escapes from englacial or subglacial storage. Such icings have been interpreted as evidence for warm-based subglacial conditions, but several are now known to occur in front of cold-based glaciers. In this study, we investigate the drainage system of Tellbreen, a 3.5 km long glacier in central Spitsbergen, where a large proglacial icing develops each winter, to determine the location and geometry of storage elements. Digital elevation models (DEMs) of the glacier surface and bed were constructed using maps, differential GPS and ground penetrating radar (GPR). Rates of surface lowering indicate that the glacier has a long-term mass balance of −0.6 ± 0.2 m/year. Englacial and subglacial drainage channels were mapped using GPR, showing that Tellbreen has a diverse drainage system that is capable of storing, transporting and releasing water year round. In the upper part of the glacier, drainage is mainly via supraglacial channels. These transition downglacier into shallow englacial "cut and closure" channels, formed by the incision and roof closure of supraglacial channels. Below thin ice near the terminus, these channels reach the bed and contain stored water throughout the winter months. Even though no signs of temperate ice were detected and the bed is below pressure-melting point, Tellbreen has a surface-fed, channelized subglacial drainage system, which allows significant storage and delayed discharge.

AbstarctLocations of subglacial lakes discovered under fast-moving West Antarctic ice streams tend to be associated with topographic features of the subglacial bed or with areas that have strong variations in basal conditions. Inversion of ice-stream surface velocity indicates that basal conditions under ice streams can be highly variable and that there can be widespread regions where basal traction is high. To seek an explanation for why lakes appear to be sited near areas with high basal traction, we use numerical models to simulate ice-stream dynamics, thermodynamics and subglacial water flow. We demonstrate that the ice flow over high basal traction areas produces favourable conditions for the ponding of meltwater. Energy dissipation associated with ice sliding over a region with high basal traction constitutes a water source supplying a lake, and ice-thickness perturbations induced by ice flow over variable traction create local minima in hydraulic potential. Variations in thermodynamic processes caused by such ice flow could be responsible for limiting the horizontal extent of the subglacial lakes.

Tidewater glaciers deposit sediment at their terminus, thereby reducing the relative water depth. Reduced water depth can lead to increased glacier stability through decreased rates of iceberg calving, glacier thinning and submarine melting. Here we investigate sedimentation processes at the termini of Kronebreen and Kongsvegen, Svalbard. We mapped the fjord floor bathymetry in August 2009 and calculate sedimentation rates based on our bathymetry and that from a similar study in 2005. A grounding-line fan is developing near the current position of the subglacial stream. An older, abandoned grounding-line fan that likely formed between ∼1987 and 2001 is degrading near the middle of the ice front. Our findings indicate that sediment gravity flows reduce the height of the sediment mound forming at the glacier terminus. The future impact of glacimarine sedimentation processes on glacier stability will depend on the net balance between the observed gravity flows and sediment deposition.

Subglacial lakes beneath Antarctica’s fast-moving ice streams are known to undergo ∼1 km3 volume changes on annual timescales. Focusing on the MacAyeal Ice Stream (MacIS) lake system, we create a simple model for the response of subglacial water distribution to lake discharge events through assimilation of lake volume changes estimated from Ice, Cloud and land Elevation Satellite (ICESat) laser altimetry. We construct a steady-state water transport model in which known subglacial lakes are treated as either sinks or sources depending on the ICESat-derived filling or draining rates. The modeled volume change rates of five large subglacial lakes in the downstream portion of MacIS are shown to be consistent with observed filling rates if the dynamics of all upstream lakes are considered. However, the variable filling rate of the northernmost lake suggests the presence of an undetected lake of similar size upstream. Overall, we show that, for this fast-flowing ice stream, most subglacial lakes receive >90% of their water from distant distributed sources throughout the catchment, and we confirm that water is transported from regions of net basal melt to regions of net basal freezing. Our study provides a geophysically based means of validating subglacial water models in Antarctica and is a potential way to parameterize subglacial lake discharge events in large-scale ice-sheet models where adequate data are available.

Ice streams influence ice sheet mass balance and stability but key aspects of their behaviour remain poorly understood. This paper reviews and discusses one very important aspect: what controls their location in an ice sheet? Seven potential controls on ice stream location are identified from the literature: topographic focusing, topographic steps, macro-scale bed roughness, calving margins, subglacial geology, geothermal heat flux and subglacial meltwater routing. For each control, the theoretical basis for its link to rapid ice flow is introduced, followed by discussion of the evidence of its influence on the location of both contemporary and palaeo-ice streams. Based on this new synthesis, topographic focusing, subglacial geology, meltwater routing and calving margins appear to be most commonly associated with fast ice flow. It is clear that rather than a single control, however, there exist a number of potential controls of varying influence. We propose a hierarchy of factors, with those occurring at the top of the hierarchy exerting a stronger influence on ice stream location, and where present beneath an ice sheet, are very likely to be associated with fast flow. Those factors occurring lower down the hierarchy are less commonly associated with ice streaming but appear to be influential in the absence of more common controls. In such a hierarchy topographic focusing in the presence of a calving margin is the primary control. In the absence of this, ice streams will preferentially occur in areas with favourable subglacial meltwater routing and subglacial geology. In the absence of these, bed roughness, geothermal heat flux and topographic steps may promote ice streaming. Significantly, the primary controls on a given ice stream location are likely to influence its spatial and temporal dynamics. Ice streams governed by the presence of meltwater routing and/or calving processes might exhibit more variable behaviour because these controls can vary over relatively short time-scales compared to controls that vary over longer time-scales, e.g. geothermal heat flux, subglacial roughness, geology and topography. Identifying the controls on ice stream location is therefore of paramount importance when understanding ice stream longevity and their past and future activity.

A hydrologically coupled flowband model of ‘higher order’ ice dynamics is used to explore perturbations in response to supraglacial water drainage and subglacial flooding. The subglacial drainage system includes interacting ‘fast’ and ‘slow’ drainage elements. The fast drainage system is assumed to be composed of ice-walled conduits and the slow system of a macroporous water sheet. Under high subglacial water pressures, flexure of the overlying ice is modelled using elastic beam theory. A regularized Coulomb friction law describes basal boundary conditions that enable hydrologically driven acceleration. We demonstrate the modelled interactions between hydrology and ice dynamics by means of three observationally inspired examples: (i) simulations of meltwater drainage at an Alpine-type glacier produce seasonal and diurnal variability, and exhibit drainage evolution characteristic of the so-called ‘spring transition’; (ii) horizontal and vertical diurnal accelerations are modelled in response to summer meltwater input at a Greenland-type outlet glacier; and (iii) short-lived perturbations to basal water pressure and ice-flow speed are modelled in response to the prescribed drainage of a supraglacial lake. Our model supports the suggestion that a channelized drainage system can form beneath the margins of the Greenland ice sheet, and may contribute to reducing the dynamic impact of floods derived from supraglacial lakes.

Subglacial conditions of large polar ice sheets remain poorly understood, despite recent advances in satellite observation. Major uncertainties related to basal conditions, such as the temperature field, are due to an insufficient knowledge of geothermal heat flow. Here, a hybrid method is presented that combines numerical modeling of the ice sheet thermodynamics with a priori information using a simple assimilation technique. Additional data are essentially vertical temperature profiles measured in the ice sheet at selected spots, as well as the distribution of subglacial lakes. In this way, geothermal heat-flow datasets are improved to yield calculated temperatures in accord with observations in areas where information is available. Results of the sensitivity experiments show that 55% of the grounded part of the Antarctic ice sheet is at pressure melting point. Calculated basal melt rates are approximately 65 Gt year − 1 , which is 3% of the total surface accumulation. Although these sensitivity experiments exhibit small variations in basal melt rates, the impact on the ice age of basal layers is quite important. In areas that are characterized by relatively low melt rates, this may lead to standard deviations up to 50% (in a model run over a period of 10 6 years). Subglacial water flow is concentrated underneath the large outlet glaciers of Antarctica. The larger subglacial lakes are perched at the head of these subglacial water systems. An exception, however, is the lake system of the Recovery Ice Stream Catchment, that receives a substantial amount of subglacial melt water from the upstream catchment.

The transport of subglacial water beneath the East Antarctic Ice Sheet is an enigmatic and difficult to observe process which may affect the flow of the overlying ice and mixing of the oceans in the sub ice shelf cavities, and ultimately global climate. Periodic outbursts are a critical mechanism in this process. Recent analysis of satellite data has inferred a subglacial hydraulic discharge totaling 2 km 3 traveling some 260 km along the ice-bed interface of the Adventure Subglacial Trench between 1996 and 1998 (Wingham et al., 2006. Rapid discharge connects Antarctic subglacial lakes. Nature 440, 1033–1036). Using radar echo sounding data from the Adventure Subglacial Trench region in conjunction with the previously reported satellite observations, along with some basic modeling, we calculate a mass budget and infer a flow mechanism for the 1996–1998 event. The volume released from the source lake exceeded the volume received by the destination lakes by ~ 1.1 km 3 . This discrepancy indicates that some water must have escaped downstream from the lowest destination lake from 1997 onward. The downstream release of water from the destination lakes continued until at least 2003, several years after the 1998 cessation of surface subsidence at the source lake. By 2003 a total of 1.5 km 3 or nearly 75% of the water released by the source lake had traveled downstream from the destination lakes. The temporal evolution of discharge from the outlet can be simulated with the classic ice-walled semicircular channel model, if and only if the retreat of the source lake shoreline is taken into account. Further downstream, the ice bedrock geometry along the inferred flow path downstream includes many sections where thermal erosion of the overlying ice would not be sustainable. Along these reaches mechanical lifting of the ice roof and/or erosion of a sedimentary substrate by a broad shallow water system would be most effective means of sustaining the discharge. A distributed system is also consistent with the 3-month delay between water release at the source lake and water arrival at the destination lake. Observations of intermittent flat bright bed reflections in radar data acquired along the flow path are consistent with the presence of a broad shallow water system. Ultimately the presence of large subglacial lakes along the flow path of the 1996–1998 Adventure Subglacial Trench flow path delayed the arrival of water to points downstream by approximately 12 months.

A global positioning system and ground penetrating radar surveys is used to produce digital elevation models of the surface and bed of Brewster Glacier. These are used to derive maps of subglacial hydraulic potential and drainage system structure using three different assumptions about the subglacial water pressure (Pw): (i) Pw = ice overburden; (ii) Pw = half ice overburden; (iii) Pw = atmospheric. Additionally, 16 dye-tracing experiments at 12 locations were performed through a summer melt season. Dye return curve shape, together with calculations of transit velocity, dispersivity and storage, are used to infer the likely morphology of the subglacial drainage system. Taken together, the data indicate that the glacier is underlain by a channelised but hydraulically inefficient drainage system in the early summer in which water pressures are close to ice overburden. By mid-summer, water pressures are closer to half-ice overburden and the channelised drainage system is more hydraulically efficient. Surface streams that enter the glacier close to the location of major subglacial drainage pathways are routed quickly to the channels and then to the glacier snout. Streams that enter the glacier further away from the drainage pathways are routed slowly to the channels and then to the snout because they first flow through a distributed drainage system. Copyright © 2008 John Wiley & Sons, Ltd.

We examine patterns of localized surface elevation change in lower Mercer and Whillans Ice Streams, West Antarctica, which we interpret as subglacial water movement through a system of lakes and channels. We detect and measure the lake activity using repeat-track laser altimetry from ICESat and image differencing from MODIS image pairs. A hydrostatic-potential map for the region shows that the lakes are distributed across three distinct hydrologic regimes. Our analysis shows that, within these regimes, some of the subglacial lakes appear to be linked, with drainage events in one reservoir causing filling and follow-on drainage in adjacent lakes. We also observe changes near ice raft ‘a’ in lower Whillans Ice Stream, and interpret them as evidence of subglacial water and other changes at the bed. The study provides quantitative information about the properties of this complex subglacial hydrologic system, and a relatively unstudied component of ice-sheet mass balance: subglacial drainage across the grounding line.

Large ice streams and outlet glaciers drain Greenland and Antarctica. An observed acceleration of ice velocity in one of these outlet glaciers, Byrd Glacier, East Antarctica coincides with a large water discharge from two subglacial lakes, allowing direct attribution of the change in glacier dynamics to the water drainage network beneath the ice.

The extent of basin glacierization has important implications for the hydrograph in part, because snow, firn and ice impart different delays in water transport through the system. Here, the significance of subglacial drainage morphology in modulating the hydrograph is examined with a one-dimensional physically based model. The conceptual model of subglacial drainage comprises both 'fast' and 'slow' elements, respectively associated with summer and winter drainage regimes. The additional possibility of a permeable glacier substrate is taken into account by allowing water transport in a subglacial aquifer. Forced by prescribed rates of melt-water delivery to the glacier bed, the model predicts glacier discharge by drainage system provenance. The effects of (1) 'hard' versus 'soft' glacier beds, (2) subsurface permeability and groundwater flow, and (3) glacier geometry are then investigated. Hydrograph character, in the form of peak timing and amplitude, symmetry with respect to the forcing, and the amplitude of diurnal fluctuations, is affected by the partitioning of water through the various flow elements. Hard beds and impermeable substrates maximize the discharge routed through the fast-drainage system in the simulations, generally resulting in higher seasonal discharge maxima and stronger diurnal variations in discharge. High hydraulic transmissivities, either at the glacier bed or in underlying strata, hinder the development of the fast-drainage system in the simulations, producing hydrographs of lower amplitude. Glacier geometry has a modest effect, with adverse bed slopes, very thick or very thin ice and short glacier lengths favouring prolonged drainage through the slow system. These results suggest that the morphology and evolution of the subglacial drainage system may play a significant role in determining the character of the hydrograph from glacierized basins.