C31C-0503
Subglacial drainage and morphological switching
Subglacial drainage plays an important role in controlling dynamics between the overlying ice and underlying bed. Here, we discuss different morphologies of subglacial water flow and how morphological transitions occur. That is, how one form changes into another based on flow conditions. For example, a distributed system with lower water velocity can morphologically switch to a concentrated system with higher water velocity. We consider subglacial sheets switching to a channel. We demonstrate that lateral diffusion of heat is an important and previously neglected term for distributed water sheets and modify existing theory to include protrusions that partially support the mass of the overlying glacier and include both the effects of regelation and creep normal to the bed as closure mechanisms. Furthermore, a stress recursion algorithm is introduced because the local driving stress is not necessarily equal to the total effective pressure driving sheet closure. We construct a local driving stress for each protrusion size. This recursion is based on previous work for glacier sliding and subglacial cavitation. Our principal results are that turbulent diffusion of heat laterally in a subglacial water system can suppress development of channels. Calculations show that our theory produces closure velocities up to several orders of magnitude larger than previously expected. With higher closure velocities, we expect sheets to be stable to greater water depth. These closure velocities can be dominated by creep or regelation depending on overall effective pressure, water depth, and protrusion characteristics. Finally, we relate stability of sheets and channels to water flow beneath large ice sheets.
C31C-0504
Stratigraphy and Geomorphology of Late Pleistocene Moraine at the Mouth of Taylor Valley, Antarctica: Implications for the Melting History of the West Antarctic Ice Sheet During the Last Deglaciation
The nature of the West Antarctic Ice Sheet margin at its maximum late Pleistocene extent in the McMurdo Dry Valleys region is controversial. The prevailing view sees the WAIS margin abutting deep proglacial lakes which implies anomalously warm climate and significant effect on the ice margin. This is problematic because GPR profiles in and around Taylor Valley reveal mainly fluvial, not glacilacustrine deposits. We examined a 5 km-scale sediment ridge that forms the divide at the mouth of Taylor Valley, focusing on a 200 m wide, 15 m deep channel system that meanders across it. We collected more than 15 km of GPR at multiple frequencies and 3 cores that provide 38 m of sediment. The GPR shows several widespread undulating reflectors that likely were caused by passage of a grounded lobe of the WAIS into Taylor Valley in the late Pleistocene. Several channels were eroded into the uppermost buried surface, the deepest of which is 100 m wide and 20 m deep. The channels are filled with stratified fluvial and some lacustrine sediment to an elevation that corresponds to that of the highest deltas found using GPR in the interior Fryxell basin, today at 82 m above sea level. The surface channel system was eroded into these stratified sediments. During its last retreat from Taylor Valley, the WAIS was fronted by a shallow glacial lake implying that melting was limited and retreat was largely terrestrial. When the ice cleared the valley, WAIS meltwater carved the buried channels in the valley-mouth ridge lowering the threshold below its current altitude of 74 m asl. This precludes a deep, valley-wide proglacial lake. Further in deglaciation, WAIS melting changes caused fluctuations in channel and lake-level, filling and eroding sediment from the channel. The evidence forms the basis for a melting history of the WAIS margin during the last deglaciation.
C31C-0505
Effect of Climate Cycling and Meltwater Plumbing on Ice-Sheet Grounding-Line Migration
Ice-shelf response to climate cycling introduces hysteretic behavior in outlet glaciers, with the evolution of meltwater drainage networks enhancing this effect. The 2002 collapse of the Larsen B Ice Shelf (one of the floating extensions formed from glaciers flowing off the Antarctic Peninsula into the Weddell Sea) followed several years of shelf-thinning, accelerated flow, and ultimately the atypical warmth of the 2001-02 summer melt season with mean monthly temperatures rising to 4°C and meltwater production up three-fold to ~40 cm/yr (Shepherd et al., 2003; Rignot et al., 2004; van den Broeke, 2005; Vieli et al., 2006; van Lipzig et al., 2008). Yet, for at least the past half century, ice shelves along the margins of the Greenland Ice Sheet have been exposed to summer mean temperatures that hover in the 3-11°C range and melt rates that can exceed 250 cm/yr (Box et al., 2006; Vinther et al., 2006; Hanna et al., 2008; Joughin et al., 2008). We suggest that this paradox can be explained by the advection of englacial meltwater channels that develop upglacier (Zwally et al., 2002; Das et al., 2008), thereby limiting meltwater ponding within crevasses and subsequent ice-shelf failure that is likely primed by surface and basal melting, years of thinning, as well as shear-margin weakening (Rott et al., 1998; Scambos et al., 2000; MacAyeal et al., 2003; Scambos et al., 2004; Vieli et al., 2007). Ultimately, the rapid loss and delayed establishment of ice-shelf buttressing of glacier flow amidst climate cycling leads to an asymmetric response in outlet glacier dynamics with implications for the future of the Larsen ice shelves as well as the hysteretic behavior of ice sheets.
C31C-0506
Dynamic subglacial lakes and West Antarctic ice streams
In November 2007 we initiated field investigations on Mercer and Whillans ice streams of several subglacial lakes recently discovered by analysis of ICESat laser altimetry data. The primary objective of this project is to constrain the temporal dynamics of these lakes and their impact on ice stream flow variability. During the first of three field seasons, we installed 10 continuously recording GPS stations, including one with near real- time satellite data uplink. The stations are located over two medium-size subglacial lakes ('Lake Mercer' and 'Lake Whillans' of Fricker et al., 2007), two small subglacial lakes ('Lake 7' and 'Lake 14' of Fricker et al., 2007) and in several topographic lows, which may correspond to subglacial drainage conduits. In addition, we collected nearly 100 km of ice-penetrating radar data. The latter suggest that lakes Mercer and Whillans still contain no less than 8-9 meters of water, even after recent drainage events detected by ICESat. Active seismic surveys are planned in 2008-09 to establish the depth of these subglacial lake basins. Initial GPS results from Lake Mercer show that the lake surface subsided through the beginning of 2008 when it started to lift up at a rate of ~1cm day. This is consistent with the interpretation from the ICESat repeat-track measurements that the lake activity switched from draining to filling between November 2007 and March 2008, and constrains the timing of the switch much more accurately than is possible from the 4-6 month temporal sampling of ICESat. Improved modeling of hydropotentials indicates that the two major subglacial lakes instrumented by us (Lake Mercer and Lake Whillans) are located in separate subglacial watersheds. This may help explain the observed differences in their temporal behavior.
C31C-0507
Radar echo sounding data driving a model for subglacial water transport from the Dome C Lakes to the onset(s) of fast glacier flow.
he distribution and evolution of subglacial water from Concordia Ridge to Adventure Trench and points beyond? The production of subglacial near the ice divides and the transport of that water from the interior to the onset of fast glacier flow is a critical yet poorly understood component of ice sheet dynamics and the response of the ice sheets to climate change. Using a strain inversion coupled with a temperature model we estimate a water budget for the upstream-most portions of the basins feeding Concordia Lake, Vincennes Lake and ultimately the Adventure Subglacial Trench water system. Using the water budget, the ice surface topography and ice bedrock topography we then test the viability of differing flow systems, and estimate residence time of water in the various subglacial lakes of the region. The strain inversion and temperature model indicates a basin wide average melt rate of 0.84 – 1.20 mm / yr for the Vincennes and Concordia Basins respectively, a figure in accord with other published order of-magnitude-estimates. Although most of the ice bed interface in the region is at or near the pressure melting point of water, the distribution of meltwater production is heterogeneous with melt rates over 2 mm/ yr in the central Vincennes Basin and shoreline of Lake Concordia. Indeed Lake Concordia sources over 25% of its water from within 20 km of its shoreline. The source region for Lake Vincennes is by contrast more distributed about the basin. Along the steepest topography furthest upstream a thin film on the order of a millimeter should be capable of transporting most of the water produced. Bedrock fractures, if present may also aide in the transport of subglacial water in this region. Along the major lake chains in this area, however a higher capacity system is required even when flow is steady state. Recent work on further downstream has demonstrated that the classic ice-walled semicircular Rothlisberger channel is not a viable system throughout much of Antarctica. A broad distributed system akin to a braided stream has been demonstrated to be effective further downstream, but may be complicated by lack of sediment in the upper-most reaches. The ultimate system whether episodic or steady state, channelized or laterally migrating will affect the arrival, and release of water along the whole flow path and determine at what point the system starts to impact the flow of overlying ice.
C31C-0508
Greenland Ice Sheet Seasonal Speedup Coupled With Surface Hydrology
We use interferometric synthetic aperture radar data from western Greenland to quantify temporal variations in ice sheet flow and to characterise the ice sheet surface hydrology. In contrast to a recent study, our data reveal a non-uniform pattern of summertime velocity increase that, in places, extends over 100 km inland. Ice speedup is intimately linked to the routing of supraglacial water, and the magnitude of seasonal flow variations is positively correlated with the area of surface hydrological catchments. During late summer, ice beneath the largest catchments flows on average ~50 % faster than in winter, but beneath small catchments there is little or no change. Our results suggest that mass losses from the Greenland Ice Sheet under a warming climate will be governed by the extent to which coupling between hydrology and flow evolves, and that ground-based experiments to study velocity fluctuations should be sited with care.
C31C-0509
Increased Future Sea Level Rise due to Rapid Decay of the Greenland Ice Sheet?
In the Fourth Assessment Report (AR4) of the United Nations Intergovernmental Panel on Climate Change (IPCC), an increase of the mean global sea level by 18-59 cm for the 21st century is projected for the six SRES marker scenarios B1, B2, A1B, A1T, A2 and A1FI. The main causes for this sea level rise are thermal expansion of sea water and melting of glaciers and small ice caps, and to a lesser extent changes of the surface mass balance of the Greenland and Antarctic ice sheets. However, recent observations suggest that ice flow dynamics could lead to additional sea level rise, and this problem is explicitly stated in the AR4. These conjectured dynamical processes are (i) surface-meltwater-induced acceleration of basal sliding, and (ii) increased ice discharge due to reduced buttressing from surrounding ice shelves. The former process is probably more relevant for the Greenland ice sheet, whereas the latter may affect the stability of the West Antarctic ice sheet. On the observational side, recent results from satellite gravity measurements for the period 2002-2005 indicate surprisingly large mass losses of 239±23 km3 a-1 (0.66± 0.06 mm a-1 sea level equivalent) for the Greenland ice sheet. Furthermore, major outlet glaciers (Jacobshavn ice stream, Kangerdlugssuaq and Helheim glaciers) have sped up drastically during the last 15 years. It is attempted to quantify the range of uncertainty of future sea level rise due to dynamical processes of the Greenland ice sheet by numerical simulations with a high-resolution version of the 3-D dynamic/thermodynamic model SICOPOLIS. Results suggest that ice-dynamical processes can speed up the decay of the Greenland ice sheet significantly, but not catastrophically, in the 21st century and beyond.
C31C-0510
Hydraulic Fracture Along Glacier Beds by Turbulent Flow of Meltwater
The problem of hydraulic fracture has been studied extensively, with focus ranging from enhanced hydrocarbon flow to boreholes, to water-driven glacial cracking, to magma eruption through Earth's crust. Although some of this work has addressed fast-flowing fracture, the work applied to glaciers has, so far, focused either on static or relatively long timescale conditions. However, glaciological observations suggest that the fluid-induced fracture process may occur quickly, possibly driven by turbulently flowing water during crack growth. Here, we take the approximation of a fully turbulent flow into an elastic ice medium with small fracture toughness to derive an approximate expression for the crack-tip speed. We accomplish this by first showing that a Manning channel model for wall resistance to turbulent flow leads to the same mathematical structure as for resistance to laminar flow of a power-law viscous fluid. We then make use of the asymptotic crack solution for that case by Desroches et al. [Proc. R. Soc. Lond. A, 1994], and finally estimate the pressure scale appropriate for a finite crack. Comparison of this estimated solution with an exact self-similar solution of Adachi and Detournay [Int. J. Numer. Anal. Meth. Geomech., 2002] validates the approximation. To apply this model, we use parameter values thought appropriate for a basal crack driven by the rapid drainage of a surface meltwater lake near the margin of the Greenland Ice Sheet (Das et al. [Science, 2008]). Thus, we take a maximum excess crack inlet pressure of 0.9 MPa, corresponding to neglect of any hydraulic head loss in flow from the glacier surface to crack entry at the bed, a horizontal basal crack length of 1 km, and a wall roughness scale for flow resistance of 10 mm, and hence estimate a crack-tip speed of about 8 m/s. Loss of ten percent of the surface head on descent to the bed would reduce that speed by slightly more than ten percent. Making various plate theory and linear elastic fracture mechanics approximations perhaps relevant to this setting, we additionally model both vertical and horizontal surface displacements and find rough agreement with the meter-scale displacements observed through GPS by Das et al. [Science, 2008].
C31C-0511
Seasonal Evolution of Supra-glacial Lakes Across the Greenland Ice Sheet
We used 268 MODIS satellite images spanning the melt seasons 2003 and 2005-2007 to investigate the seasonal evolution of supra-glacial lakes in three different regions of the Greenland ice sheet. Lake area estimates were obtained by developing an automated classification method for their identification based on 250 m resolution MODIS surface reflectance images. Our dataset reveal widespread supra-glacial lake formation and drainage across the Greenland ice sheet, with a 2-3 weeks delay in the evolution of total supra-glacial lake area in the northern study areas compared to the south-western study area. The onset of lake growth varies by up to one month inter-annually, and lakes form and drain at progressively higher altitudes during the melt season. The annual peak in total lake area is positively correlated with modelled annual runoff across all study areas. Our results indicate that, in a warmer climate, supra-glacial lakes on the surface of the Greenland ice sheet can be expected to form earlier in the melt season and at higher altitudes than is presently the case. In consequence, the area and time period over which connections between the ice sheet surface and base may be established (Das et al 2008) will increase, potentially increasing the rate of ice sheet discharge and its sea level contribution (Zwally et al 2002). Das, S., Joughin, M., Behn, M., Howat, I., King, M., Lizarralde, D., Bhatia, M., 2008. Fracture propagation to the base of the Greenland Ice Sheet during supra-glacial lake drainage, Science, 5877, p.778-781. Zwally, H.J., Abdalati, W., Herring, T., Larson, K., Saba, J., Steffen, K., 2002. Surface Melt-Induced Acceleration of Greenland Ice-Sheet Flow, Science, 297, p.218-221.
C31C-0512
Influence of the Hydrogeological Setting on Englacial Conduit Morphology
Glaciological ideas about the character and evolution of englacial drainage systems have been deeply influenced by the theoretical model developed by Shreve (1972). The Shreve model is based on three main assumptions: (1) englacial drainage is in steady state; (2) englacial water will flow along the steepest hydraulic gradient within the glacier; and (3) pressure head equals the pressure of the surrounding ice minus a small component due to melting of the walls. The Shreve model has been widely adopted as a fundamental component of englacial drainage theory. There is no evidence, however, that the model provides a realistic picture of actual glacier drainage systems. To evaluate Shreve theory, we used speleological techniques to directly survey englacial conduits. We mapped a total of 8.5 km of passage in 45 distinct englacial conduits in temperate, polythermal, cold-based and debris-covered glaciers between 2005 and 2008. New information reported here is supplemented by published data on 40 other englacial conduits located worldwide and investigated to ice depths of 200 m using speleological techniques. In all cases, englacial conduits consisted of a single unbraching conduit. Englacial conduit morphologies were found to be intimately linked to the orientation of a glacier's principal stresses or the presence of pre-existing lines of high hydraulic conductivity. If a sufficient supply of water is available, hydrofracturing forms vertical conduits in zones of longitudinal extension and subhorizontal conduits where longitudinal stresses are compressive. On unfractured glacier surfaces, relatively shallow subhorizontal conduits with migrating nickpoints form by cut- and-closure provided channel incision is significantly faster than surface lowering. Conduits can also form along permeable debris-filled crevasse traces that connect supraglacial lake basins of different potential. Only conduits formed by extensional hydrofracture were found to be connected to glacier beds. Our results suggest that a Shreve-type englacial drainage system probably doesn"t exist and implies that englacial conduits can only penetrate through thick ice to recharge the bed of the Greenland Ice Sheet where supraglacial water bodies either intersect, or are advected through, zones of acceleration.
C31C-0513
A calving law for ice sheet models; Investigating the role of surface melt on dynamics of Greenland outlet glaciers
alving of icebergs accounts for perhaps as much as half the ice transferred from the Greenland Ice Sheet into the surrounding ocean, and virtually all of the ice loss from the Antarctic Ice Sheet. We have formulated a calving model that can be readily incorporated into time-evolving numerical ice-flow models. Our model is based on downward penetration of water-filled surface crevasses and upward propagation of basal crevasses. A calving event occurs when the depth of the surface crevasse (which increases as melting progresses through the summer) reaches the height of the basal crevasse. Our numerical ice sheet model is able to reproduce observed seasonal changes of Greenland outlet glaciers, such as fluctuations in flow speed and terminus positions. We have applied the model to Helheim Glacier on the east coast, and Petermann Glacier in the northwest. Our model suggests that rapid retreat of the claving front is highly affected by the amplified calving rate due to increasing water level in surface crevasses during warmer summers. Our results show little response to seasonally enhanced basal lubrication from surface melt. This modeling study provides insights into the role of surface and basal hydrology to ice sheet dynamics and on how to incorporate calving in ice sheet models and therefore advances our ability to predict future ice sheet change.
C31C-0514
Spatiotemporal dynamics of pure ice streams due to a multivalued sliding law
It is shown that a multivalued sliding law is naturally implied by
the transverse spatial structure of an ice stream velocity field.
It is then demonstrated that such a sliding law can lead to some
interesting stream-like patterns and time-oscillatory solutions.
A spontaneous generation of rapid stream-like solutions within a
slow ice-sheet flow, separated by narrow internal boundary-layers
(shear margins), is found using a combination of simple 1D model
and horizontal two dimensional numerical simulations over a
homogeneous bed and including longitudinal shear stresses.
The stress structure and detailed dynamical balance within the
different regimes around the stream solution are analyzed.
Different qualitative behaviors are obtained by changing a single
physical parameter, a mass source magnitude, leading to changes from
a slow creeping flow to a relaxation oscillation of the stream
pattern, and to steady ice-stream like solution.
http://www.people.fas.harvard.edu/~sayag/ST2008-movies-l.html
C31C-0515
Elevation Changes Over Large Antarctic Subglacial Lakes From ICESat Repeat Altimetry 2003-2008
Antarctica's recently described Recovery Lakes group (Bell et al., 2007, lakes A-D, about 13,300 km2) and
well-known Lake Vostok (15,690 km2) are crossed by numerous Ice, Cloud, and land Elevation Satellite
(ICESat) altimetry profiles. Since those laser measurements began in 2003, more than 13 repeats of each
profile have been completed, and preliminary elevation change results from ICESat are available. Because of
Lake Vostok's overall size and the Recovery Lakes relationship to the acceleration of ice into East
Antarctica's Recovery Ice Stream, analyses of lake levels have immediate relevance to broader ice sheet
stability assessments. Our study uses ICESat crossovers to assess elevation changes across these two
subglacial lake sites to compensate for inexact repeat profiles that cross the subglacial lake topography at
varying locations and angles.
Lake Vostok's relative stability, documented by a multi-year GPS network around Vostok Station (Richter et
al., 2008), makes it an excellent target for assessing repeat altimetry accuracy; ICESat results are compatible
with the GPS network results but reveal the difficulty in determining an elevation time series with cm-level
accuracy. Recovery Lake A appears to be rising, about 5 cm over 2004-2007, whereas Lake B shows no
clear pattern with respect to off lake elevations. Recovery Lakes C and D are also currently being evaluated
in preparation for Norway-US IPY traverse and planned airborne geophysical surveys by British and US
teams. Updated ICESat elevation time series results will be available prior to IPY fieldwork initiation.
Bell, R.E., M. Studinger, C.A. Shuman, M.A. Fahnestock, and I. Joughin, (2007), Large subglacial lakes in
East Antarctica at the onset of fast-flowing ice streams, Nature, vol. 445, pp. 904-907,
doi:10.1038/nature05554
Richter, A., S. Popov, R. Dietrich, V. Lukin, M. Fritsche, V. Y. Lipenkov, A. Matveev, J. Wendt, A. Yuskevich,
and V. Masolov (2008), Observational evidence on the stability of the hydro-glaciological regime of subglacial
Lake Vostok, Geophysical Research Letters, doi:10.1029/2008GL033397
http://traverse.npolar.no/
C31C-0516
Observing Outlet Glacier Motion Using High Rate GPS
Observations of short-term variations in the flow speed of marine-terminating outlet glaciers are essential for understanding the dynamics of these systems in relation to changes in surface melting and calving. Due to their inaccessiblity and highly crevassed and unstable surfaces, there is little existing ice motion data for polar outlet glaciers at high-temporal and spatial resolution over multi-week timescales. Here we present the results, successes, and planned improvements of a proof- of-concept study at Store Glacier,West Greenland in 2008 where we deployed both dual frequency and inexpensive L1 single frequency receivers. The dual-frequency system, was deployed roughly 15 km from the ice front near a large supraglacial lake and recorded at 1 s epochs. Three single frequency receivers, recording at 10 s epochs, were installed within 5 km of the ice front and transmitted their data to a base station on the side of the glacier. Dual frequency data was post processed using GAMIT's kinematic software package TRACK. However, multiple attempts at post processing the L1 phase data were unsuccessful due to multipath cycleslip intensity, massive clock drift, and data corruption, so that only the pseudorange data could be utilized. By combining the high-rate motion data with concurrent meteorology,time-lapse photography and seismographic data, we assess how variations in motion correspond with changes in meltwater production, calving and sea ice conditions at the front. Based on this initial deployment, we propose future improvements including increased data collection robustness, use of chokering antennae for multipath mitigation, broadcast GPS clock correction, and dual frequency disposable rover solutions for drastically improved positional analysis at the glacier front.
C31C-0517
Hydrological Modeling of Ice-Sheet Outlet Glaciers
Outlet glacier stability is recognized as a key control on ice-sheet mass balance. Observation of the development, drainage and dynamic effects of surface meltwater ponds near the equilibrium line of the Greenland ice sheet and on other polar ice caps has ignited interest in the coupling between surface climatology, englacial/subglacial hydrology and outlet glacier dynamics. Toward a more comprehensive model description of the hydromechanical processes that link surface conditions with basal processes, we present a two-dimensional flowband model of glacier hydraulics intended for coupling with a high-order thermomechanical ice-flow model. Surface meltwater is assumed to be stored primarily in ponds, which drain englacially if stress conditions permit. Water that reaches the bed encounters a dynamic subglacial drainage system, whose capacity evolves in response to water input. The impact of meltwater delivery to the bed therefore depends on the state of the subglacial drainage system and may vary in the presence of a seasonally evolving drainage system. The model is tailored to Belcher Glacier, a large tidewater outlet glacier of the Devon Island ice cap, Arctic Canada. As a polar tidewater glacier featuring the seasonal development and drainage of surface meltwater ponds, Belcher Glacier has much in common with some of the better- known outlet glaciers along the western margin of the Greenland ice sheet. We demonstrate the model behavior under conditions appropriate to ice-sheet outlet glaciers, and attempt to quantify the supraglacial water fluxes required to produce a significant impact on basal dynamics as a function of glacier geometry and thermal structure.
C31C-0518
Sub-daily glacier flow variations at Helheim Glacier, East Greenland, using GPS
Observations that are high resolution in both space and time at outlet glaciers are key for improving our understanding of the physical processes that govern glacier flow variations. We collected high-rate GPS observations simultaneously at several locations distributed along and across Helheim Glacier, East Greenland, during three Arctic summers between 2006--8, as well as other geophysical observations, to study glacier earthquakes and dynamics. GPS-derived position estimates of centimeter-level precision reveal the surface expression of glaciological signals, occurring from sub-hourly to daily time-scales and beyond, such as iceberg calving and ocean tidal forcing modulations. We will investigate a suite of relevant signals contained in the GPS time series and discuss how those signals may place new constraints on models for glacier flow.
C31C-0519
Velocity changes in the ablation zone of the Greenland ice sheet.
Continuous Global Positioning (GPS) observations are used to study variations in the flow of the western ablation zone of the Greenland ice sheet. Velocities increase by a factor 4 within days of increased melt water production. Over a longer period of 17 years annual ice velocity in the region decreased slightly, suggesting an adjusting hydraulic system, where increased meltwater input increases the efficieny. Results along the K-transect are now available for a 3 year period with hourly low precision GPS measurements. At the same time weather station are operated at three sites including Sonic Height Ranger instruments which are used to study hourly ablation changes. In this presentation we will address the interannual variations in the flow characteristics.
C31C-0520
Seasonal and diurnal melt-induced flow dynamics at a land terminating outlet of the Greenland Ice Sheet
It has recently been unequivocally established that surface generated meltwater can penetrate ~1km of polythermal ice within 40km of the margin of the Greenland Ice Sheet (GrIS), hydrologically lubricating the bed and inducing significant flow acceleration. Moreover, recent satellite and geodetic observations indicate that the coupling between surface melt forcing and ice flow response is potent, leading to between 50 to 400% (seasonal to weekly) ice speedup across extensive grounded, land-terminating sectors of the western GIS margin. We present data from a permanent network of kinematic GPS and AWS deployed across the wider Russell Glacier Catchment in 2007. These results indicate strong diurnal cycles of horizontal and vertical displacement that are synchronous with each other but which lag surface ablation by a few hours. The fluctuations are strongest near the ice margin but are observed, even if dampened, some 100 km upglacier at 1500 m a.s.l., well above the Equilibrium Line Altitude. Such phase-locked diurnal coupling between uplift, flow and ablation indicates that surface generated meltwater is directly accessing the bed of the ice sheet over a larger area than previously considered. Such a result has significance not only for land- terminating sectors of the GrIS but also marine outlet reservoirs since a well lubricated bed is a prerequisite for effective longitudinal coupling and the potential inland migration of enhanced flow, which could the increase dynamic drawdown of the ice-sheet interior under a warming climate.
C31C-0521
The Importance of Ablation on ice Flow and Meteorite Exhumation: the Physical Modeling Approach at Frontier Mountain, Antarctica
The highest concentration of meteorites yet discovered on Earth is found in the ice sheet covering Antarctica. Major meteorite accumulation zones often occur in front of submerged or emerged bedrock obstacles, where the meteorite-bearing ice slows down, is uplifted by the buttressing effect and exhumed and concentrated by wind ablation ("ice-flow model"). Meteorite traps have also been discovered in the downstream side of major emerged bedrock barriers, as in the Frontier Mountain (FM) region, a nunatak outcropping in the Northern Victoria Land, Antarctica, 250 km from the Italian Terra Nova Base. However, recent detailed glaciological analyses indicate that also for this site the "ice-flow model" remains the best concentration explanation, being present an important submerged barrier in the main blue-ice field. A this site, during the last fifteen years, more than 1000 meteorites have been there collected. During the various field expeditions, different data were acquired to allow a detailed study of the local glaciodynamics. This large data set was used to constrain boundary conditions for performing a set of physical experiments reproducing the main glaciodynamics characteristics of the FM region. In addition to previous result, now we performed a series of experiments that consider also the ablation effect in order to better reproduce the natural environment. Analog experiments were performed at the Tectonic Modelling Laboratory of the CNR-IGG (Florence, Italy). Polydimethilsyloxane (PDMS) is used to simulate glacial flow in analog models; this material properly simulates the rheological behaviour of ice (Corti et al, 2003; Corti et al, 2008). Models are built inside a Plexiglas box with dimensions of 70cm x 20cm x 10cm. The models are scaled to nature conditions allowing a comparison between laboratories and natural models. The geometrical scaling ratio was of 2*10-5, such that 1 cm in the model represented about 500 m in nature. In order to measure the progressive ice deformation during the experiments, passive grids of carbon-black particles are printed both inside and on the model surface using the unbaked photocopy method. In a first series of experiment we tested the flow variation for submerged obstacle with dimensions 3cm x 10cm x 1÷5 cm placed inside the Plexiglas box which is inclined of 3°; the PDMS is allowed to flow by opening the front end of the box. Afterward we introduced the ablation effect in the model with emerged obstacle for two natural case: upstream and downstream of the bedrock obstacle. In these models, ablation was simulated by physically removing pieces of PDMS from the surface with a small knife at regular intervals of 30 minutes in order to maintain a constant depression on the free surface of the ice. Later we used the Frontier Mountain bedrock topographic model used for previous work with addition of ablation downstream the mountain outcrop, as in the natural environment. Experimental results show how the flow field and variations in the topography of the free surface and internal layers of the ice are strongly influenced by the presence and height of bedrock obstacles, but only limited uplift of internal layers is observed in these experiments. The exhumation of deep material embedded in the ice is observed only if ablation is be included in the physical models. In this case, the analogue ice replenishes the area of ablation (simulated by material removal), thereby allowing deep layers to move vertically to the surface and severely altering the local ice flow pattern.
C31C-0522
Water Flow and Lake Drainage Beneath Antarctica
We use 5 km resolution surface and bed DEMs of Antarctica to calculate the subglacial hydraulic potential, and location of drainage catchments and major drainage pathways for the Antarctic Ice Sheet. We find close correlations between sinks in the hydraulic potential, the location of major drainage pathways and the position of known subglacial lakes. We use a thermo-mechanical ice sheet dynamics model with an assumed geothermal heat-flux to calculate energy supply and melt rates beneath the ice sheet. Accumulating this water along the drainage pathways allows us to calculate the steady state water flux to all known lakes (which range between ~ 0.1 and ~ 10m3 s-1) and in the drainage pathways as they enter the ocean (which range from ~ 1 to ~ 100m3 s-1). For different assumed lake drainage event discharges, we estimate a range of jokulhlaup frequencies for each lake. For the observed 1996-8 Adventure Subglacial Trench lake drainage event of ~ 1.8 km3, we estimate a flood frequency of 25 years. Finally, we use Nye's (1976) theory to model the time dependent discharge associated with the Adventure Trench jokulhlaup and can match theory to the observed surface altimetry data for realistic values of initial conduit diameter and roughness.
C31C-0523
Analysis of Spatial and Temporal Variations in Strain Rates Near Swiss Camp, Greenland
We present results from a two-year study of strain-rate variations along a flow line on the western margin of the Greenland ice sheet. We used baseline network solutions to investigate variations in longitudinal (along- flow) strain rates over the 2006 and 2007 melt seasons. Analyses revealed high-magnitude, short-duration events of increased longitudinal strain early in the melt season coincident with intensive melt, suggesting a hydrologic link. Results from 2006 show that longitudinal strain rates became variable shortly after the onset of melt (day 186) changing up to ~ 15 × 10-4 a-1 over background rates within 24 hours. The onset of melting occurred earlier in 2007 (day 153) and was also followed closely by strain-rate deviation. The data reveal rapid (hours to days), high-magnitude (two to ten times greater than background strain) changes in longitudinal strain rates (hereafter referred to as 'high-strain' events) that occurred both on the small-scale (affecting 1-4 baselines,<10km along flow) and less frequently (two per season) on the large-scale (affecting 5 or more baselines, ~10 km or more along flow). Events were likely caused by drainage of supraglacial meltwater that penetrated to the base of the ice, raising the basal water pressure leading to reduced basal resistive stress and rapid local acceleration. The local change in stress also caused non-local effects in adjacent ice including longitudinal compression of the ice down glacier and longitudinal extension up glacier with respect to the event origin. High-strain events altered longitudinal strain rates more than 15 km along flow from the site of initiation and indicate that short-term altered stress conditions are not confined to the ablation zone. Results suggest that seasonal high-strain events have the ability to alter longitudinal baseline length, allowing a greater ice flux to lower elevations where melting occurs for a larger portion of the year. However, the cumulative seasonal effects of both large-scale and small-scale strain events are modest, and indicate that seasonal changes in strain rates have only a minor effect on the overall stability of the ice sheet at this location. Nevertheless, it is possible that over much longer timescales these seasonal changes may become more important with increasing temperatures and available melt.
C31C-0524
Premelting and the Water Budget in Polycrystalline Ice
A number of mechanisms, generally classified as premelting are responsible for the presence of liquid water at ice interfaces at temperatures well below 0°C . Premelting includes the familiar colligative effects of ions and other impurities, which lower the chemical potential of the liquid solvent, and the Gibbs-Thomson effect which describes the lowering of the melting point for a solid convex into its melt. Such phenomena are known to influence the amount of water in natural and laboratory polycrystalline ice and to control the thermal, chemical, and material transport properties. Thus, liquid water within the solid ice matrix influences the behavior of terrestrial ice over a wide range of length and time scales, from the macroscopic behavior of temperate glacier ice to the distribution of climate proxies within polar ice sheets. Using optical microscopy observations of ice near its melting temperature, rough bounds have been put on the length scales and dihedral angle associated with the liquid network in ice. However, these techniques cannot resolve whether the boundary between any two grains is wet or dry. For this, a more refined light scattering method has been developed. This method and the results are described both in the context of the basic physics and the application to the geophysical setting. The importance of this approach is broad, with implications ranging from the understanding of the role of intermolecular forces in the wetting properties of the ice/ice interface to constructing a budget for the total amount of water in an ice sheet. Additionally, basic applications of grain boundary melting are important in fields from metallurgy and materials science to mineral physics and geoengineering.
C31C-0525
Assessing ice-sheet and sea-level changes: Implementing dynamic subglacial processes in a high-order ice sheet model
Current predictions of ice sheet mass balance and discharge variability including sea-level rise, are severely limited because subglacial processes are not reliably implemented in large-scale ice sheet models. One of the most limiting factors is the challenging nature of simulating the fast and transient flow of ice streams that drain the interior and discharge large volumes of ice into the polar ocean. This difficulty is partly due to uncertainties in the nature of subglacial processes, but also due to the complexity in including local basal processes at the scale computationally required to run a whole ice sheet model. We aim to improve the predictive capability of a community ice sheet model (GLIMMER) for the Antarctic Ice Sheet by implementing higher order ice flow physics and introducing dynamic subglacial processes that include hydrologically controlled shear strength evolution in a till layer with Coulomb plastic rheology. Here, we focus on the development of a subglacial processes model coupled to a large-scale ice sheet model. The coupling between both models is done via the determination of the bed strength, which is assessed from water availability within the till layer. Initial experiments have been carried on a simple idealized domain. This allow us to do sensitivity tests on a range of model parameters (e.g., sediment type, thickness and distribution) and to do some model tuning before applying the calculation to the whole ice sheet domain. In addition, a new geology mask has been defined for Antarctica, differentiating three classes of basal geology using bed elevation together with observed ice velocities. Additional modelling developments will include the generalization of the subglacial processes model by coupling pore water pressure to a regional hydrology model. We anticipate that our model will be able to reproduce the inland thinning observed along the Amundsen Coast as well as the stagnation of ice streams in the Ross sector. We therefore expect our model to produce reliable sea-level rise prediction for the 21st century.
C31C-0526
Coupling Glacial Hydrology into a High-order Numerical Ice Model
A two-dimensional thermomechanical flow-band model is presented that comprises components solving the momentum equations, the advective-diffusive heat equation, and the evolution of the free surface. Longitudinal stretching and lateral shearing play an essential role in the dynamics of large outlet glaciers, especially near the margins and when basal sliding is considerable. Therefore, this model properly accounts for these high-order stress gradients. Model performance is compared with benchmark solutions using model intercomparison exercises and analytical solutions under simplifying assumptions. The model is applied to the Belcher Glacier on the Devon Island ice cap in the Canadian Arctic. Sharing similarities with many Greenland outlet glaciers, this is a large, fast-flowing, tidewater glacier. As such, various basal sliding laws and iceberg calving parameterizations have been incorporated and examined within the model. Surface meltwater, through supraglacial/englacial drainage networks, has an important contribution in the dynamics of polar tidewater glaciers. We therefore incorporate key glaciohydraulic processes into the ice-flow model. These include drainage of supraglacially-stored water through englacial fractures. Under permissible stress and water influx conditions, meltwater can drain to the bed, providing lubrication to the sole of the glacier. We demonstrate how this process affects model behavior under geometry and meltwater conditions appropriate for polar glaciers.
C31C-0527
Investigating Crevasse Structure Impact on Glacial Sub-Surface Ice Temperature Distribution with Implications for Moulin Formation
Ice flow acceleration in the ablation zone of outlet glaciers in Greenland has been linked to an increase in infiltration of surface melt through moulins. In order to understand the potential for moulin formation within cracks and crevasses in glacial ice and their impact on melt infiltration rates it is important to gain knowledge about how crevasse structure influences sub-surface ice cold content. A tower of seven wireless sensors was deployed in a moulin at the base of a crevasse in Svinasfellsjokull, glacier in Iceland as an analog to those in Greenland to measure temperature and incoming solar radiation at depth. Seven Crossbow® Environmental Motes (MEP410 Models) measuring temperature and irradiance (integrated from 0.4- 0.6μm) were deployed and inserted in a moulin at a depth of 156.8 cm for three days. Temperature measurements indicate greater diurnal fluctuations near the surface of 0.39 °C from the near surface mote, while sub-surface temperature remain stable near mean temperature of -0.05 °C at a depth of 139.8 cm. Incoming solar radiation measurements showed diurnal variation as expected near the surface at 5.30 W/m2 in the top of the sensor tower, whereas no variation occurred at other depths as the average was 0 W/m2. A computational fluid dynamic model has been calibrated by measured mote temperatures. The model was run accounting for diffusion and convective processes and forced with meteorological data to explore how crevasse density, depth, and geometry influence sub-surface ice temperature distribution.