C31B-0486
Accumulation Patterns and Basal Conditions from Radar Observations Along the US- ITASE Traverse in East Antarctica
We present short-pulse radio echo reflection profiles recorded along a largely cross-flow transect during the 2006-08 US-ITASE traverse from Taylor Dome to South Pole. We used a 200-MHz system to focus on firn strata within the upper 100 meters and a 4-MHz system to focus on englacial strata, as well as the bed. Within the catchment area of Byrd Glacier unusual cross-cutting englacial horizons, alternating with metamorphic layers representing hiatuses in accumulation, suggest changing patterns of accumulation, possibly caused by long term wind shifts. Profiles recorded by both systems from south of the Byrd catchment area to Titan Dome are conspicuously characterized by buried megadunal type stratigraphy, consistent with the traverse being downflow from the apparent eastern edge of the modern megadunes region. The 200-MHz profiles reveal buried and partially buried foreset beds extending over 25 km in length, up to 50 m thick and sandwiched between 4 to 15 m thick unstratified layers up to 60 km long. These beds are antidunal deposits, and the unstratified layers are glazed, zero accumulation zones of extreme, long term metamorphism. The deeper, 4-MHz profiles show prograding beds occur to 1700 m depth, up to 250 m thick along a segment on the northeast fringe of the megadunes. These impressive features, both near the surface and at depth, show the concentrated way snow has been, and now accumulates, in the megadunes area and well east of it. If these beds formed during high accumulation Holocene warming periods, then we need to understand how they could be interleaved with features characteristic of zero accumulation. The 4- MHz detection of prograded beds must be made possible by embedded, volcanic high conductivity layers, which reveals that these beds contain datable layers. We have also plotted the power received from some 300,000 bed echoes versus ice thickness along the traverse. We use these data to estimate the average 4-MHz signal absorption rate within the ice, which allows us to calculate bed reflectivity and interpret basal properties. The results suggest many areas of thawed basal conditions, generally thought to be rare in East Antarctica. They also depict bed reflectivity at several candidate subglacial lake sites based on satellite imagery and reveal at least one possibly new subglacial lake.
C31B-0487
Megapackets and Megapseudosynclines Profiled With GPR East of the Megadunes Region, East Antarctica
We have used GPR during the ITASE07 program to profile sequential packet formation east of the megadunes region. Packets presently refer to the 8 to 15 m thick prograding beds sandwiched between 1 to 4 m thick metamorphic layers previously profiled with GPR in the megadunes region. The prograding beds migrate over the metamorphic layers. Much larger megapackets occur along a 530 km southerly transect that started at latitude 81.7 degrees, far west of the Transantarctic Mountains. The first 130 km was orthogonal to ice flow, in an east fringe bordering the megadunes region, and parallel to dunelike ridges. Along this segment our 90 m deep 200 MHz profile reveals metamorphic layers, one of which may extend 140 km, but no well defined prograded packets, as expected for the transect orientation. The next 400 km diverged east from the megadunes to a maximum distance of 220 km. In contrast with the 2 to 6 km megadune wavelengths and their 8 m maximum amplitudes, hills along this segment are spaced 10 to 30 km with local elevation differences up to 40 m; our 4 MHz profile suggests bedrock controls this topography. Packets originate at the south-facing windward slopes. Sequential metamorphic layers up to 60 km long occur, some with horizons from buried glaze. Several spectacular megapackets occur, one of which contains 25 km of prograding beds up to 50 m thick and associated with metamorphic layers up to 15 m thick. The entire transect has a few depositional megapseudosynclines well over 100 m thick and from which progradation extends north. These features depress and override metamorphic layers. The 4 MHz profile reveals even larger megapackets, including one 250 m thick that reaches more than 1000 m deep, and a 70 km metamorphic unconformity that dips over 200 m. Glaze, underlain by metamorphism, covers about 46 percent of the entire transect. By 86 degrees latitude, at 530 km, the metamorphic layers fade, and simple layering forms after another 200 km. The 25 km prograded deposits must represent extremely high deposition rates, given the concentration at windward faces. The megapseudosynclines should be ideal core sites to obtain deposition history. Clearly, packet stratigraphy extends well beyond the megadunes area.
C31B-0488
Coupling snow densification and melt-water retention in a large-scale ice sheet model.
The physics of snow and how the annual snow changes during a melting season is important for the surface mass balance of the Greenland Ice Sheet. Densification when meltwater is present will generally go much faster than the dry snow densification due to percolation and refreezing of meltwater in the snow-pack and it needs to be considered in ice sheet models in order to calculate the ablation more correctly. Simple parameterizations are used to calculate surface melt, snow densification and meltwater retention of the annual accumulation rate expressed as surface boundary conditions in a large-scale ice sheet model of Greenland. Coupling the snow densification and meltwater retention processes makes it possible to separate volume and mass changes of the surface layer. To calculate the surface melt contribution to runoff only mass changes of the surface layer is used. Experiments for present-day and future scenarios show that snow depth at the onset of melting, mean annual surface air temperature and the density of the annual snow layer are key factors controlling the quantity and location of meltwater runoff above the equilibrium line in Greenland. Possible future changes in the volume of the Greenland Ice Sheet, as changes in the total air content in the firn, will also be presented.
C31B-0489
Hydrologic drainage of the Greenland Ice Sheet
A simple hydrologic drainage network for the Greenland Ice Sheet is modeled from available DEMs of bedrock and surface topography and assumptions of hydrostatic water pressure, uniform hydraulic conductivity, and no conduit flow within the ice sheet. As such it is a first-order model best suited for broad- scale hydrological assessment. Results identify 293 distinct hydrologic basins together with their "realized" (wet) and "unrealized" (dry) drainage patterns. Intersection with 1991-2000 Polar MM5 mesoscale climate model hindcasts of meltwater runoff suggests that these basins route varying amounts of water to the ice edge. While basins in the Jakobshavn region produce the greatest total volume of meltwater, area- normalized runoff is highest in Julianehab (55 cm/yr, vs. 12 cm/yr in Jakobshavn); and lowest in Tunu and a newly proposed "Eastern Humboldt" hydroclimatic region (3-4 cm/yr). The extent to which meltwater truly reaches the ice margin as modeled is difficult to test. However, the simulated flow outlet locations show general, if qualitative, agreement with the locations of 460 observed meltwater outlets (proglacial lakes, streams, and rivers; and sediment plumes into fjörds) mapped continuously along the ice-sheet perimeter. On average, about 36% of the modeled drainage network was activated (i.e. received water) over the 1991- 2000 study period. Remaining areas, barring dynamic changes to ice-surface topography, would activate if surface melt penetrates deeper into the ice-sheet interior. Both new datasets are freely available for scientific use at the National Snow and Ice Data center [metadata submitted 8/2008; final URL to be provided upon publication].
C31B-0490
Change in Firn Densification Rates to 80 m Depth Across the Percolation Zone of Western Greenland
We conducted ground based georadar surveys at thirteen locations along a 70 km long transect of the EGIG line in the percolation zone of the western part of the Greenland Ice Sheet. The purpose of these surveys is both to gain an understanding of the hydrologic pathways of surface generated meltwater as well as to measure densification of the upper 80 m of the percolation zone, and how densification rates change with elevation. To determine the density of a layer we assume a lack of liquid water and use the CRIM equation to solve for overall density from electromagnetic (EM) velocity. In this study we used three methods to solve for the EM velocity 1) Common Mid-Point (CMP) reflection analysis solving for velocity with a ray-based model inversion technique, 2) multi-offset, single line acquisition solving for velocity with a reflection tomography inversion technique, and 3) impedance analysis of single offset data to determine radar velocity variation with depth along a single trace. We show that near surface density varies laterally over tens of meters and that density curves calculated from CMPs are representative of measured core densities. The CMP density curves are smoothed over depth and represent volumes that are orders of magnitude larger than traditional core density curves; therefore they may be better measures of densification rates for regions of the Ice Sheet. We also found that firn density increases at significantly higher rates at lower elevation sites and that the rates at lower sites deviate substantially from standard firn densification curves.
C31B-0491
Present-day West Antarctic ice-mass change estimate by the constrained inversion of GRACE and InSAR data
We estimate the ice-mass loss in West Antarctica from the Gravity Recovery and Climate Experiment (GRACE) using an inverse gravimetric approach. The inversion is constrained by the InSAR data to minimize the ambiguity of the resulting mass estimates. We determine the spatial resolution of GRACE over Antarctica and use it to calculate unconstrained (GRACE only) mass balance estimates for a reduced number of drainage basins. We compare InSAR, constrained and unconstrained mass change rates, estimate their uncertainties and discuss the influence of glacial-isostatic adjustment on our results.
C31B-0492
Rapid Response of Sediment Plumes to Greenland Ice-Sheet Surface Melt
Quantifying the mass balance of the Greenland ice-sheet in a warming climate intensifies the need for hydrologic observations of meltwater exiting the ice-sheet. Traditional observational methods such as stream discharge measurements are sparse due to the remoteness of Greenland, therefore alternative methods must be explored. We hypothesize that hydrologic pulses of ice-sheet runoff are registered as sediment plumes in downstream estuaries. This study uses Moderate Resolution Imaging Spectroradiometer (MODIS) satellite imagery calibrated by field data to quantify the area and maximum Suspended Sediment Concentration (SSC) of plumes; and explores the terrestrial and oceanic drivers of plume formation in the Kangerlussuaq Fjord, southwest Greenland. Terrestrial pulses of surface meltwater are the predominant driver of sediment plumes rather than ocean tidal cycles. A one-day lag between the timing of melt pulses on the ice-sheet and subsequent SSC pulses in the estuary suggests a causal relationship between the two and an unhindered transfer of surface meltwater out of the ice-sheet. Supraglacial lake drainages are also associated with plume development, but their temporal coincidence with surface-melt episodes confounds partitioning of their relative importance to ice-sheet drainage. We conclude that remote sensing of sediment plumes offers a useful tool for studying hydrologic mass losses from the Greenland Ice Sheet.
C31B-0493
Temporal and Spatial Variability in Snow Accumulation at WAIS Divide over Recent Centuries
Temporal and spatial variability in snow accumulation was evaluated using ice cores collected in 2000 (ITASE 00-1) and 2005 (WDC05A, WDC05Q) from the West Antarctic Ice Sheet Divide (WAIS Divide) project site. The ice cores were dated based on annual layer counting of multiple glaciochemical measurements resulting in mean annual accumulation rates of 0.200, 0.204, and 0.221 mweq yr-1 for WDC05A, WDC05Q, and ITASE 00-1, respectively. An analysis of variance of accumulation in the ice core array was used to determine the small scale spatial variability (SSV), allowing for quantification of uncertainty in an individual accumulation record. The spatial variability for the WAIS Divide project site was calculated to be 0.030 mweq yr-1, or approximately 15% of the average annual accumulation. An accumulation record representative of the WAIS Divide local area over recent centuries was developed using a principal component analysis of the array to identify the coherent accumulation signal. The WAIS Divide local area accumulation record did not exhibit a trend over recent centuries, with 14% inter-annual variability (1-standard deviation of the mean) and SSV of 0.017 mweq yr-1. Correlations between the WAIS Divide local accumulation record and atmospheric indices (e.g., Antarctic Oscillation and El-Nino Southern Oscillation) exhibited periods when the records oscillate in and out of phase. Thus, identifying relationships between WAIS Divide accumulation records and geophysical processes over recent centuries may be problematic.
C31B-0494
Mass-balance measurements from a network of automatic weather stations in the ablation zone of the Greenland Ice Sheet
The most accurate way to record continuous mass balance variations for a specific location is by placing an automatic weather station (AWS). In spite of this, the ablation zone of the Greenland Ice Sheet had a poor spatial coverage of these measurement systems, as changing ice surfaces, strong winds, etc. may call for frequent visits, and therewith high logistical expenses. However, we are in the process of building a comprehensive network of AWSs in the ablation zone (part of the PROMICE programme for monitoring the Greenland Ice Sheet), increasing the number of permanent transects around the ice sheet from two to nine. We will present preliminary mass-balance data of the stations placed during the previous two summers, focusing on the summer of 2008. After completion of the network in 2009, the measurements will serve as input for a melt model run over the entire ice sheet. Another ambition is to use the data to validate regional climate models. We will discuss the uncertainties of such modeling in the ablation zone, and possibly compare model results for 2008 to our observations.
C31B-0495
ICESat elevation changes over Antarctica and comparison with GRACE mass change
The ICESat and GRACE missions have been collecting coincident data since 2003, providing valuable information into the elevation and mass changes of the Earth. The observations are particularly valuable for regions such as Antarctica where the near-polar orbits of both satellites ensure a high density of measurements. The first goal of this study is to examine some of the many corrections and processing strategies for ICESat that can have a significant influence on elevation change estimates. For the ICESat data, major sources of uncertainty include the calculation and understanding of inter-campaign biases, and a comparison of the relative merits and errors between height rates derived from crossover versus repeat track analysis. Comparisons will examine a variety of spatio-temporal scales, including multi-year and seasonal trends and basin-scale areas. The second goal of this study is to discuss similarities and differences of the mass change estimates generated from both ICESat and GRACE missions. The measurements from ICESat can be converted to mass change estimates, with high spatial resolution; however, this relies on an accurate glacial isostatic adjustment (GIA) model and assumptions regarding firn density. GRACE can directly observe mass changes at monthly time intervals, but the interpretation of the results is limited by the spatial resolution of the measurements (~500 km) and the proper removal of the large mass effects of GIA. The results show that the two independent data sets possess strong spatial correlations. The total ICESat- derived surface mass variability is less than that determined by GRACE, although they match within their error bounds. These differences highlight the need for further model tuning and investigation. The similarities that do exist indicate a strong potential for using the combination of both missions for future refinement of both GIA and surface mass balance estimates, and highlight both missions as key measurement tools for understanding critical climate processes.
C31B-0496
GRACE Observes Small-Scale Mass Loss in Greenland
Using gravity data from the GRACE satellites between February 2003 and January 2008, we examine changes in Greenland's mass distribution on a regional scale. During this period, Greenland lost mass at a mean rate of 179±25 Gt/yr, equivalent to a global mean sea level change of 0.5±0.1 mm/yr. Rates increase over time and are driven by mass loss during the summers, which vary substantially over the years. The largest mass losses occurred along the southeastern and northwestern coast in the summers of 2005 and 2007, when the ice sheet lost 279 Gt and 328 Gt of ice respectively within 2 months. In 2007, a substantial mass loss is observed during summer at elevations above 2000 m, for the first time since the start of the GRACE observations.
C31B-0497
Surface Meltwater Runoff and Retention in the Accumulation Zone of the Greenland Ice Sheet
Surface meltwater transport on the Greenland Ice Sheet has important implications for mass balance. Melt runoff is the principal loss term in the surface mass balance, and any meltwater that reaches the englacial and subglacial hydraulic systems can contribute to sliding and dynamic losses from the ice sheet. In the accumulation zone of a glacier or ice sheet, some fraction of melt is retained each year by percolation into the underlying cold snow and subsequent refreezing of that melt. The runoff limit is defined as the elevation below which all surface melt is able to leave the ice sheet and count as mass loss. Most ice sheet mass balance models make use of meltwater retention parameterizations to determine the runoff limit, but the runoff limit has never been directly measured. Lateral routing of meltwater is important for runoff calculations because any melt produced at the surface must be able to leave the ice sheet to count as mass loss. A physically based snowpack model combining day-time surface melt percolation with nighttime refreezing was used to simulate the effects of multi-year cycles of melt, infiltration and refreezing on snowpack physical properties and downslope flow of meltwater. Increasing accumulation during the melt season leads to a higher frequency of melt layers in the near-surface snow and a greater fraction of surface melt traveling downslope. The temperature conditions in the snowpack control the lateral distance this melt can travel. Previous meltwater retention models have taken into account only seasonal melt and refreeze cycles, but the daily cycles in this model indicate that nighttime refreezing has a large effect on lateral melt transport by inhibiting the distance melt can travel down the ice sheet. Recent field data and snow temperature measurements from the Greenland Climate Network indicate the snowpack refreezes most nights during the melt season in the accumulation zone. Our results suggest the maximum lateral distance melt can travel is only a few kilometers per melt season. These findings indicate much less runoff from the accumulation zone of Greenland than previously estimated.
C31B-0498
Spatial and Temporal Trends in Snow Accumulation From Radio Echo Eounding, Summit, Greenland
Current estimates of snow accumulation over Greenland have large errors (20-25%) because they are derived from a relatively sparse network of point measurements (Ohmura and Reeh, 1991; Bales et al., 2001). To determine whether the Greenland ice sheet mass is increasing or decreasing and how this will affect the global sea level, the Center for Remote Sensing of Ice Sheets (CReSIS) at the University of Kansas has developed several ice penetrating radar systems. Ground based ultra-wideband radar (500-2000 MHz) operated near Summit, Greenland, in July 2005, is used to map near-surface internal layers with 10 cm free- space resolution. This high resolution allows for visual inspection of accumulation layers to a depth of over 200 meters. Radar transects connecting the GRIP and GISP2 ice cores reveal continuous reflection horizons that allow for the transfer of age-depth relationships obtained from the ice cores to the continuous radar reflections. Accurately dated and spatially continuous isochrones are valuable for calibration and verification of ice sheet models. The observed isochrones provide a detailed description of spatial and temporal variations in accumulation rate over the past 500 years and constrain the selection of parameters and climate history used to force numerical models.
C31B-0499
Does Warming of the North Atlantic over the Last Decade Explain the Acceleration of Outlet Glaciers in Southeast Greenland?
The Greenland Ice Sheet's contribution to sea level rise more than doubled in the last five years, mostly because of increased mass flux rates from outlet glaciers in southeast Greenland. These outlet glaciers terminate at tidewater in deep fjords, which provides an intimate connection between the ice sheet and the ocean and raises the possibility that ocean warming was the trigger for recent changes in ice dynamics. Until recently, however, the oceanic role was unknown since there was no evidence that the warm waters of tropical origin, found offshore along the continental margins of southeastern and western Greenland, could cross the cold, fresh Arctic waters found on the shelf, and penetrate deep into fjords. We provide evidence that warm offshore waters both penetrate and circulate deep inside Sermilik Fjord, the 100 km long fjord in East Greenland where Helheim Glacier terminates, based on a series oceanographic surveys conducted in the summer of 2008. Furthermore, the depth at which these waters are found, as well as the observed spatial and temporal variability all indicate that the warm waters play an active role in the fjord-glacier estuarine system and suggest that they come into contact with the glacier's terminus. Then, using historical oceanographic and glaciological data we argue that the timing of Helheim Glacier's acceleration is consistent with the variability in warm water properties found offshore. Finally, because Sermilik Fjord and Helheim Glacier are typical of many fjord-glacier systems in southeast Greenland, we propose that glacier-ocean interactions can explain a significant fraction of the increased mass flux from the Greenland Ice Sheet.
C31B-0500
Snowpack and Firn Densification Gradients in the Percolation Zone of the Greenland Ice Sheet: Implications for Estimates of ice Sheet Mass Balance
The proportion of surface generated meltwater that subsequently refreezes in the snowpack and firn plays a critical role in controlling the mass balance of polythermal ice masses. In Greenland, changes in the volumes of meltwater that refreeze in the superimposed and percolation zones are likely in response to any future climate change. However, processes of meltwater percolation and refreezing may cause a re-distribution of mass through densification, resulting in surface elevation change but with no associated mass loss. Seasonal densification gradients caused by variations in altitude (and thus temperature) must therefore be accounted for to reduce errors associated with mass balance estimates derived from satellite and airborne altimetry platforms. Here, we investigate the spatial variability in seasonal densification in the snowpack and firn along a 57 km transect of the EGIG line (~67.5N, 47W) in the percolation zone of the Greenland Ice Sheet. Snowpack and firn densities (derived from shallow cores) were obtained from seven sites along the transect at approximately 10 km intervals and spanning a 350 m elevation range between 1680 and 2050 m. As expected, there was no significant variation in spring snowpack density (prior to melt) but following the summer processes of melt, percolation and refreezing, density decreased by 36 kg m-3 for every 100 m increase in elevation. Measurements over three melt-seasons at three sites spanning an 155 m elevation range reveal substantial variability in the seasonal densification gradient. The implications of the observed densification gradients for ice sheet elevation and thus mass balance estimates in a warming climate are discussed.
C31B-0501
Warm Atlantic water drives Greenland Ice Sheet discharge dynamics
Greenland outlet glaciers terminating in fjords experience seasonal fluctuations as well as abrupt episodes of rapid retreat and speed-up. The cause of abrupt speed-up events is not firmly established, but synchronous occurrences suggest that it is related to Arctic warming. Here, we report major warming of water masses in Kangerdlugssuaq Fjord, East Greenland, immediately prior to the fast retreat and speed-up of Kangerdlugssuaq Glacier in 2004-05. Our hydrographic data show that this event occurred when Atlantic water entered the fjord and increased temperature of surface water by 4°C and deep water by 1°C. On the basis of meteorological records and satellite-derived sea surface temperatures, which fluctuate by up to 4°C in periods of 2-3 years, we infer that inflow of Atlantic water is controlled by the direction and intensity of prevailing winds that force coastal and offshore currents. Our results demonstrate that Greenland Ice Sheet discharge dynamics are modulated by North Atlantic climate variability, which is identified by shifts in the position of atmospheric low pressure over the Labrador and Irminger seas. A persisting westerly position of the Icelandic Low since 1999 may explain why winters in Greenland have been particularly mild during the last decade and it is feasible that widespread and synchronous discharge fluctuations from outlet glaciers, which resulted in high rates of ice loss in southeast Greenland, are a consequence of this synoptic condition.
C31B-0502
Using Spaceborne Laser Altimetry to Assess Elevation Change in Arctic Canada: Implications for the Major Ice Sheets
Although tasked primarily with monitoring spatial and temporal changes in the topography of the Greenland and Antarctic ice sheets, the Ice, Cloud and land Elevation Satellite (ICESat) has provided a unique platform from which to observe other ice caps and glaciers at high latitudes, including those in Arctic Canada, since 2003. Notably smaller than their ice-sheet counterparts, these regions may still make significant contributions to sea-level change in the next few decades. Despite the advantages that ICESat has in observing these areas, coverage remains sparse, due to relatively low latitudes, in certain cases, and to the presence of clouds, which varies seasonally. Furthermore, the steep surface slopes that characterize such ice caps and glaciers introduce significant errors into the derived elevations. Similar conditions exist at the margins of the larger ice sheets, where limited data coverage and high slopes also combine to inhibit the assessment and interpretation of elevation change observed by ICESat. We obtain our estimates of local elevation change rates using a repeat-track analysis method, in which ICESat elevations from multiple mapping campaigns are binned along reference ground tracks, and illustrate it with results from Axel Heiberg Island, in the Canadian Arctic Archipelago. We also employ ASTER imagery to classify the various ice-covered surfaces, and to extract digital elevation models, which are used to constrain estimates of local slope in conjunction with those for elevation change. We identify and discuss specific error sources, their impact on the interpretation of elevation change across the island, and the potential implications for study of the coastal margins of the Greenland and Antarctic ice sheets with ICESat.