C53A-01 INVITED
Seasonal Ice Acceleration Near the Greenland Ice Sheet Margin
Seasonal changes in ice velocity, depending upon their extent both spatially and temporally, may be important considerations when assessing the stability of inland glacier ice. Work by Zwally and others (2002) suggested potential ice sheet instability in Greenland due to the penetration of summer melt water from the surface of the ice sheet to the bed, where it promotes basal motion. The implication is that this process could lead to rapid disintegration of the Greenland ice sheet as surface melt becomes more common in a warming climate. The recent IPCC 4th Assessment Report points out that such dynamic responses are currently not used in projections of sea level rise; consequently, current estimates of future sea level rise may be too low. We present recent field data collected from the vicinity of Swiss Camp, a site near the equilibrium line on the west coast of Greenland. These data show short duration, episodic changes in velocity in both ablation and accumulation areas. The accelerations are synchronous over the scale of a few kilometers, but are not correlated over larger areas. The duration of these events are significantly shorter than those reported by Zwally et al. (2002), lasting at most two weeks; most are shorter. Events are consistent with localized, periodic melt water input to the basal water system. We explore the implications of these observations in terms of the subglacial hydrologic system near the ice sheet margin and ice sheet stability in a warming climate.
C53A-02
Direct Observations of Melt-Water Lake Drainage and the Establishment of an Efficient Surface to Basal Water Connection on the Greenland Ice Sheet
Melt water lakes are recurrent features on the surface of the Greenland Ice Sheet margin that collect a large fraction of the annual surface melt across the ablation region. Many of these lakes fill and drain seasonally and are hypothesized to be a significant source of surface melt water to the ice sheet bed. We present results from field campaigns during the summers of 2006 and 2007 to investigate the filling and draining of two lakes, and the dynamic response of the ice sheet to drainage events. Measurements include air temperature, lake-water level, seismicity and local ice motion. One of the instrumented lakes was observed to be actively discharging water through a meltwater-cut channel in the side of the lake basin, which followed a deeply incised (5-10 m) supraglacial stream for nearly a kilometer before cascading into a moulin. The second instrumented lake drained catastrophically through a series of fractures and moulins that opened beneath the lake and that were subsequently mapped in the field following drainage. At this site, the 2.7-km-diameter lake, holding on the order of 0.03 km3 of water, drained entirely through 1 km of ice thickness in less than 2 hours. The peak rate of water flow during this event exceeds the average flow over Niagara Falls. This drainage event coincided with increased seismicity as well as rapid glacier uplift (1.2 m) and horizontal acceleration to nearly 8 km/yr as measured on the ice surface near the lake shoreline. Subsequent subsidence and deceleration of the ice sheet occurred over the following 24 hours. These observations provide evidence for the injection of surface melt water directly to the ice sheet bed, and also indicate the presence of an efficient basal drainage system that can quickly disperse large inputs of surface melt water.
C53A-03
InSAR and GPS Observations Show Seasonal Speedup of Ice Flow in Greenland Following the Onset of Summer Melting
We have assembled a comprehensive set of InSAR and GPS observations that reveal both spatial and temporal changes in velocity during the summer melt season along a several-hundred kilometer stretch of the ice-sheet margin near Jakobshavn Isbrae, Greenland. In the bare ice zone, we obtain InSAR (speckle/feature tracking) results throughout the melt season that agree well with results from two continuous GPS stations located at roughly 1000 meters elevation. Over much of the slow-moving (100 m/yr) bare-ice zone, the InSAR data show summer speedups of 50-to-100 m/yr averaged over 24 days. We also detect seasonal speedups of similar magnitude on Jakobshavn Isbrae and several smaller fast moving (> 1 km/yr) outlet glaciers. In relative terms, however, the outlet glaciers speedups represent increases of less than 10 % relative to their annual means. Thus, proportionately the slow-moving inland ice is far more sensitive to seasonal speedup than are the rapidly flowing outlet glaciers, making it unlikely that recently reported large (> 1 km/yr) speedups on Jakobshavn and other outlet glaciers can be directly attributed to enhanced basal lubrication from increased surface melt. Similarly, the GPS data also reveal a period of generally enhanced flow extending through the melt season, punctuated by shorter-term speedups lasting a few days. These shorter-term accelerations correlate well with periods of increased surface melt that we inferred from positive-degree-day values measured at the GPS sites. In addition, the short-term accelerations coincide well with GPS-measured periods of peak uplift rates of the ice-sheet surface. The strong correlation of seasonal velocity with melt and uplift rates suggests that surface melt makes its way to bed rapidly, providing enhanced lubrication to regions of the ice sheet extending up to at least 1000 meters elevation. Furthermore, the spatially uniform nature of the speedup in the upper bare-ice zone, where a sparse distribution of moulins delivers water to the bed, suggests the presence of a well distributed sub-glacial drainage network.
C53A-04
Kinematic Constraints on Greenland Contribution to Sea Level Rise in the Next Century
Present-day contributions to sea level rise from Greenland, Antarctica, and all other glaciers and ice caps, are moderately well known, but prediction of future sea level rise is complicated by lack of knowledge of future ice- dynamic response from marine-based outlet glaciers and ice sheets. Much attention has been focused on Greenland's potential future sea level contribution, because it is more vulnerable to warming and warming-related effects than Antarctica and contains a greater total possible sea level equivalent than the category of glaciers and ice caps. Warming-induced sea level contributions from Greenland of 2 to 6 m have been postulated but no time scales have been attached. We present a simple kinematic analysis aimed at discovering what constraints may be imposed on Greenland ice discharge by fundamental glacier mechanics, and use this analysis to place limits on Greenland's contribution to sea level over the next century, the longest time scale over which we believe quantitative predictions of ice dynamics can reasonably be made. We find that sea level contributions from Greenland of 2 m or more during the next century are highly implausible, based on very simple kinematic constraints, including Greenland's subglacial topography.
C53A-05
Evidence for organized subglacial meltwater drainage beneath the Pine Island and Marguerite paleo-ice streams, Antarctica following the Last Glacial Maximum
Swath bathymetry records from the onset areas of two paleo-ice streams, the Pine Island and Marguerite ice streams, show geomorphic features suggestive of organized meltwater drainage systems. Radial, anastomosing and straight channels that link basins that are in excess of 800 meters deep characterize these systems. There is evidence that these drainage systems were active at the end of the Last Glacial Maximum, in the form of a well sorted silt unit, interpreted as a meltwater deposit, resting above till. This Radiocarbon ages from Marguerite Bay indicate that the ice sheet retreated rapidly from the bay about 9,000 years ago. Since that time the bay has remained ice-free, as evidenced by a surface unit of diatomaceous mud. In Pine Island Bay, the same stratigraphic succession occurs, but the meltwater deposits are the youngest deposits in the inner bay and a single radiocarbon date indicates a modern age for these deposits. These observations suggest that subglacial meltwater may have played a role in the retreat of these ice streams. In the case of Pine Island Bay, meltwater discharge occurred in modern times.
C53A-06 INVITED
More About R Channels
Much of our understanding of how subglacial drainage systems operate is based on key papers by Röthlisberger and Shreve that describe the steady-state hydraulics of circular ice-walled water conduits. Subsequent studies have extended these results to include transient conditions---relevant to diurnally- and seasonally-forced drainage systems---and to semi-circular bed-floored channels. In passing from circular to semi-circular conduits it is usual to adopt Nye's creep closure expression for circular tunnels and to assume that creep closure proceeds as it would for a circular ice-walled conduit. This assumption is only valid if ice can slip freely over the subglacial bed and it completely breaks down if free sliding is inhibited. Even for glaciers and ice streams that are sliding at fast rates the frictional resistance to conduit closure can be substantial and it is therefore impossible to maintain a semi-circular channel in a steady-state. The resulting difference between the rate of creep closure at the conduit ceiling and at the sidewalls forces the cross-sectional geometry of the conduit to evolve. At the time of writing it is unclear whether or not non-circular steady-state geometries can exist but either possibility is interesting. If there is no steady-state geometry then subglacial conduits must necessarily operate in an episodic manner. If steady-state geometries do exist then steady-state subglacial conduits must manifest along-path variations in their height-to-width aspect ratio in response to along-path changes in effective pressure and viscous dissipation.
C53A-07
Subglacial water distribution and transfer in Antarctica mapped with ICESat and image differencing
We have used repeat-track satellite laser altimeter (ICESat) data acquired between 2003-2007 over several drainage basins in Antarctica to map the locations of pooled subglacial water and monitor the transfer of water between reservoirs. The altimeter results are supported by satellite image differencing using contemporaneous MODIS imagery. We have made estimates of the temporal variation in water volumes in the lakes using two independent methods and examined linkages between lakes within the same hydrologic regime. We also consider the source of the water and how it affects local ice stream dynamics. Mapping subglacial water distribution and movement in Antarctica is important for improving models of the Antarctic ice sheet system. Monitoring subglacial outflows from the ice sheet margins is also important for quantifying freshwater flux to the ocean and understanding ice-ocean interactions.
C53A-08
Subglacial water transport throughout Antarctica from ICESAT laser altimetry
A survey of the Antarctic ice sheet using satellite laser altimetry has detected 46 small regions of surface uplift or drawdown in twelve different glacier drainages around Antarctica. Surface displacements are measured relative to the best-fitting plane passing through multiple (5-11) elevation measurements on the same repeat-track, allowing correction for across-track slopes. Volume displacements are derived by interpolating displacements from multiple tracks to a common grid. These ECAs (Elevation Change Anomalies) range from less than four km to more than 60 km across, with vertical displacements ranging from a few decimeters to over ten meters. Typical volume displacements are on the order of 0.05 cubic kilometers over the three-year survey, and the largest displacement is more than 1.4 cubic kilometers. Although the majority of the ECAs are within the Filchner- Ronne catchment, others (including those discussed by Fricker and others, 2007), are found in the Ross Embayment, in the drainages of Byrd Glacier and Lambert Glaciera, and in the interior of Wilkes Land. As have other researchers who have observed ECAs, we take these features to result from water motion at the bed. In all cases where the ice sheet velocity structure is known, the ECAs are in regions of ice stream or tributary flow, which implies that they are associated with melting bed conditions. Some of the ECAs appear to be downstream of linear features in the ice sheet surface, suggesting that they are associated with local minima in the hydraulic potential at the bed. Others have no clear association with surface topography. The relatively small number of ECAs precludes drawing strong conclusions about spatial and temporal correlations between filling and drainage events. However, a few conclusions are clear: Because adjacent ECAs are more likely to have correlated filling or drainage rates than to have anticorrelated filling or drainage rates, it does not appear that water is conserved among the ECAs. This suggests that the ECAs exchange water with other water systems at the bed. Of critical importance for our understanding of ice stream dynamics is whether inflation of the ECAs represents a withdrawal of water from the system that lubricates fast basal motion. If the correlation of drainage and filling over distances of hundreds of kilometers is not coincidental, it suggests either that large-scale channel systems connect the ECAs, or that filling or drainage is triggered by large-scale velocity variations.