Cryosphere [C]

C53A   MCW:3001   Friday  1340h

Glacier and Ice Sheet Hydrology: Processes in Subglacial Environments II

Presiding: M Tranter, Bristol Glaciology Centre; A Fountain, Portland State University; M Studinger, Lamont-Doherty Earth Observatory

C53A-01 INVITED

A Complex Sub-Glacial Water System Beneath Whillans and Mercer Ice Streams Mapped Using ICESat

* Fricker, H A (hafricker@ucsd.edu) , Scripps Institution of Oceanography, 9500 Gilman Drive, La Jolla, CA 92093-0225, United States
Scambos, T (teds@icehouse.colorado.edu) , National Snow and Ice Data Center, CIRES, Campus Box 449; 1540 30th St., University of Colorado, Boulder, CO 80309-0449, United States
Bindschadler, R (Robert.A.Bindschadler@nasa.gov) , NASA Goddard Space Flight Center, Code 614 Bldg. 33, Rm. A112, Greenbelt, MD 20771, United States
Padman, L (padman@esr.org) , Earth & Space Research, 3350 SW Cascade Ave, Corvallis, OR 97333-1536, United States

Sequential elevation profiles from October 2003 to June 2006 collected by the ICESat laser altimeter over the Whillans and Mercer ice streams, West Antarctica, have revealed several (at least 10) localized elevation changes interpreted as subglacial water movement. Vertical motion and areal extent of some of the features are confirmed by differencing of MODIS satellite images over the period 2000-2005. The largest of these is a ~300 km$^{2}$ region which experienced a rapid 9 m decrease in surface elevation over the period February 2003 to November 2005. We interpret this as the surface expression of the draining of a previously unknown subglacial lake of at least 2.5 km$^{3}$ volume. Overall, the data indicate an active subglacial system in the two ice streams, which is highly variable on short timescales. It is likely that this water system exerts an important control on ice flow and mass balance.

C53A-02

Unsteady Glacier Flow Revealed by Multi-Source Satellite Data

* Magnï¿½sson, E (eyjolfm@raunvis.hi.is) , Institute of Meteorology and Geophysics, University of Innsbruck, Innrain 52, Innsbruck, A-6020 Austria
* Magnï¿½sson, E (eyjolfm@raunvis.hi.is) , Institute of Earth Sciences, University of Iceland, Sturlugata 7 ï¿½ Askja, Reykjavï¿½k, 101 Iceland
Rott, H (helmut.rott@uibk.ac.at) , Institute of Meteorology and Geophysics, University of Innsbruck, Innrain 52, Innsbruck, A-6020 Austria
Bjï¿½rnsson, H (hb@raunvis.hi.is) , Institute of Earth Sciences, University of Iceland, Sturlugata 7 ï¿½ Askja, Reykjavï¿½k, 101 Iceland
Roberts, M J (matthew@vedur.is) , Icelandic Meteorological Office, Bï¿½stadavegur 9, Reykjavï¿½k, 105 Iceland
Berthier, E (eberthier@eos.ubc.ca) , University of British Columbia, 6339 Stores Road, Vancouver, V6T1Z4 Canada
Pï¿½lsson, F (fp@raunvis.hi.is) , Institute of Earth Sciences, University of Iceland, Sturlugata 7 ï¿½ Askja, Reykjavï¿½k, 101 Iceland
Geirsson, H (dori@vedur.is) , Icelandic Meteorological Office, Bï¿½stadavegur 9, Reykjavï¿½k, 105 Iceland
Gudmundsson, S (sg@raunvis.hi.is) , Institute of Earth Sciences, University of Iceland, Sturlugata 7 ï¿½ Askja, Reykjavï¿½k, 101 Iceland
Bennett, R (rab@geo.arizona.edu) , Department of Geosciences, University of Arizona, Gould-Simpson Building #77, 1040 E 4th Street, Tucson, AZ 85721-0, United States
Sturkell, E (sturkell@hi.is) , Institute of Earth Sciences, University of Iceland, Sturlugata 7 ï¿½ Askja, Reykjavï¿½k, 101 Iceland

We use three types of data to study the flow dynamics of Skeidarï¿½rjï¿½kull, an outlet glacier of the Vatnajï¿½kull ice cap, Iceland. Firstly, by combining InSAR measurements from ascending and descending orbits and mass continuity, a three-dimensional flow field is derived for the glacier from late December 1995. By using the derived horizontal flow direction, and assuming mass continuity, we then derive the three-dimensional flow field over 24 hours, for single interferograms, for 23 periods between 1995 and 2000. These data are from the ERS1/2 tandem mission, obtained within the ERS AO projects VECTRA and AO3.239. Secondly, we derive the annual, horizontal flow field by cross-correlating optical satellite images acquired at the end of the ablation season (August or September) for the years 1999-2005. Images from SPOT5, ASTER and LANDSAT sensors were combined for that purpose. Thirdly, we present continuous GPS data-sets from three stations that were deployed on Skeidarï¿½rjï¿½kull in spring 2006 and will remain on the glacier until late 2006. The combined data-sets reveal consistent and stable ice-surface velocities; however, episodes of short-lived high velocity are apparent. Changes in basal water-pressure, caused by either water input from the surface (rainfall or intense melting), or drainage of subglacial or ice-marginal lakes seem to trigger these events. Comparison with the annual velocities derived from the optical imagery indicates that a significant part of the ice flux occurs during such speed-up events.

C53A-03

Melting and Water Distribution Beneath Rutford Ice Stream, West Antarctica from radar and seismic data.

* King, E C (ecki@bas.ac.uk) , British Antarctic Survey, High Cross, Madingley Road, Cambridge, CB3 0ET United Kingdom

I collected radar data in the onset region of Rutford Ice Stream where previous seismic surveys indicated free water at the bed in a possible canal system. The objectives were to use the radar to map the bed topography in detail and to determine if there was a correlation between seismic and radar reflectivity which would allow mapping of water distribution. I used the British Antarctic Survey's Delores system (Deep Look Radio Echo Sounder), which is a 1 MHz monopulse radar. The data show internal reflections through the full 2900 m thickness of the ice stream and a bed reflection with very good signal to noise ratio. There are large variations in the strength of the radar reflection from the bed, which I interpret as an indicator of water distribution. The highest reflection strengths are found on the down-stream flanks of bedrock highs. The peaks in reflection strength coincide with maxima in the hydraulic potential gradient. Thus the water distribution is controlled by the interaction between surface slope and bedrock slope, as expected from theory. Isochronal reflectors show fairly uniform vertical separations through most of the thickness of the ice stream. The exception is near the bed where the thickness of the lowermost ice layer varies between 200 and 600m. Some of this variation may be due to deformation of the soft ice at the bottom of the ice stream accommodating flow around and over bedrock obstructions, however it is likely that some of the variation is due to basal melting.

C53A-04

Flow-Switching Between Rutford Ice Stream and Carlson Inlet, West Antarctica

* Vaughan, D G (dgv@bas.ac.uk) , British Antarctic Survey, Madingley Road, Cambridge, CB3 0ET United Kingdom
Corr, H F (hfjc@bas.ac.uk) , British Antarctic Survey, Madingley Road, Cambridge, CB3 0ET United Kingdom
Smith, A M (amsm@bas.ac.uk) , British Antarctic Survey, Madingley Road, Cambridge, CB3 0ET United Kingdom
Pritchard, H D (hprit@bas.ac.uk) , British Antarctic Survey, Madingley Road, Cambridge, CB3 0ET United Kingdom

Despite efforts to find it, Antarctica has yet to yield evidence for rapid flow-switching in between ice streams, or rerouting of ice-flow over short periods. The Rutford Ice Stream (RIS) ï¿½ Carlson Inlet (CI) system in West Antarctica is a strong candidate for such behaviour. Currently, RIS is an active ice stream, while CI is stagnant, but several lines of evidence suggest that CI was recently active: its bed is composed of partially de-watered sediments which are sculpted along ice-flow directions, its basin has a positive mass balance and is inflating at a rate > 20 cm per year, and basal hydrologic calculations indicate that all subglacial water is currently diverted away from CI and beneath RIS. There is, however, no conclusive evidence that would allow us to date the switch- off of CI, although radar data suggest it was more than 320 years bp. A sensitivity test using a new high-resolution bed elevation map indicates that the sub-glacial hydrology is finely poised. Thickness changes of only 50 ï¿½ 100 m (2-4% of ice thickness) would be required to re-direct drainage beneath CI, and alter the balance between the ice streams. A 10% increase in flowrate of RIS could cause such a thickness change in just a few decades. The findings argue that subtle features in the subglacial hydraulic potential could dramatically alter large-scale ice- stream flow, and thus the evolution of the ice sheet.

C53A-05

Discharge of New Subglacial Lake on Whillians Ice Stream: Implication for Ice Stream Flow Dynamics.

* Sergienko, O V (olga@neptune.gsfc.nasa.gov) , NASA Goddard Space Flight Center, Code 614,NASA Goddard Space Flight Center, Greenbelt, MD 20770, United States
Fricker, H A (hafricker@ucsd.edu) , Institute for Geophysics & Planetary Physics, Scripps Institution of oceanography, 9500 Gliman Drive, La Jolla, San Diego, CA 92039, United States
Bindschadler, R A (Robert.A.Bindschadler@nasa.gov) , NASA Goddard Space Flight Center, Code 614,NASA Goddard Space Flight Center, Greenbelt, MD 20770, United States
Vornberger, P L (patricia@igloo.nasa.gov) , SAIC, 4600 Powder Mill Road, Beltstville, MD 20705
MacAyeal, D R (drm7@midway.uchicago.edu) , SAIC, 4600 Powder Mill Road, Beltstville, MD 20705
MacAyeal, D R (drm7@midway.uchicago.edu) , University of Chicago, 5734 S. Ellis Ave, Chicago, IL 60637

One of the surprise discoveries made possible by the ICESat laser altimeter mission of 2004-2006 is the presence of a large subglacial lake below the grounding zone of Whillians Ice Stream (dubbed here ï¿½Lake Helen' after the discoverer, Helen Fricker). What is even more surprising is the fact that this lake discharged a substantial portion of its volume during the ICESat mission, and changes in lake volume and surface elevation of the ice stream are documented in exquisite detail [Fricker et al., in press]. The presence and apparent dynamism of large subglacial lakes in the grounding zone of a major ice stream raises questions about their effects on ice-stream dynamics. Being liquid and movable, water modifies basal friction spatially and temporally. Melting due to shear heating and geothermal flux reduces basal traction, making the ice stream move fast. However, when water collects in a depression to form a lake, it potentially deprives the surrounding bed of lubricating water, and additionally makes the ice surface flat, thereby locally decreasing the ice stream driving stress. We study the effect of formation and discharge of a subglacial lake at the mouth of and ice stream using a two dimensional, vertically integrated, ice-stream model. The model is forced by the basal friction, ice thickness and surface elevation. The basal friction is obtained by inversion of the ice surface velocity, ice thickness and surface elevation come from observations. To simulate the lake formation we introduce zero basal friction and ï¿½inflateï¿½ the basal elevation of the ice stream at the site of the lake. Sensitivity studies of the response of the surrounding ice stream and ice shelf flow are performed to delineate the influence of near-grounding-line subglacial water storage for ice streams in general.

C53A-06

Floating Fraction of Basal Ice, a Possible Mechanism for Ice Streams.

* Hofstede, C M (coen.hofstede@maine.edu) , University of Maine, Department of Earth Sciences Climate Change Institute, Orono, ME 04473, United States

Increased melt on major ice sheets will lead to increasing sea level. The two major ice masses on Earth, Greenland and Antarctica, seem to react faster to global warming than originally was thought, by increased draining of ice into the sea. Most of the draining takes place through fast moving calving glaciers called ice streams. A not well understood condition for ice streams to occur is the presence of basal water. We present a possible mechanism for ice stream draining by introducing the floating fraction of basal ice in a flow line model. We will show that this fraction is really an uncoupling percentage of the ice mass from the bed. Two limiting situations occur in the value of the uncoupling percentage. In one case, the ice mass is completely uncoupled which will result in an ice shelf profile. In the second case, the ice mass is completely coupled which will result in a parabolic shaped ice sheet. As ice streams are transitional between sheet-flow and shelf-flow, the uncoupling percentage or the floating fraction has to have its value between the two limiting situations. We will first present a test case in which the floating fraction is known along a flow line, leading to an ice sheet profile. We will use the produced ice profile as input to see if we can reproduce our floating fraction. To see if our model makes any sense in the real world, the model will be tested with a field data set for Whillans Ice Stream. The results show us that the model is a good first order approximation but has shortcomings due to simplifications in ice rheology. The beauty of the model lies in the fact that it is simple and thus quick and that the floating fraction provides a gradual transition from ice sheet to shelf flow, representing an ice stream.

C53A-07

Subglacial water sheets: Depth and stability are compatible

* Creyts, T T (tcreyts@eos.ubc.ca) , Earth and Ocean Sciences, University of British Columbia, 6339 Stores Road, Vancouver, BC V6T 1Z4 Canada
Schoof, C (cschoof@eos.ubc.ca) , Earth and Ocean Sciences, University of British Columbia, 6339 Stores Road, Vancouver, BC V6T 1Z4 Canada
Clarke, G K (clarke@eos.ubc.ca) , Earth and Ocean Sciences, University of British Columbia, 6339 Stores Road, Vancouver, BC V6T 1Z4 Canada

There exists a broad spectrum of subglacial hydraulic configurations. These are typically categorized based on parameters such as cross-sectional area, along-flow morphology, and water discharge. Common examples of categories include cavities linked through small orifices and subglacial hydraulic sheets that are much wider than deep. Sheets have often been discounted in the glaciological literature because they decay to channels or other drainage configurations above a critical water depth ($\leq$ 4 mm) (Walder, 1982). We show that under reasonable conditions much thicker sheets can remain stable. We present a new formulation of water sheets based on insight gained from glacier sliding. In this case, we consider an ice ceiling that intrudes into a water sheet because viscous creep and regelation act normal to the ice--bed interface. We modify a stress renormalization technique originally formulated for the study of linked cavities. In our renormalization, obstacles, such as sediment grains of different sizes, support some of the weight of overlying ice with water pressure supporting the remainder. Obstacles are divided into size classes with each different class bearing a different percentage of the weight of the overlying ice. This formulation yields a number of simultaneous equations: one velocity equation for each size class, and one total stress balance. Intrusion velocities are dependent on the effective pressure, which is the ice overburden pressure less the water pressure. For zero effective pressure, no closure occurs. When effective pressure is high, results show that instantaneous closure velocities can range up to approximately $0.4$ m h$^{-1}$. However, these velocities are not sustained indefinitely. As ice intrudes farther into the sheet (i.e., water depth decreases), the ice rests on more obstacles, stress concentrations disperse, and velocities decrease. Stress solution distributions can vary considerably for each obstacle size class. Thus, a true controlling obstacle size is ill-defined and not necessarily justifiable. In addition, results suggest that water sheets can be stable to much greater depths because intrusion velocities are coupled to effective pressure. This conclusion results from the inclusion of both regelation and viscous creep into the governing equations. Creep acts rapidly and is the controlling mechanism for high effective pressures. Regelation is the controlling mechanism when effective pressures are low. For these low stress cases, sheets close slowly and are more stable than a simple viscous closure model suggests.