Cryosphere [C]

C42A MCC:2002 Thursday

The Dynamics of Glacier System Response: Tidewater Glaciers and the Ice Streams and Outlet Glaciers of Greenland and Antarctica II

Presiding: M Truffer, University of Alaska Fairbanks; T Murray, Department of Geography, University of Wales, Swansea

C42A-01 INVITED

Columbia Glacier, Alaska, at Mid-Retreat

* Pfeffer, W T (pfeffer@tintin.colorado.edu) , Institute of Arctic and Alpine Research, University of Colorado, 1560 30th St, Boulder, CO 80303 United States

Columbia Glacier, which terminates in a grounded calving margin in Prince William Sound in central coastal Alaska, has retreated 15 km in the last 25 years, and with annually-averaged speeds in excess of 25 meters per day, it is among the world's fastest glaciers. Discharge flux of ice to the ocean has now exceeded 7 cubic kilometers of ice per year, making it also the largest single glacial contributor to sea level in North America and a prime example of the potent influence of non-linear glacier response to climate forcing. It is a model for similar dynamic responses which may already be occurring in Greenland and Antarctica. Columbia Glacier has been carefully observed, principally by aerial photogrammetry, since the onset of its retreat in the early 1980s. This record is now being supplemented by terrestrial photogrammetry, optical surveying, seismic observations, and oceanographic observations. These observations document the evolving geometry, velocity and strain rate during the retreat and give insights into the iceberg calving process, changing force balance during retreat, and the processes responsible for the apparent irreversibility of tidewater retreat. The retreat is now at its midpoint, with another 15 km of retreat expected before the terminus reaches shallow water. As the glacier terminus passes the principal constriction in its channel and the deepest water yet encountered (in excess of 500 m), Columbia Glacier may now undergo further changes unlike those observed up to this time. This talk provides an overview of observations of Columbia Glacier made up to this time, a summary of present understanding of tidewater retreat, a discussion of the similarities and differences between tidewater retreat and ice sheet outlet glacier dynamics, and informed speculation on the near future for Columbia Glacier.

C42A-02

Post Little Ice Age Collapse of the Glacier Bay Icefield, Alaska

* Motyka, R J (jfrjm@uas.alaska.edu) , Geophysical Institute, University of Alaska, PO Box 757320, Fairbanks, AK 99775 United States
Larsen, C F (chris@gi.alaska.edu) , Geophysical Institute, University of Alaska, PO Box 757320, Fairbanks, AK 99775 United States

Glacier Bay provides an excellent example of the tidewater glacier cycle proposed by Austin Post. It has a complete record of an advancing phase, stability, rapid calving and drawdown, lengthy retreat, and then readvance behind protective sediments. Glacier Bay currently consists of numerous discrete glaciers and small isolated icefields, but it recently contained a huge continuous icefield up to 2 km thick that covered more than 6000 km$^{2}$ at the peak of the Little Ice Age (LIA) (1750 AD). Rapid calving and associated upstream drawdown lead to its collapse. In less than 160 yrs, the main trunk of the icefield retreated 120 km in fjords as deep as 500 m. We evaluated the LIA volume and topography of the Glacier Bay Icefield based on mapping of trimlines, lateral moraines and terminal moraines. We used light aircraft to identify these geomorphic markers as well as analysis of vertical airphotos, hydrographs, seismic profiles, and the SRTM digital elevation model. Our reconstruction indicates an ice volume loss of over 3000 km$^{3}$ during the post-LIA collapse. This localized ice wastage represents the largest post-LIA deglaciation known to us, and is greater than the volume lost from all Alaskan and neighboring Canadian Glaciers from 1955-2002, greater in volume than the Larsen B 2002 ice shelf collapse, comparable in volume to Lake Huron, and equivalent to a global rise in sea level (SLE) of 8 mm. The collapse of the Glacier Bay icefield stranded many tributary glaciers. Some were entirely isolated from any source of accumulation and are now simply wasting away (e.g., Burroughs Glacier). Other glaciers in the region have had their accumulations areas severely reduced as the icefields feeding the LIA tidewater glaciers disappeared (e.g., Casement and Brady Glaciers). The vast loss of ice has lead to some of the highest rates of glacier rebound presently occurring in the world (32 mm/yr) with total uplift since the 18th century of as much as 5.8 m. Facilitated by infill of fjords from erosion and remobilization of subglacial sediments, several glaciers are now in the advancing phase of the tidewater glacier cycle despite the regional trend of glacier wastage. Post-LIA Glacier Bay could provide an analogue to collapse of other tidewater glacier systems and outlet glaciers from polar ice sheets.

C42A-03

Repeated Rapid Retreats of Bering Glacier by Disarticulation - The Cyclic Dynamic Response of an Alaskan Glacier System

* Molnia, B F (bmolnia@usgs.gov) , U.S. Geological Survey, 12201 Sunrise Valley Drive MS 926A, Reston, VA 20192 United States

Bering Glacier is the largest glacier in continental North America, with an area greater than 5,000 square kilometers. Alone, it accounts for more than 6% of the glacier-covered area of Alaska and perhaps 15% of Alaska's glacier ice. In places, its bed is more than 250 m below sea level. It is also the largest surging temperate glacier on Earth. Surges, some with maximum ice displacements exceeding 13 km, occurred at least five times during the 20th century. Analysis of aerial photography, dating from 1936, and satellite imagery, dating from 1972, documents that following each of the last four surges, Bering Glacier experienced post-surge, decadal-scale, cyclic episodes of rapid retreat. In each instance, the primary mechanism responsible for the rapid retreat was a buoyancy-driven process, here named 'disarticulation.' Abundant imagery exists to carefully document the post-surge retreat pattern during the two most recent cycles, 1967-1992 and 1996-present. In each cycle retreat began by calving. However, within a few years, as the piedmont lobe rapidly thinned, often by more than 20 m/yr, the dominant process transitioned to disarticulation, and the rate of retreat greatly increased. Disarticulation events occur when the thinning, low-gradient, distal end of the glacier's piedmont lobe reaches a state of buoyancy and separates from its bed. As it begins to float, large tabular pieces of ice up to a kilometer in maximum dimension passively separate from the terminus. Separation usually occurs along old crevasse and fracture planes and may begin at distances of more than 2 km behind the terminus. Often, hundreds of large icebergs simultaneously separate. Disarticulation events were also identified as being underway on photographs from 1936 and 1948. These followed surge that ended in the 1920s and 1940. In both 1936 and 1948, disarticulation was occurring at the same location. This location was also a focal point for disarticulation during the last two post-surge cycles. The post-1967 surge cycle spanned 25 years and resulted in a maximum of 10.7 km of terminus recession. Maximum annual recession exceeded 2.5 km. In the post-1996 cycle, maximum retreat is more than 6 km. Disarticulation is not unique to Bering Glacier, and probably not unique to Alaska. A 2005 aerial survey of Alaskan coastal glaciers identified more than a dozen glaciers, many former tidewater and calving glaciers, including Grand Plateau, Alsek, Bear, and Excelsior Glaciers that were rapidly retreating through disarticulation.

C42A-04

Exploring similarities between tidewater and ice sheet outlet glaciers

* Truffer, M (truffer@gi.alaska.edu) , Geophysical Institute, 903 Koyukuk Dr, Fairbanks, AK 99775-7320 United States
Motyka, R (roman.motyka@uas.alaska.edu) , Geophysical Institute, 903 Koyukuk Dr, Fairbanks, AK 99775-7320 United States
Echelmeyer, K (kechel@gi.alaska.edu) , Geophysical Institute, 903 Koyukuk Dr, Fairbanks, AK 99775-7320 United States

Tidewater and outlet glaciers pose a serious challenge to glaciologists because of their complicated behavior. Yet, they exhibit some of the largest changes observed in glacial systems, often defying regional trends of nearby land-terminating ice masses. For example the Chugach Range of Alaska, while generally losing ice mass, contains one growing tidewater glacier. Also, about 50% of the current mass loss there is due to the rapidly retreating Columbia Glacier. In Glacier Bay, several tidewater glaciers are advancing while land terminating and lake calving glaciers are rapidly wasting away. We will explore some of the common features found between temperate tidewater glaciers and the cold or polythermal outlet glaciers of the big ice sheets. These commonalities include patterns of ice draw-down, particularly far upstream, and acceleration of ice flow during the retreat phase that propagates much farther upstream than suggested by effects of longitudinal coupling. Rapid retreats are initiated at the glacier terminus and we suggest that they are linked to warming ocean temperatures and glacial freshwater runoff. We will also outline the modeling challenges for the upstream propagation of a drawdown event. These require a full treatment of the three dimensional Stokes equations. However, it is not clear how to treat the basal boundary condition and the possibility of temporal changes at the ice base. A particular challenge for outlet glacier systems is posed by free surfaces, such as the grounding line or the cold-temperate-surface (CTS). One striking feature of tidewater glaciers is their evolution through a tidewater glacier cycle of rapid retreat and slow advance as proposed by Austin Post. This cycle is well documented for Alaska's tidewater systems, and we suggest that the possibility of such a cycle should be explored for ice sheet outlet glaciers. The often asynchronous behavior of these systems poses a challenge for the assessment of regional or global ice volume changes. Similarly, ice sheet mass balance can be dominated by a few rapidly changing outlet systems. Such systems must be carefully examined for their disproportionate effects on volume change and sea level studies.

C42A-05

Glacier Flow in a Rapidly Changing Climate: the Antarctic Peninsula West Coast

* Pritchard, H (H.Pritchard@bas.ac.uk) , British Antarctic Survey, High Cross, Madingley Road, Cambridge, CB3 0ET United Kingdom
Vaughan, D G (D.G.Vaughan@bas.ac.uk) , British Antarctic Survey, High Cross, Madingley Road, Cambridge, CB3 0ET United Kingdom

The most dynamic Antarctic glacial regime is that of the mountainous Antarctic Peninsula that covers only 3% of the continental area but receives 18% of its snowfall. Over the last few decades, the peninsula has been the fastest warming region on Earth and this is linked to the virtual loss of seven ice shelves, retreat of 87% of its tidewater glacier fronts, a 74% increase in annual positive degree-days (PDDs) and a consequent rise in summer meltwater runoff. In addition, precipitation has increased by 20%. All of these factors can be expected to increase the flux of ice through the glacial system, and possibly to drive it out of balance. The flow of most glaciers on the peninsula has, however, never been measured and very little is known about the dynamic response to these forcings. We present repeated radar remote sensing measurements of flow on 337 glaciers on the west coast of the Antarctic Peninsula through 4 summers and 75 glaciers through 8 summers between 1992 and 2005. We find interannual changes in flow rate that are both widespread and significant. This demonstrates the high sensitivity of the peninsula glacial regime to its changing environment and allows us to requantify the region's contribution to sea level rise.

C42A-06

Century-Scale Ice Stream Variability and Stability: A case study of Kamb Ice Stream, West Antarctica.

* Catania, G A (gcatania@ig.utexas.edu) , Institute for Geophysics, 4412 Spicewood Springs Rd. Unit 600 University of Texas, Austin, TX 78759 United States
Scambos, T (teds@icehouse.colorado.edu) , National Snow and Ice Data Center, CIRES, University of Colorado, Boulder, CO 80309 United States
Conway, H (conway@ess.washington.edu) , Dept. of Earth and Space Sciences, University of Washington Box 351310, Seattle, WA 95195 United States
Raymond, C (charlie@ess.washington.edu) , Dept. of Earth and Space Sciences, University of Washington Box 351310, Seattle, WA 95195 United States

Observational and model studies are synthesized to obtain a 700 year ice flow history for the Kamb Ice Stream (KIS) region of West Antarctica which is used to infer the nature and pattern of the KIS shutdown. This ice flow history reveals a dominance of short-term (order of 100 years) variability in ice stream position and discharge that is linked to changes in subglacial conditions and ice thickness. Evidence suggests that the trunk of KIS was wide, thin and temporarily ungrounded between ~550-350 years ago [Catania et al., {\it Journal of Glaciol.}, in press]. Since grounded conditions currently exist throughout the trunk region we suggest that rapid changes in basal conditions were possible during the last few hundred years prior to shutdown. Simultaneous with re-grounding in the trunk region was a narrowing of the ice stream width. Such large events in the history of the ice stream may be linked to the sudden loss of fully lubricated basal conditions during the transition from ungrounded to grounded conditions and may have initiated the eventual shutdown in the ice stream trunk some 200 years later. We also observe that the short-term variability of KIS appears to be in part, controlled by neighboring Whillans Ice Stream (WIS). Examples of this include; 1) grounding throughout the trunk likely occurring because of a prior diversion in flow direction of WIS to the north, 2) migration of the northern WIS margin that was coincident with the KIS shutdown and 3) a switch in flow direction of a tributary into KIS (now flowing into WIS) [Conway et al., {\it Nature}, {\bf 419 (6906)}, 465-467, 2002]. Such interdependence suggests that conclusions regarding the behavior of an individual ice stream cannot be examined in isolation. In the context of long-term (1000 years) ice stream history, the observed short-term variability of one particular ice stream may be seen as "noise" within a system that is in constant flux but one that may maintain a stable mass balance over large temporal and spatial scales.

C42A-07

Classification of Surface Structures of Jakobshavns Isbrae Through Time and Thoughts on Recent Changes

* Williams, S , Cooperative Institute for Research in Environmental Sciences, University of Colorado Boulder, Boulder, CO 80309-0449 United States
Herzfeld, U C (herzfeld@iceberg.colorado.edu) , Cooperative Institute for Research in Environmental Sciences, University of Colorado Boulder, Boulder, CO 80309-0449 United States
Mayer, H (mayerh@mechanik.tu-darmstadt.de) , Institut fuer Mechanik III, Technische Universitaet Darmstadt, Hochschulstr. 1, Darmstadt, 64289 Germany

Jakobshavns Isbrae, arguably Earth's continuously fastest-moving glacier and considered by many as the prototype of an autonomous dynamical system, has exhibited signs of rapid and accelerating retreat, starting in 1999. We combine and synoptically analyze data from our field expeditions, satellite and aerial observations to investigate this phenomenon. The mathematical focus of this presentation is the application of a newly developed tool for automated geostatistical classification of surface structures. The glaciologic objective is to describe, map and quantify regions and types of dynamic changes and ultimately to contribute to our understanding of changes in the outlet glaciers of the Greenland Ice Sheet and their possible relationship to climatic warming.