C41A-0037
Calving of Talyor Glacier, Dry Valleys, Antarctica
Calving of tide-water glaciers has received considerable attention, with seismic arrays in Alaska, Greenland, and Antarctica devoted to their observation. In these environments, ice cliffs are directly coupled to oceanic temperatures. The land-based polar glaciers of the McMurdo Dry Valleys in Antarctica represent a simpler environment unaffected directly by water contact where other factors can be isolated. In particular, summer calving events of Taylor Glacier are observed to consist of precursory activity including crack growth, cliff overhang, and active seismicity at least 1 hour before collapse. We propose that collapse occurs only after a stress threshold has been crossed, evident from ‘pre-calving' of ice from the cliff base 1-3 days prior to the major event. We provide photographic, seismic, and temperature data to illustrate the thermal and stress landscape for land-based calving of polar glaciers. http://earthweb.ess.washington.edu/~joshuadc/
C41A-0038
Calving Theory and the Thinning, Retreat, and Disarticulation of Bear Glacier, Alaska
Bear Glacier, with an area of ~225 km2, is a 25-km long valley glacier located in the southern Kenai Mountains. When first mapped in 1909, it had a piedmont lobe with an area of >30 km2. By the mid- 1980s, the terminus had retreated from 1-3 km, thinned by >150 m, and was actively calving small icebergs into an ice-marginal lake. Through the end of the 20th century, the piedmont lobe continued to thin and narrow. Active calving decreased and then ceased as the thickness of the thinning glacier approached Tnb, the thickness of neutral buoyancy (the floatation thickness). As thinning continued, much of the terminus began to float. Passive calving, characterized by the release of large tabular icebergs from Bear's low gradient, low elevation terminus, became the dominant retreat process. Between 2000 and 2007, part of the terminus retreated about 3.6 km. Now, 0.5 km-long icebergs frequently separate from Bear's western terminus. An August 2006 bathymetric survey identified many locations with depths >75 m in Bear Glacier's ice-marginal lake. Calving theory can be used to explain much of Bear's observed behavior. Current calving theory suggests that active calving is initiated when a glacier terminus thins to a critical thickness, (i.e. Tnb plus an additional but limited thickness of ice such as 50 m [Tnb+ 50 m]). The thinning of Bear Glacier during much of the 20th century resulted in it approaching this thickness. In a freshwater lake with a depth of ~75 m, active calving would be initiated when Tnb + 50 m was about 120 m, continuing as the glacier thinned to <100 m. With further thinning, Bear Glacier would transition to a regime dominated by passive calving (disarticulation). This condition was initiated sometime after the mid-1990s, when Bear Glacier became significantly thinner than Tnb. Glacier ice floats because its density is at least 8% less than that of freshwater or saltwater. As Bear Glacier thinned to Tnb (approximately 85 m), active calving decreased and parts of the terminus began to float. Concurrently, lake water began to appear in crevasses and at low points on the glacier's surface. By 2002, most of Bear's terminus was afloat. With little or no piedmont lobe gradient and virtually no terminus margin height, active calving ceased and large icebergs began to be produced through passive calving (disarticulation), generally separating along former crevasse scars and fractures. These are the condition that resulted in the catastrophic retreat that characterizes Bear Glacier's 21st century behavior.
C41A-0039
Glacier Dominated Shelf Stratigraphy Interpreted From High-Resolution Seismic Reflection Profiles Within the Yakutat Sea Valley Region, Gulf of Alaska
The detailed analysis of the shelf stratigraphy along the northern Gulf of Alaska margin has the potential to reveal critical insights into past glacial and climate conditions in this modern temperate glacimarine environment. The shelf stratigraphy in and around the Yakutat Sea Valley is of interest as it can provide insights into the interaction between the Malaspina Glacier and glaciers within Disenchantment Bay, including Hubbard Glacier, during times of glacial expansion along the shelf. The interpretation of seismic facies, depositional environments, submarine landforms, and glacial sequence stratigraphy from the analysis of high-resolution seismic reflection profiles and multibeam bathymetry can be used to generate such insights. Here we put forward a hypothesis that the Malaspina Glacier may be more dominant than Hubbard Glacier in the erosion and deposition of sediment within Yakutat Bay and along the Yakutat Sea Valley during times of ice expansion. This hypothesis is supported by the orientation and distribution of submarine landforms within Yakutat Bay and the seismic stratigraphy of glacial influenced strata in and around the Yakutat Sea Valley. The orientation of approximately 215 meters of seismic strata infilling the sea valley indicates the Malaspina Glacier may be the most significant source of sediment input along the central portion of the Yakutat Sea Valley since the LGM. These strata contain multiple distally thinning wedges of seismically chaotic sediment interbedded within a continuously stratified seismic facies. The wedges of chaotic facies are interpreted to represent sediment deposited at a proximal grounding line position from a source along the northern or northwestern edge of the Yakutat Sea Valley. Sequence stratigraphic analysis of glacimarine deposits adjacent to and below the Yakutat Sea Valley is preliminary and is hindered by abrupt seismic facies changes and ambiguity when defining distinct erosional surfaces from reflection termination patterns. Nonetheless, differences in reflection amplitude and continuity of particular reflections indicate that multiple sequences can be identified and that glacimarine sequences are preserved below the Yakutat Sea- Valley. This preliminary interpretation indicates that there is a potential to constrain the timing of the development of the sea valley, which will provide greater insight into the depositional, glacial, tectonic, and climate history along this part of the Gulf of Alaska shelf.
C41A-0040
Investigation Into the Origin of Submarine Channelized Deposits Found on the Continental Shelf of Southern Alaska
The interaction between tectonic exhumation and climatic events along the southern coast of Alaska provides a unique setting in which the interplay of these significant processes can be evaluated. High precipitation rates and topographic relief in conjunction with the cool, temperate nature of the Alaskan coast produce large-scale glaciation across the margin. As such, glacial erosion processes control much of the sediment yield to the continental shelf during advance and retreat stages that have persisted throughout the last ~2.5 myr. Numerous large-scale channel structures preserved within continental shelf sediments may serve as markers of glacial advances or retreats and may illustrate significant glacial-marine processes. We have undertaken an integrated study into the origin of these structures using 1979 industry seismic data provided courtesy of the USGS and NSF-funded, high-resolution seismic data. Channel geometries range in scale from 513 m wide by 60 m deep, up to 4700 m wide and 355 m deep. Differentiation of the channels from the surrounding sediment is possible due to the hummocky, unstratified seismic character of the deposits filling the channels. Bright reflectors in the seismic data, likely representing glacial retreat surfaces, tend to terminate at channel margins. Packages of stacked or overlapping channels are common, suggesting reactivation of the channel surfaces. Although the exact nature of channel formation is still ambiguous, their considerable size and abundance speaks to their potentially significant role in piecing together a geologic history of the margin. Preliminary results suggest that glacially driven physical processes during advance and/or retreat are responsible for their formation. Interpretation of the seismic data will provide a means to determine if the channels are related to sub-ice movement of water or ice proximal erosional processes. Concurrently, with correct interpretation of channels and major erosional horizons, age associations amongst the channels are possible. In using the channels and major erosional horizons as markers in time and space, our goal will ultimately be to gain a better understanding of the degree of influence that the numerous glacial cycles, relative to one another, have had on the delivery and accumulation of sediment along the margin.
C41A-0041
A 700 year-long Record of Glacial Surging and Associate Flooding: Bering Glacier, Alaska.
Bering Glacier, Alaska is one of the largest, but until recently, most poorly studied dynamic glaciers in North America, and is largely known for its dramatic surging events, six of which have occurred in historic times. A primary late Holocene history of the Bering has been established from on-shore studies of glacial termini position and evidence of glacial advances, but the Little Ice Age (LIA) record of glacial surging and associated flooding has not been examined. A 14m-long jumbo core collected on the adjacent continental shelf reveals a 700-year-long record of flood deposition. The core was dated using 210Pb chronology and five radiocarbon dates and can be separated into three distinct lithologic units based on examination of x-radiographs and physical properties: 1) the uppermost unit dates from /sim120 yr BP to the present and is characterized by bioturbated mud interbedded with faintly laminated, thick (5-20cm) low-density beds, 2) the middle unit dates from /sim120 /- 300 yr BP and includes abundant laminated-to-interbedded low- and high-density beds with some evidence of bioturbation, and 3) a lowermost unit post dates 300 yr BP and is composed of rare laminated beds grading into mottled to massive mud. In each of these units, the laminated lithofacies from this oxic mid- shelf location indicates both flood and gravity flow deposition. Based on terrestrial studies (Wiles et al, 1999), from 850 /- 400 yr BP, the terminus was at a slightly advanced position relative to the present and was at its Neoglacial maximum extent, which was close to the modern coastline, during the LIA (400 /- /sim200 yr BP). The thick low-density, clay-rich beds in the uppermost unit correlate with historic outburst floods associated with known surge events. During the LIA, bioturbated intervals are rare and thin while laminated intervals are common. Given average radiocarbon-established sedimentation rates of 2-3 cm/y, this style of interbedding indicates frequent flood deposition at the core's location. This would suggest that the decadal-long quiescent periods between outburst floods typical of the last century were not prevalent during the LIA. The infrequent deposition of flood layers in the lowermost unit could be attributed to the diversion of glacial drainage to the eastern Kaliakh, Tsiu, and Tsivat Rivers (Muller and Fleisher 1995) instead of present day Seal River. The observation of more frequent flooding events during the LIA differs appreciably from the past century, suggesting that Bering Glacier's hydrology during this period was altered relative to the present conditions that result in quasi-periodic surges.
C41A-0042
Remote Sensing Applications for the Bering Glacier, Alaska
Satellite remote sensing is an invaluable tool to monitor and characterize the Bering Glacier System. The Bering Glacier is located in coastal, south-central Alaska and is the largest and longest glacier in continental North America, with an area of approximately 5,175 km2, and a length of 190 km. It is also the largest surging glacier in America, having surged at least five times during the twentieth century. The last great surge occurred in 1993- 1995. Bering Glacier alone covers more than 6 percent of the glacier covered area of Alaska and may contain 15- 20 percent of Alaska's total glacier ice. Applications of glacier remote sensing include but are not limited to: mapping extent and features, ice velocities through sequential observations, glacier terminus locations, snow line location, glacier albedo, changes in glacier volume, iceberg surveys and calving rates, hydrographic and water quality parameters in ice marginal lakes, and land cover classification maps. Historical remote sensing images provide a much needed geospatial time record of the dynamic changes Bering Glacier has undergone, including changes due to its surge behavior and response to climate change. Remote sensing images dating back to the early 1990s have been used to map the glacier terminus retreat of approximately five to seven kilometers which has resulted in Vitus Lake increasing in volume approximately 260 percent since 1995 to the current (2006) volume of 9.4 km3 of water. Using elevation data obtained from remote sensing and GPS surface points, we have determined that the glacier elevation has decreased approximately 150 m in elevation at the terminus and 30 m at a position 300 m below the present (2006) equilibrium line (~1,300 m) since 1972. Satellite observations have recorded the upward migration in altitude of the equilibrium line to its present position (slightly > 1,200 m). The decrease in glacier volume, obtained using remote sensing derived elevation data, from 1957 to 2004 is estimated at approximately 104 km3. Remote sensing data has also mapped the sediment rich (rock flower) water flowing into Vitus Lake providing insight into the hydrologic circulation of the Bering Glacier system, showing major glacier discharge from the Abandoned River, Arrowhead Point, and Lamire Bay in the area of Vitus Lake west of Taggland.
C41A-0043
Hubbard Glacier Update: Another Closure of Russell Fjord in the Making?
Hubbard Glacier is located near the community of Yakutat in southeastern Alaska. It is the largest non-polar temperate tidewater glacier in the world and has been advancing since 1890 AD, currently at a rate of 35 m a-1. Hubbard Glacier has twice closed off Russell Fjord creating enormous glacier dam lakes, once in 1986 and again 2002. Both dams failed catastrophically producing two of the largest outburst floods in historic times. Past closures were facilitated by the terminus pushing glaciomarine sediments above tidewater near Gilbert Point (where past dams have closed Russell Fjord), thus limiting calving losses and allowing the glacier to advance rapidly across the 200-300 m gap. A new push moraine is currently emerging in the same location as past dam forming events, causing concern that a new closure may be eminent, perhaps as early as this winter (2007-08). Such an occurrence is of concern to local inhabitants because sustained damming of Russell Fjord will cause the lake to overflow into the Situk River, dramatically changing the landscape, creating floods, destroying fish habitats, and threatening structures. In this poster we will present results of October 2007 field measurements and remote sensing investigations on the evolution of the glacier push moraine and provide updates on the potential for another closure of Russell Fjord.
C41A-0044
Studies of contemporary glacier basal ice cryostructures to identify buried basal ice in the permafrost: an example from the Matanuska Glacier, Alaska.
In the permafrost, massive ice bodies occur as buried glacier ice, aufeis ice, recrystalized snow, massive segregated ice, injection ice, ice wedges or ice formed in underground cavities ("pool ice", "thermokarst-cave ice"). The origin of massive ice bodies in the permafrost bears considerable implications for the reconstructions of paleoenvironments and paleoclimates. Our work aims to help the permafrost scientists working on massive icy sediments to distinguish buried basal glacier ice from other types of buried ice. To do so, the properties and structure of contemporary basal ice must be well known. Field investigations at the Matanuska Glacier (Chugach range, South-central Alaska), consisted in descriptions and sampling of natural basal ice exposures. We have used the basal ice facies classification of Lawson (1979) which is simple, easy to use in the field and provides a good framework for the description of basal ice exposures. Cores were extracted and brought back to the laboratory for water and grain-size analyses. The sediments forming the cryostructure were mostly polymodal, poorly sorted gravelly silt to gravelly fine sand, with mud contents generally over 50%. These data will be used to calibrate three-dimensional (3D) models produced from micro-tomographic scans of basal ice which will produce quantitative estimates of volumetric ice and sediments contents of basal ice cryostructures. Ultimately, visual qualitative and quantitative characterization of the basal ice components of 3D models together with field observations and laboratory analysis will allow for a new micro-facies and cryostructures classification of the basal ice. Our work will also have applications in glaciology, glacial geology, geomorphology, Quaternary and paleo-climatological studies based on inferences made from the structure of basal glacier ice. This paper presents the internal composition of the basal ice facies in terms of cryostructures assemblages (Fortier et al.: 2007) and sedimentological properties. Fortier, D., Kanevskiy, M, Stephani, E., Dillon, M., Shur, Y. 2007. Facies and cryostructures of glacier basal ice as an object of permafrost study, an example from the Matanuska Glacier, Alaska. Canadian Quaternary Association Conference, Ottawa, June 2007: 75. Lawson, D.E. 1979. Sedimentological analysis of the western terminus region of the Matanuska Glacier, Alaska. Cold Regions Engineering and Research Laboratory, Hanover, N.H., Report 79-9.
C41A-0045
Quantification of Dead-ice Melting in Ice-Cored Moraines at the High-Arctic Glacier Holmströmbreen, Svalbard
An extensive dead-ice area has developed at the stagnant snout of the Holmströmbreen glacier on Svalbard following its Little Ice Age maximum. Dead-ice appears mainly as ice-cored moraines, ice-cored eskers and ice- cored kames. The most common dead-ice landform is sediment gravity flows on ice-cored slopes surrounding a large ice-walled, moraine-dammed lake. The lake finally receives the sediment from the resedimentation processes. Dead-ice melting is described and quantified through field studies and analyses of high-resolution, multi-temporal aerial photographs and satellite imagery. Field measurements of backwasting of ice-cored slopes indicate short-term melting rates of c. 9.2 cm/day. Long-term downwasting rates indicate a surface lowering of ice-cored moraines of c. 0.9 m/yr from 1984-2004. Different measures for dead-ice melting are assessed in relation to the temperature record from Svalbard since the termination of the Little Ice Age. The most prominent impact of dead-ice melting is the evolution of the ice-walled lake with an area increasing near-exponentially over the last 40 years. As long as backwasting and mass movement processes prevent build-up of an insulating debris-cover and expose ice-cores to melting, the de-icing continues even though the area is characterized by continuous permafrost.
C41A-0046
Seismic and Glaciological Implications of the 2001 Tsar Mountain Rockslide on a Clemenceau Icefield Group Cirque Glacier, BC, Canada
A long runout rockslide off the south ridge of Tsar Mountain was detected on a Clemenceau Icefield Group cirque glacier. The size, morphology, sedimentology, timing, and possible mechanisms for this rockslide were determined using satellite imaging and airborne photography in combination with analysis of seismic data and automatic snow pillow climate data. Seismic waveform data for the region suggest that several large mass wasting events might have taken place on nearby glaciers over the past 30 years. The Tsar rockslide has a rhomboidal shape with prominent levees around the perimeter. With a length of 1.3 km, footprint area of 1.37 km2, an estimated volume of 1 x 107 m3, and a fahrböschung of 13.8° (tan 0.24), it is within the expected range of values for catastrophic rockslides on other glaciers. Slide debris ranges from sand and pebble size to boulders with diameters up to 25 m. The larger clasts appear concentrated on, or close to, the levees and some reverse grading was observed. Sediment thickness varies from 30 m in the levees to an estimated 1-10 m at the centre. No disturbance of the glacier surface or its overridden medial moraines can be detected, suggesting limited interaction between the base of the slide and the substrate. The rockslide covers 24% of the cirque glacier, but the mass balance effects might be limited because of the pre- existing debris cover. However, an increase in basal shear stress from the addition of mass might have resulted in an acceleration of the glacier. Satellite imaging narrowed down the timing of the rockslide between 9 July and 11 August 2001. Of the 13 seismic events recorded within this period, only one had its epicentre close to the cirque, suggesting that the event took place on 13 July 2001 at 18:09 local time. This coincides with the end of winter snowmelt, night-time freezing after a week of high temperatures, and a precipitation event in the order of 20 mm/day. Twenty-two similar shallow seismic events occurred between 1986 and 2007, of which two in the rockslide cirque, and five on the neighbouring Shackleton Glacier. Even though not all of these mass wasting events result in long runout rockslides, a glacier's sediment yield will be significantly affected by this intermittent supraglacial sediment supply.
C41A-0047
Ice-rafted Landslide Deposits Versus `Regular` Supraglacial Debris --- an Issue for Paleoclimatic Reconstruction?
Parts of many glaciers in high mountains are covered by debris. The debris may originate from recurrent rock falls or from less common, but much larger landslides. Although relatively rare, rock avalanches rapidly distribute thin sheets of blocky debris over large portions of a glacier`'s surface. The debris sheets affect glacier regimen by decreasing ablation and increasing glacier velocity. In some instances, they may trigger surges. A supraglacial debris sheet is modified as it is carried downvalley by the glacier and ultimately is redeposited at the glacier`'s lateral margins and snout as a moraine. Although these moraines differ in lithology and sedimentology from those built by climatically driven glacier advances, they have been misidentified as such in many mountain ranges, including the Karakoram, Italian Alps and Southern Alps of New Zealand, leading to erroneous conclusions about Holocene climatic fluctuations. Work in the Karakoram and elsewhere has improved our ability to differentiate Quaternary moraines and landforms produced by rock avalanches. Few workers however, have attempted to distinguish moraines consisting of catastrophic landslide debris transported, en masse, by glaciers, from moraines formed during, or at the culminations of, glacier advances. To provide a better basis for interpreting moraines derived from landslides, we collected sedimentologic data from the debris sheet of an earthquake-triggered rock avalanche that fell onto Black Rapids Glacier, Alaska, in November 2002. Macrofabric and matrix texture distinguish the rock avalanche deposit from supraglacial debris deposited by rock fall, sampled in the vicinity of the landslide. On a landscape scale, the rock avalanche deposit has a coarse boulder rim, crude inverse grading with a blocky surface cap, a matrix of silt and sand, and conspicuous striping, with alternating bands of finer and coarser monolithologic material oriented parallel to the direction of debris transport. Similar features have been reported from other rock avalanches. Clasts are generally larger, more poorly sorted, and more angular than rock fall clasts transported longer distances at the surface of a medial moraine. Contoured fabric diagrams indicate a strong preferential orientation of elongate boulders parallel to fine and coarse stripes at the surface of the proximal part of the debris sheet. Clast macrofabrics from distal parts of the landslide deposit are weaker, and no transverse fabrics were measured. The debris sheet is being transported downvalley by Black Rapids Glacier at a rate of about 30-35 m per year, and in the five years since its emplacement has become elevated ~15-20 m above adjacent debris-free glacier ice. It remains unclear how the characteristic physical properties of the debris sheet will be modified prior to its emplacement as lateral and end moraines later in this century and in the next.
C41A-0048
Stream Flow Contributions of Rock Glaciers in the Southern Sierra Nevada Mountains of California
Rock glaciers are an important contributor to base flow and a source of nitrate in high mountain environments but we presently have little understanding of the magnitude of this contribution or their biogeochemical behavior. In this study we investigate 2 ice-cored rock glaciers located in the Southern Sierra Nevada and determine the contribution they provide to surface flow in their respective watersheds. Water samples were collected at the outlet and downstream of the rock glaciers in the summer and fall of 2006 and 2007 and stream flow was monitored continuously at the outlet of the rock glacier from July 2006-2007. Water samples were analyzed for stable isotopes ( δ 18O and δ) and major ions. Results show a distinct shift from snow to subsurface dominated flows occurred seasonally from July to October in 2006 and also intra-annually from 2006 to 2007 as δ D increased from -125‰l to -110‰l, when the subsurface flows originated from the melting ice-core of the rock glacier. This earlier shift to subsurface dominated stream flow corresponded with a lower than normal snow pack. Using the stable isotope ( δ D) and conductivity, a two- component hydrograph mixing model provided further details indicating that snow melt accounted for >60% of the stream flow, but declined to <40% in October 2006 and July 2007. Nitrate concentrations generally increased seasonally from 5.3 μ eq L-1 to 20.5 μ eq L-1 suggesting the existence of unique microbial communities within rock glaciers. However, these nitrate values are significantly lower than results from the Rocky Mountains where atmospheric contributions to nitrate formation are higher.
C41A-0049
Surface Velocities of Himalayan Glaciers: Implications for Glacial Erosion Potential During Climatic Change
Mountain glaciers in the high elevations (> 3.5 km) of the Himalaya are very efficient erosion agents. Glacier size and thus the area affected by glacial erosion are controlled by climatic conditions. Understanding the impact of climate change and variability on glacial budgets and erosion requires knowledge of the erosive potential of glaciers, which is inferred to scale with ice flux. Here, we use ASTER satellite imagery in combination with the orthorectification and correlation tool COSI-Corr to derive horizontal surface velocities of glaciers from several regions across the Himalayan-Karakoram domain. Our results show that glaciers in the Eastern and Central Himalaya, where precipitation is mainly supplied by the Indian Summer Monsoon, are relatively slow, with velocities usually below 50-60 m/a. In contrast, glaciers in the Western Himalaya and Karakoram, receive a significant amount of precipitation during the winter months and are considerably faster with velocities often exceeding 80-100 m/a. This discrepancy is visible among glaciers of different size and orientation although local slope and catchment area effects may cause velocity excursions. A relatively sharp gradient appears to exist in the catchment area of the Sutlej River in the NW Himalaya of India at approximately 79°E. To the east, glaciers in the Garhwal Himalaya – among them Gangotri glacier, the largest in the Indian Himalaya – have mean velocities of around 20-40 m/a, whereas glaciers in the much drier Lahul region to the west attain mean velocities of around 30-60 m/a. Importantly, the Sutlej River valley marks a climatic transition zone from an annual summer-rainfall maximum (more than 75% of annual rainfall during the summer) to the east to a winter-rainfall maximum (more than 60% of annual rainfall during the winter) to the west. These observations corroborate the notion of a significant climatic boundary in this part of the Himalaya, which may have shifted west- and northward during past episodes of intensified summer monsoons, such as during the early Holocene. Our findings show that glaciers dominated by summer accumulation have generally lower surface velocities, hence ice-flux, and thus only limited potential to erode underlying bedrock. However, their counterparts in the Western Himalaya and Karakoram, receive moisture during all seasons, have a higher ice- flux, are more likely to have grown during an intensified winter or summer monsoon and thereby play a more important role in sculpting landscapes. Yet, if an intensified monsoon coincides with lower temperatures, such as during MIS 3-4, even glaciers in the Eastern and Central Himalaya should have had favourable conditions to advance toward much lower elevations, with higher ice flux.
C41A-0050
Measurements of Fast ice Flow of the Malaspina Glacier to Explore Connections Between Glacial Erosion and Crustal Deformation in the St. Elias Mountains, Alaska
In the St. Elias range of southern Alaska, dynamic tectonics and massive, fast-moving temperate glaciers provide an ideal opportunity to explore linkages between concentrated glacial erosion and localized deformation. With the potential for rapid erosion, the Seward Throat funnels a large volume of ice through the St. Elias Mountains, slicing through inactive and active structures alike and spreading out to form the Malaspina piedmont glacier. While glacial erosion processes are complex, erosion rates tend to scale with the sliding velocity, making it a useful indicator of erosive potential. In simple glacial valleys, conservation of mass dictates that reaches with high ice speeds tend to be relatively shallow; interestingly, these areas where the glacier bed is relatively high are precisely where the erosion is expected to be fastest. Hence, bedrock uplift must be concentrated for the elevated parts of the glacier bed to maintain their positions. Systematic studies of glacier speed and ice thickness hold considerable promise for elucidating patterns of active tectonics. We investigate the variation in velocity, and thus the probable variation in erosion rate, along a portion of the glacier's length. In late July 2007, 10 targets were placed on the lower Seward Throat and surveyed for 8 days. 7 targets were placed along 6 km of a flow line to define the velocity variation through steep and gentle sections of the glacier. Preliminary determination of surface velocities reveal large longitudinal strain rates with velocities varying from ~1400 m/year to ~1800 m/year within 1500 m with surface slopes ranging from ~1° to 3°. Previously calculated balance velocities, which suggest that most of the motion results from sliding, compare encouragingly. Assuming a 1 bar basal shear stress to estimate the ice thickness, basal sliding comprises between 83% and 95% of the surface velocity. To extend the spatial coverage of our measurements, our survey results will be examined in the context of detailed, unpublished velocity measurements collected by the USGS (Robert Krimmel and Austin Post) in the 1970s, as well as surface velocities derived from Radarsat-1 interferometric synthetic aperture radar data collected from 2000 to the present. The InSAR velocities were derived using a speckle tracking technique over 24- day periods, describing the interannual as well as seasonal variability in speed. The velocities from different time periods are influenced by climate and possible surge behavior; hence we expect temporal variations in the absolute values of the velocities but not in the spatial pattern. Analysis of the surface strain rates will enable us to improve calculations of basal velocities by relaxing assumptions such as uniform basal shear stress and negligible sidewall drag. An enhanced understanding of the glacier's dynamics, combined with information about ice depth from upcoming air-borne ice penetrating radar measurements, will enable us to better investigate areas with high erosive potential in the context of the active deformation and seismic activity in the region studied by the St. Elias Erosion/Tectonics Project.
C41A-0051
A Poleward Transition From Destructive to Constructive Glacio-Climatic Control on Mountain Building
Alpine glaciers have been hypothesized to have sufficient erosive power to control the ultimate height of actively developing mountains regardless of the rate at which uplift occurs. This process is commonly labeled the "glacial buzzsaw" and has been corroborated by the observation of remarkable correlation between glacial equilibrium line altitude (ELA) and mean elevation in several mid- to high latitude active orogens independent of highly variable late Cenozoic exhumation rates. In contrast, extremely low late Cenozoic erosion rates have been demonstrated below frozen-based subpolar to polar glaciers and ice sheets with the ice acting to protect the landscape from further erosion potentially leading to increased relief and mean elevation in actively uplifting mountain ranges. In cases where efficiency of glacial erosion during late Cenozoic climate change is strong, then following onset of late Cenozoic climate change and ELA lowering, a significant increase in erosion rates is predicted, the magnitude of the rate change being higher if the orogen responds by active uplift driven by accretion as opposed to passive isostatic rebound. Alternatively, if after onset of climatic cooling, glacial erosion is inefficient, owing to, say, the inception of cold-based glaciers and/or slow-moving ice sheets, then no increase in erosion rates is predicted. To investigate the efficiency of glacial erosion and its effects on mountain building we have undertaken low- temperature thermochronologic analysis (fission-track and (U-Th)/He dating) at various latitudes along the Chilean Patagonian Andes: a high latitude active orogen with a well-documented late Cenozoic tectonic, climatic, and glacial history. New data from regional transects from 38°S to 49°S reveal that the highest rates and magnitudes of late Cenozoic erosion are restricted to the main divide and its windward western flank. Here age-elevation relationships demonstrate a marked and consistent acceleration in erosion at 8 to 6 Ma to rates between 0.4 and 0.6 mm/yr coeval with the timing of onset of major Patagonian glaciation between ca. 5 and 7 Ma, but well after initial surface uplift of the Patagonian Andes at around 17-14 Ma. The estimated erosion rates are consistent with conceptual predictions for either (1) moderately efficient glacial erosion acting on an active orogen, or (2) very efficient glacial erosion acting on an inactive orogen responding by passive isostatic rebound. In contrast, new thermochronometric data from further south (49°S to 56°S) show no evidence of enhanced late Cenozoic bedrock erosion along the main divide with old (>10 Ma) apatite (U-Th)/He ages despite the apparent presence of widespread glaciation. These data imply that that long-term glacial erosion has been less efficient further south. We speculate that during each of the many glacial cycles in the late Cenozoic, the Patagonian Andes south of ca. 45°S glaciation was dominated by either frozen-based glaciation and/or slow moving ice caps. Inefficient glacial erosion at these more southern latitudes also explains well the anomalously high non-volcanic summit elevations along the actively uplifting main divide within the South Patagonian icefield between 46°S and 51°S now situated well above the modern and last glacial maximum ELA.
C41A-0052
Riverine habitat and relative size of fluvial and glaciated valleys in central Idaho
Our recent investigation of glaciated and fluvial valleys in central Idaho compared differences in valley relief and width in otherwise similar geologic and geomorphic settings (Amerson et al. in press). We found that local valley relief and width in glaciated valleys is greater than in fluvial valleys. Local valley cross-sectional area in glaciated valleys increases with drainage area with, however, glacial valleys have greater cross-sectional area. Hillslope angle distribution differed between glaciated and fluvial valleys also; glaciated valleys have steeper hillslopes overall. The differences in valley form and in hillslope angle distribution in glaciated and fluvial valleys influences the channel type and therefore distribution of aquatic habitat in the river flowing through them. The distribution of anadromous salmonids (Oncorynchus spp.) is influenced by differences in stream gradient, pool density, and side-pool abundance. Glaciated and fluvial valleys have differences in the amount of low-gradient terrain (<5%), much of which is adjacent to streams (i.e. riparian zones). The broad flat valley bottoms of glaciated valleys have more low gradient terrain and therefore a greater total length of low-gradient reaches when contrasted with fluvial basins. Field observations in low order glaciated streams show that low-gradient areas consist of pool-riffle channels that are typically found in higher-order streams, which have floodplains that support a complex of side channels created by downed, large coniferous trees. Pool-riffle channels with abundant wood provide ideal habitat for anadromous fishes and the valley bottom morphology of glaciated valleys appears to help set the stage for productive fish habitat in upland regions of Idaho.
C41A-0053
Relief History and Coupling of Glacial Valley and Hillslope Erosion in the Teton Range, Wyoming
Alpine landscapes are the product of coupling and competition between erosion and weathering mechanisms. In the Teton Range, Wyoming, a unique, rugged landscape has evolved through a combination of block tilting, fluvial and glacial incision, cirque retreat, and physical weathering and mass wasting of ridges. We are studying the interaction and effectiveness of these processes using field observations and detrital thermochronology, with particular focus on changes associated with Late Cenozoic climate change. The Tetons are an ideal location for this investigation, because the patterns of net rock uplift and incision since onset of normal faulting at 9 Ma are structurally constrained. To investigate the erosion pattern since glacial advances began, we have used apatite (U-Th)/He detrital thermochronology of moraine and modern river sands to identify where sediment is sourced based on an established bedrock age-elevation gradient and basin hypsometry. Preliminary results from modern river sediment in Garnet Canyon are plagued by poor sample quality, but show a disproportionately large component of young ages, presumably sourced from low altitudes. This may reflect glacial incision, consistent with an increase in relief associated with Late Cenozoic global cooling that has been observed in numerous locations. However, evidence of hillslope denudation implies that glacial erosion may have brought the Tetons close to a topographic steady state. Estimates of mass flux based on surveys of talus fans in Garnet Canyon suggest ridges have recently eroded at close to the rate of long term rock uplift, whereas the long term average rate of peak and ridge erosion must have been much slower. We interpret that glacial incision increased relief up to a point, but then over steepened hillslopes, such that the landscape reached a threshold for mass wasting. Observations of densely spaced joints and fractures in the bedrock suggest material properties may have facilitated reaching this threshold. The Teton landscape thus illustrates the complexity of integrated erosional mechanisms through transitional climate conditions.
C41A-0054
A laboratory apparatus to measure clast-bed contact forces
Glacier dynamics, sediment transport, and erosion are controlled in part by processes occurring at the interface between basal ice and bedrock. One critical parameter is the contact force between a clast and the bedrock. This force affects many processes such as basal friction which regulates sliding speed, slip resistance which influences basal shear stress and may cause micro-seismic events associated with slip instabilities, abrasion which controls rates of erosion, landscape evolution, and production of sediments. Despite field and laboratory evidences indicating that contact forces may be up to one order of magnitude higher than estimated from leading theories, no studies have yet measured with precision the magnitude of contact forces and how contact forces vary as a function of key glaciological variables such as basal melt rate and effective pressure. An apparatus was designed to make two independent measurements: (1) the contact force between a clast and a hard bed as a function of melt rate and effective pressure; (2) the drag force on an identical clast away from the bed as a function of the ice speed. The contact force differs from the drag force because of the presence of the bed which modifies the ice flow field. Measurement (2) is necessary to estimate the rheological properties of the ice and to quantify wall- (bed) effects on the drag force. The apparatus consists of a hydraulic press that pressurizes an ice cylinder, 24~cm high and 20~cm in diameter, to 1.0 - 1.5~MPa. The ice cylinder is contained inside a polycarbonate vessel. Above and below the ice cylinder are three disks: an aluminum disk sandwiched between two Delrin disks. The aluminum disks are hollow and used to circulate a fluid at a controlled temperature. The Delrin disks are used to isolate the ice from the cold room and to control the flow of heat to the ice block. The ice is kept at the melting temperature by circulating a fluid in channels inside the polycarbonate vessel and in the top aluminum disk. The fluid is pumped from an external bath at a controlled rate and temperature. The melt rate at the ice-bed (bottom Delrin disk) interface is controlled by a fluid circulating in the bottom aluminum disk at a temperature slightly above the melting temperature. The fluid is pumped to the disk by a second external bath. To measure the drag force ice exerts on a clast, a Delrin sphere is suspended by Kevlar filaments to a load cell above the ice cylinder. The sphere remains immobile as ice moves downward by melting at the ice-bed interface. To measure the contact force, a Delrin sphere sits above a ceramic rod running through the center of the three bottom disks. The rod transmits the contact force to a load cell beneath the apparatus. Temperature is monitored with thermistors at the ice-bed interface and in the ice. Ice motion is measured with a LVDT. Water pressure at the ice-bed interface is measured with two piezometers. Three drains on the outside perimeter of the Delrin bottom disk remove melt water. The effective pressure underneath the sphere is controlled by a valve that opens or closes a drain near the center of the bottom Delrin disk.
C41A-0055
Permafrost degradation in the source area of the Yellow River, Northeastern Tibet
Frozen ground was investigated in 2003-2006 to evaluate the present-day distribution and ongoing degradation of permafrost in the source area of the Yellow River, located at the northeastern margin of the Tibetan Plateau. Distribution of permafrost was examined by seismic, electrical and/or thermal soundings at 18 sites between 3250 m and 4800 m ASL. Temporal variations in ground thermal and hydrological regimes were also investigated for two years at Madoi observatory (4273 m ASL), by automatic and manual observations of air and ground (0-8 m deep) temperatures, precipitation, snow depth, near-surface soil moisture and groundwater level. High P-wave velocities (>2 km/s) and relatively high DC resistivities (650-1100 ohmm) below a thin uppermost layer showed that permafrost 10-30 m in thickness occurred above 4300 m ASL. In contrast, low P-wave velocities (<1 km/s) throughout the sediments indicated that permafrost was absent below 4000 m ASL. On widespread alluvial plains between 4200 m and 4300 m ASL, permafrost was lacking or significantly degraded. Negative values of the mean annual surface temperature (MAST) also indicated widespread permafrost only above 4300 m ASL under the present climatic condition. The seasonal frost penetration reached a maximum depth of 2.6 m at the observatory. Intermittent and very shallow snow cover favored frost penetration. The ground between 4 m and 8 m in depth was kept at slightly positive temperatures (0-4°C) throughout two years, although the presence of permafrost at this site was suggested by a few reports in the 1980s. Assuming that the inter-annual variation in MAST follows that in the mean annual air temperature, permafrost is estimated to have significantly thawed on the alluvial plains at 4200-4300 m ASL during the last half-century. The resulting degradation of the permafrost is assumed to have extended 3000 km2 on the alluvial plains in the source area.
C41A-0056
A Laboratory Scale Study of the Fluid and Energy Dynamics in Non-sorted Circle Systems of Alaska
Non-sorted circle ecosystems are naturally stable and abundant ecosystems throughout the Arctic tundra. These are formed by processes of differential frost heave and are an important component of the water cycle in tundra regions. Even though the soil surface is basically bare, these systems are very stable. The stability of the system equilibrium and associated water and energy fluxes is, however, poorly understood. Water and ice are critical to non-sorted circle formation and interact in complex ways with the surrounding vegetation, soils, atmosphere and hydrological systems. Climate warming is expected to significantly affect such complex interaction in these ecosystems. It is also not clear how these systems will react to a larger shift in the perturbation of the soil surface. The objective of this study is to develop a laboratory scale model using soil from a non-sorted circle ecosystem to understand the effects of differential freezing on water and energy fluxes across its boundaries due to variations in temperature and insulation. We have generated a heave pattern using an insulation pattern that is similar to what is observed in the field using this laboratory setup. We have identified that preferential ice accumulation occurs underneath the soil and is sensitive to the freezing temperature and the amount of insulation. The amount of heave cannot be described by the expansion of water alone, secondary processes of water movement are very important in frost susceptible soils like the one used in the experiment. Since non- sorted circle systems are ubiquitous to Arctic Alaska, our research focus is expected to complement the existing knowledge base in pan-Arctic studies by improving the understanding of the ecosystem functioning, structure and stability at the regional scale.
C41A-0057
Late Wisconsin Permafrost Conditions Evidenced by Patterned Ground in the Saginaw Lowlands, Eastern-Central Lower Michigan
Permafrost associated with the Laurentide ice sheet has never been positively identified in the state of Michigan, in contrast to surrounding states and Canada. This study examined patterned ground in the Saginaw Lowlands of eastern-central lower Michigan to see if freeze-thaw processes in a permafrost environment had led to its formation. The area of the patterned ground is constrained by two glacial lake strands, Lake Warren and Lake Elkton-Lundy, limiting the age of the patterned ground to approximately 14.3-14.8 thousand radiocarbon years BP. Aerial photo analysis revealed a widespread area (>1020 km2) of elongate polygons in a reticulate mesh, individually measuring between 150 and 160m along the long axis and 60-90m along the short axis. Surface electrical resistivity (ER) tomography and invasive soil studies were conducted at a small sampling of mapped patterned ground locations. Polygon centers typically stood about 1 meter higher in elevation than corresponding inter-polygon swales, and often exhibited a sandy covering above the silty clay loam till below. A bowl-shaped sandy loam deposit in one polygon edge/suture was determined to be a thermokarst channel infilling following the trace of a former ice wedge. The subtle morphology and sedimentological characteristics of the patterned ground in the Saginaw Lowlands suggest that thermokarst erosion, rather than ice-wedge replacement, was the dominant geomorphic process associated with the degradation of the Late-Wisconsin permafrost in the study area.
C41A-0058
Near Realtime Monitoring of Tidewater Glacier Advance and Retreat: Hubbard Glacier, Southeast Alaska
Tidewater glaciers advance and retreat seasonally as part of the well-documented tidewater glacier cycle. Quantitative factors that influence tidewater glacier activity are often difficult to capture. The accurate collection of such data is a key to better understanding this cycle. In this paper, we describe the results of current research at Hubbard Glacier to quantify realtime glacier terminus activity using high-resolution ranging measurements and local meteorologic conditions. Hubbard Glacier is the largest non-polar tidewater glacier in the world. It encompasses an area of 3500 sq km and flows 120 km from the flanks of Mt Logan (5959 m) in the Wrangell St. Elias Mountains (Canada) to sea level where its terminus widens to ~11.5 km. In contrast to most glaciers in Southeast Alaska, Hubbard Glacier continues to advance and thicken and is predicted to continue for the foreseeable future. Our data provide unique, high-resolution (4 times daily) measurements of ice margin motion along a section of the terminus near Gilbert Point. Gilbert Point is the location of past ice dams formed by Hubbard Glacier (1986, 2002) both of which failed and generated massive outburst floods. This location is critical to understanding the potential for future ice dam formation by Hubbard Glacier. Preliminary results of monitoring since late 2006 indicate variable rates in ice motion within periods of overall advance and retreat. A seasonal advance of the glacier occurred from March to mid-June 2007, advancing 280 m in 98 days at an average of 2.85 m/day, but with daily average rates occasionally as high as 10 m/day. A period of retreat began in mid-June with a large calving event that removed 60 m of ice from the terminus in less than a day. This seasonal retreat has continued with retreat rates varying from 1.11 m/day in the first 45 days to 2.39 m/day between August 1st and 30th, 2007. http://www.GlacierResearch.org
C41A-0059
Estimation of meteorological inputs for regional modeling of glacier melt
A challenge for regional modeling of snow and glacier melt is the specification of the driving meteorological variables, including air temperature and humidity. Progress has been made over the last decade in developing methods for regional interpolation of these variables from climate station networks for modeling seasonal snowmelt. However, these approaches cannot be directly used for modeling glacier melt because they do not account for the effects of glacier boundary layer processes. For example, air temperatures within a katabatic glacier boundary layer should be lower than those interpolated from a climate station network due to the cooling associated with sensible heat transfer to the glacier. We present preliminary analyses of meteorological data from two glaciers in the southern Coast Mountains, British Columbia, to demonstrate how simple elevation- based models can be used to remove the biases associated with using regional interpolations from climate networks to estimate temperature and vapour pressure within the glacier boundary layer.
C41A-0060
Modeling Future Sea Level Rise From the Melt of Glaciers: Assessment of Uncertainties
Melting and disintegrating mountain glaciers have been identified as the second largest contributor to rising sea level after thermal expansion of the oceans (e.g., IPCC, 2007; Meier et al., 2007), and Meier et al. (2007) show that melting mountain glaciers are likely to remain the dominant glaciological contributor to rising sea level through the end of the 21st century. Current work will be presented assessing the sources of uncertainty in model-derived estimates of the probable future contributions from glacier wastage to rising sea level. To evaluate uncertainties associated with the choice of glacier mass balance model, we apply three temperature-index and two energy mass balance models to Storglaciären, a small well-measured valley glacier in northern Sweden. The five mass balance models are individually calibrated using ERA-40 reanalysis data from past years. These models are then forced during future years using statistically downscaled regional climate model outputs in order to simulate the future mass balances of Storglaciaren. The cumulative mass balance for the time period 2002 to 2100 AD in response to predicted temperature changes is found to vary between -81 and -92 m for four models but is estimated at -121 m for the fully distributed energy balance model. This demonstrates the sensitivity of the results to the choice of mass balance model. To investigate the sensitivity of projected future changes in the volume of Storglaciaren to the choice of climate model, we used temperature and precipitation outputs from different global climate models (GCMs) to force a temperature-index glacier mass balance model. The results show that the volume-change projections vary by 40% of the initial glacier volume for six different GCMs. The projections showed volume losses by 2100 AD of 50% to 90% of the initial volume of Storglaciaren. Since these volume projections are computed using a volume-area scaling approach, we further investigate the scaling approach relative to the projections obtained using an ice flow model. For this analysis we calibrate a one-dimensional ice- flow model for 6 glaciers with available surface and bed topography maps and sufficiently long records of length fluctuations and mass balance observations. The calibrated model is then forced with a hypothetical mass balance perturbation to produce 100-year volume evolutions. The same mass balance perturbations are used in the scaling approach to derive volume evolutions which are then compared to the modeled simulations of volume change. These comparisons show that the volume-area and volume-length scaling methods underestimate the volume loss by up to 50% and 15%, respectively, relative to the loss predicted by the ice-flow model. These results show that projections of future changes in glacier volume are highly sensitive to the method used to simulate the effects of changes in glacier geometry during the projections.
C41A-0061
Identifying the climatic drivers of Southern Hemisphere Ice Ages from glacier modelling in New Zealand
The timing of glacier advances and retreats during the past 30,000 years is broadly consistent between the Northern and Southern hemispheres. In particular, cosmogenic exposure age dating of moraines indicates that the timing of glacier advance and retreat cycles in New Zealand, Chile and Australia broadly matches that of the Northern Hemisphere. And yet the precession component of the insolation forcing is out-of-phase between the hemispheres. If summer insolation at high Northern latitudes is to pace global glaciation, some strong inter hemispheric connection must exist. Southern Hemisphere glaciers may respond to changes in greenhouse gas concentrations, or to shifts in oceanic and atmospheric circulation. Alternately, Southern Hemisphere glaciation may respond to some other component of the insolation forcing or vary independently and, by chance, coincide with the North. In order to draw distinctions between the various influences on Southern Hemisphere glaciations, we explore how models of Southern Hemisphere glaciers respond to a variety of forcing agents. The model includes ablation calculated using an energy balance model, including full seasonal insolation forcing, and accumulation calculated using a precipitation model including the effects of moisture transport over an orographic barrier. A dynamical ice-sheet model relates the resulting mass balance variations to changing ice extent, which is compared to the geological record. The model is used to quantify the relative importance of changes in precipitation and temperature resulting from local intrinsic variability, as a response to local insolation forcing, or resulting from ocean/atmosphere changes induced non-locally, possibly as a result of Northern Hemisphere climate.
C41A-0062
Structural Glaciology, Dynamics, and Changes of the Grand Pacific Glacier System, Saint Elias Mountains, British Columbia, Canada, and Alaska, U.S.A.
The cryosphere has been undergoing drastic and rapid changes in recent years. Mass loss is reported from most glaciers and ice sheets worldwide. At the same time, there is strong evidence for global warming. In Alaska, the majority of glaciers is retreating, whereas a considerable number of glaciers is advancing or fluctuating. So, although it may appear clear that glaciers are melting in response to climatic warming, the relationship is not as simple and straightforward as that. Glaciers are dynamical systems, and fast-moving glaciers in particular follow their own internal dynamics, although ultimately all glaciers are subject to climatic forcing. When assessing a glacier's mass balance or terminus retreat/advance in relation to climatic change, it is crucial to take into account of which dynamical type that glacier is. In remote areas, however, this is often unknown. In this context, we have been developing a structural-glaciological methodology that allows for the dynamical classification of glaciers from remote-sensing data (aerial or satellite) based on structural analysis of surface deformation structures and their patterns. Here, we present a structural overview and characterization of some dynamically differing and interesting glaciers in the Grand Pacific Glacier System of the Saint Elias Mountains in British Columbia, Canada, and Alaska, U.S.A., and assess their recent changes.
C41A-0063
Increasing Wastage of the Bering and Malaspina Glacier Systems, Alaska-Yukon, 1972 to 2006
Ice dynamics are integral to the net mass balances of the huge Bagley-Bering and Seward-Malaspina Glacier systems of south-central Alaska. Quasi-periodic surging of the main trunks and some large tributaries of these exceptionally active glacier systems are important contributors to their increasing volume losses in the present rapidly-warming climate, because surges rapidly transport ice from higher elevations, where it is "safe," to lower elevations where it subject to increased ablation. New estimates of mass losses from the Bering and Malaspina Glacier systems during 1972-2006 were derived from analysis of (i) digital elevation models (DEMs) synthesized from airborne and spaceborne interferometric synthetic aperture radar (InSAR); (ii) small-aircraft laser altimetry; and (iii) spaceborne laser altimetry acquired by ICESat. Adjustments for estimated seasonal snow accumulation were applied to datasets acquired at times subsequent to late summer. Adjustments for systematic DEM biases were also applied. The area-average lowering rate on the main-trunk of the Bering Glacier system from 1972 to 1995 was 0.9 ± 0.1 m/yr. The major 1993 to ‘95 surge moved ice rapidly from the surge reservoir into the piedmont lobe where rapid surface melting was facilitated by the heavily crevassed surface. The lowering rate accelerated to 3.0 ± 0.1 m/yr during 1995 to 2000, then moderated to 1.4 ± 0.1 m/yr during 2000 to 2003. On the Malaspina Glacier system, the area-average rate of surface lowering was 1.4 ± 0.1 m/yr during 1972 to 1999. It then increased by 30% to 1.8 ± 0.1 m/yr during 1999 to 2002. Near-concurrent surges of Agassiz Glacier (a west piedmont lobe tributary), lower Seward Glacier (main source for the central Seward lobe), and Marvine Glacier (a detached former tributary of the eastern piedmont lobe) were observed during this 3-year time span of increased surface lowering. Recent ICESat-derived elevation changes from 2003 to 2006 indicate increasing wastage on the Malaspina piedmont lobe. By contrast, its main accumulation area, upper Seward Glacier, which was drawn down by the 1999-2002 surge, is showing recovery with increasing surface elevations. Concurrently, elevations on Bagley Ice Valley are also increasing in preparation, evidently, for the next surge of the Bering Glacier system. For both of these large glacier systems we estimate a combined volume loss of 254.0 ± 16.5 km3 (water equivalent) over an area of 7734 km2 during 1972 to 2003, representing over 80% and 70% of the areas of the Bering and Malaspina Glacier systems, respectively. This is equivalent to a mean surface lowering of 31 to 35 meters. These glaciers are making an increasing contribution to globally-rising sea-level.