C11D-0525
Modelling of the flow regime of the Antarctic drainage system from Dome F to Shirase Glacier
The Antarctic ice sheet is primarily drained by large outlet glaciers and ice streams which play a decisive role in the dynamic behaviour of the ice sheet. Shirase Glacier is one of these fast-flowing glaciers, characterized by surface velocities of > 2500 m a-1 near the calving front. The Shirase Glacier catchment area extends upstream towards Dome Fuji and is therefore important for the dynamics of the vicinity of Dome Fuji. By making use of recently processed, detailed data for the surface and basal topographies, we will apply the full-Stokes model Elmer/Ice to this drainage system. The main objective is to characterize the three-dimensional flow regime and investigate the relative contributions of slow ice-sheet flow and fast ice-stream/shelf flow. This will also allow assessing the possibility of rapid dynamic decay of the area under global-warming conditions.
C11D-0526
Frozen margin dynamics at Storglaciaren, Sweden
Since the summer of 2006, more than 40 boreholes have been drilled near the frozen margin of Storglaciaren. A variety of instruments have been deployed in these holes to investigate the influence of the temperate-to-cold-bed transition on ice flow dynamics and sediment transport near the margin. Among the instruments are thermistors, slidometers, and water pressure transducers. Four boreholes were used for inclinometry, which was repeated a year later. A stake network in the same region was surveyed with a dGPS and a total station to determine surface velocities. Ground-penetrating radar was used to study the geometry of both the bed and cold-temperate transition surface (CTS). Slidometer and thermistor data indicate that, as expected, basal motion declines toward zero where the freezing isotherm intersects the bed, indicating that the basal thermal transition (BTT) is effectively a slip/no- slip transition. However, in this area the average longitudinal gradients in surface and basal velocity during summer 2007 were -0.025 a-1 and -0.0275 a-1, respectively, indicating that compressive longitudinal stress gradients arising from the BTT are not large. Additionally, radar profiles from the same transects in winter and summer suggest that the CTS (and consequently the BTT) may migrate seasonally. A slidometer that was located within temperate-based sliding ice (about 20 m from the inferred BTT location) in summer, 2007, recorded zero basal velocity during winter but activated during summer, 2008. Comparison of the mean annual and summer surface velocities indicates that such seasonal increases in velocity do not extend far into the cold-based ice at the margin. These observations suggest that the thermally-induced slip/no-slip transition near the margin of Storglaciaren is a gradual and dynamic transition and that the frozen margin does not represent a major impediment to ice flow.
C11D-0527
Mass Balance of Storglaciären
Storglaciären, a small valley glacier (3 km2) in northern Sweden, has the longest detailed mass balance programme in the world starting in 1945/46. Measurement routines have varied with time, but since 1966 winter mass balance has been derived from snow probings in a 100 × 100 m regular grid and a few density pits and summer balance from roughly 50-80 ablation stakes across the glacier. We recalculated the entire mass balance series in a consistent manner, based on a newly digitized data set of all available snow probings, a new set of digital elevation models and glacier outlines and re-evaluation of all available density data. Results show that seasonal balances differ up to 0.4 m compared to previously published results indicating that even for a glacier where abundant data are available upon which mass balances are based uncertainties can be large. We also performed a number of sensitivity experiments to systematically evaluate the sensitivity of results to assumptions on factors, such as snow density, definition of glacier outlines, number of snow probings and ablation stakes, etc.. Mass balances turned out to be sensitive to some of these assumptions. Equilibrium line altitudes varied by more than 100 m depending on how they were derived. Our results emphasize the uncertainties associated to glacier mass balance data and the need to report details on the calculation procedures to allow such uncertainties to be evaluated.
C11D-0528
Firn and Shallow Ice Profiling at Jakobshavn Glacier Using Dispersed Seismic Surface Waves
Seismic refraction methods are used commonly to determine firn and shallow ice acoustic properties. However, this practice is hampered by the difficulty in generating shear waves in firn and by the presence of dispersed compressional waves interfering with shear wave arrivals. Conventional reflection surveys used to image the ice-bed interface and subglacial geology contain surface waves that are typically discarded as noise. Surface waves propagate within one wavelength from the surface and their dispersive characteristics can be exploited to estimate shear wave velocities of the near-surface. We employed the Multichannel Analysis of Surface Waves (MASW) method along a 10 km-long reflection line acquired on Jakobshavn Glacier, Greenland, to map the velocity structure of firn and shallow ice. Seismic data were acquired using twenty-four 28 Hz vertical geophones spaced 20 m apart and 0.5 kg of buried explosives at 10 m depth. Rayleigh wave dispersion curves along the 10 km line show phase velocities ranging from 1000 m/s at 100 Hz frequency to 1800 m/s at 8 Hz frequency. The 10 km long continuous surface wave velocity profile indicates near-surface shear wave velocity estimates increasing rapidly with depth, from 1050 m/s to 1950 m/s. Lateral velocity variations reveal low relief structures that correlate with ice internal layering imaged by a coincident ground-penetrating radar profile. Preliminary analysis of surface waves contained in the seismic reflection line yielded usable dispersion curves and a velocity profile consistent with the expected velocity structure near the surface of Jakobshavn Glacier. Use of surface wave methods is suggested as an alternative to refraction surveys for firn and shallow ice imaging.
C11D-0529
Cryogenesis study of a pingo-like mound in the Akkol valley of the Russian Altai Mountains
Vertical outcrop of a pingo-like mound found in the Akkol valley of the Russian Altai Mountains is described. Several pingo-like mounds were found on the valley floor at about 2300 m ASL. They are 5 - 10 m high and up to 50 m in diameter. Part of a 5 m mound had collapsed into adjacent pond with continuous water supply from streams on the mountain ridge nearby and top 4 m section had been revealed. Highly complex combinations of segregated ice lenses were observed in the outcrop. Ice veins, which are consisted by a number of thick ice lenses, develop radially from the core of the mound. The areas in-between the ice veins had fine parallel lenticular cryostructure. Surface soil layer (about one meter) and patchy soil parts between ice lenses were made of fine till of lacustrine sediments. delta O18 values of water from these ice lenses range from -15 to -18. Spatial distribution of the isotope values was well correlated with the spatial pattern of the ice lensesf distribution. In addition to segregation of ice lenses perpendicular to the temperature gradient, contribution of relatively rapid formation of ice in radial direction from the core of the mound can be large in this three dimensional frost heave phenomenon.
C11D-0530
Recent Acceleration of Glacier Wastage in the Wrangell Mountains, Alaska, USA, From Airborne Laser Altimetry and a Model of Orographic Precipitation
We assess the glacier mass loss of Wrangell Mountains, a heavily glacierized volcanic range in south central Alaska for the periods 1957 to 2000 and 2000 to 2007. These glaciers have a total area of about 4,700 km2 and an elevation range of about 400 to 5,000 m a.s.l. Airborne laser altimetry profiles acquired on the Wrangell Mountains in 2000 and 2007 and 1957 USGS topographical maps are used to compute area averaged mass balance in this region. Mass balances of the ocean facing and inland facing side are evaluated separately. The area averaged mass balance of the entire mountain range from 1957 to 2000 is estimated to be -0.08±0.02 m /yr ice equivalent and - 0.32±0.015 m/yr ice equivalent from 2000 to 2007. Our results show indications of thickening at elevations above 3000 m a.s.l. in both periods but coverage is incomplete and data are too sparse to be conclusive at these high elevations. Sparse data coverage at high elevations, which are about 40% of the total glacierized area, and USGS map errors, particularly in accumulation areas, are the major sources of mass balance uncertainty. To investigate the plausibility of thickening in the higher reaches of the mountain range, we have used a precipitation model based on the linear theory of orographic precipitation to simulate the orographic enhancement of precipitation across the mountain range. The model uses 12 hourly precipitation and 6 hourly temperature, specific humidity and wind data from ECMWF (European Center for Medium-Range Weather Forecasts), to compute annual and winter precipitation on a 1 km grid over the study area from 1958 to 2001. The model has been validated by weather station data as well as winter mass balance data from two glaciers outside the Wrangell Mountains. We will also present the model-simulated spatial patterns and temporal trends in precipitation over the Wrangell Mountain glaciers during 1958-2001 in relation to our airborne laser altimetry measurements.
C11D-0531
Buried Glacial Basal Ice Along the Beaufort Sea Coast, Northern Alaska
A storm along the Beaufort Sea coast in late July 2008 exposed a 2.8-km-long section of coastal bluff at Barter Island, northeastern Alaska. It created a unique opportunity to describe and map the cryostratigraphy of several exposures up to 600 m long and 10 m high. During the description and mapping, we discovered large, highly deformed bodies of massive ice that had macro- and micro-structures very similar to those of basal ice we studied at the base of the Matanuska and Muldrow glaciers. The exposed visible ice was as much as 67 m thick, up to 15 m wide as individual masses, and occurred at highly irregular depths, varying from just below the active layer to a depth of 68 m. The ice masses were highly deformed with differing orientations (vertical, horizontal, inclined) and shape (straight, wavy, folded). The ice was stratified with layers and inclusions of sand and gravel, which were irregular shaped with numerous folds. The largest clast we found in the ice was 3 cm, and there are numerous cobbles in sediment containing basal ice. The ice lacked organic material. Basal ice is incorporated in the ice-rich unsorted gravelly sand deposits, which we relate to glacial till. Fold deformations also can be observed in these deposits. Several ice-rich diapir structures formed by vertically injected glacio-marine clays occurred in the lower parts of the coastal bluff. The structure of the basal ice was in stark contrast to that of the ice wedges and thermokarst-cave ice that also occurred in the bluff. The ice wedges where distinguished by the wedge-shaped macrostructure, vertically inclined foliated micro-structure, inclusion of very limited amounts of very fine mineral and organic material, and the white to gray color. The ice wedges were 13 m wide and 24 m tall, thus much smaller than the basal ice, and were observed only in the upper part of the bluff, usually right below the permafrost table. Thermokarst- cave ice was differentiated by the occurrence of horizontal or slightly inclined lenticular shape, small size of 0.11.2 m thick and 14 m long, and the clear to yellow organic-stained color. These ice bodies were always connected to the upper parts of ice wedges. Based on the similarity of micro- and macro-morphology of the exposed ice with known basal ice at the Matanuska and Muldrow glaciers, and the dissimilarity with overlying ice wedges and thermokarst-cave ice, we conclude that the massive ice is buried basal glacier ice associated with a Late Pleistocene glaciation of the continental shelf and adjacent coastal plain. Ice with similar structures has been observed at the Beaufort Sea coast in Canada and in northern Eurasia.
C11D-0532
Role of Shelf Morphology in Grounding-Line Stability: A Numerical Approach.
The early Pliocene, approximately 5 to 3 million years ago, is depicted by mean global temperatures slightly warmer than today. Bart and Iwai (in prep.), utilize seismic and biostratigraphic evidence to hypothesize that the Antarctic Peninsula's outer continental shelf overdeepened during a transitory period of increased glacial erosion in the early Pliocene. They attribute the enhanced erosion of the continental shelf to a period of regional warming on the peninsula's pacific margin that produced an associated increased flux of snow and ice. Salty, relatively warm, deep water is known to encroach onto the Antarctic continental shelf and melt the underside of ice shelves. The occurrence of this phenomenon emphasizes the susceptibility of subaqueous grounding-lines, ice shelves and outlet glaciers with marine terminations to fluctuations in ocean temperature. Assuming the early Pliocene overdeepening is a continental-wide phenomenon, this relatively recent modification of the Antarctic continental shelf may have greatly altered the way in which the marine terminus of the Antarctic Ice Sheet interacts with global thermohaline circulation. The present foredeepened continental shelf presumably permits a larger influx of relatively warm circumpolar deep water hindering major advancement of the ice sheet. Preliminary numerical modeling experiments which manipulate ocean temperature and grounding-line depth on the Ross Sea Basin continental shelf have confirmed the sensitivity of the grounding line to ocean temperature and water depth. Deeper grounding lines exhibited greater responses to changes in sea temperature. This ongoing study tests the sensitivity of an ice sheet on a particular continental shelf morphology (shallow and flat; deep and flat; shallow, gently basinward dipping; shallow, steeply basinward dipping; deep, steeply landward dipping) to a warming ocean with the following objectives:- (i) delineate the various transitions from assumed pre-Pliocene shelf (shallow and basinward dipping) to post-Pliocene shelf (deep and landward dipping) and examine the evolving relationship of ice and deepwater by comparing rates and magnitudes of grounding-line retreat. (ii) quantitatively determine ice sheet stability (e.g. at 'X' water temperature on 'Y' shelf catastrophic retreat occurs). (iii) determine the amount of sea-level fall/rise by analyzing the ice volume changes as the ice sheet waxes/wanes to temperature variations.
C11D-0533
Soil Processes on Delta and Till Deposits in Taylor Dry Valley, Antarctica
During the Last Glacial Maximum the West Antarctic Ice Sheet expanded across McMurdo Sound and dammed Taylor Valley. It is speculated that this ice dam caused the formation of proglacial Lake Washburn, which reached an elevation of approximately 300 meters above sea level. Inundation by the lake would leach Taylor Valley soils of soluble salts, and present soil salt contents should reflect accumulation since the last flooding event. Interpreting soil soluble salt content requires an understanding of active soil forming processes in the Dry Valleys and how these processes affect salt distribution and accumulation. Our findings indicate that the soil salt contents are highly correlated with fines brought in by eolian processes. The desert pavement surface traps windblown clay- and silt-size particles, along with salts, to accumulate at the soil surface. Both the fines and salts appear to be transported to depth in the soil profile by periodic wetting events and cryoturbation, thereby resulting in the accumulation of fines and salts in the upper soil horizons. Soil salt contents parallel the % volume of fines in the upper horizons of soils, and soil salt contents increase with decreasing elevation, a trend that mirrors higher fluxes of eolian material at lower elevations due to wind speed, availability of material for transport, and settling. These results indicate that variables affecting eolian deposition, such as topography and wind patterns, influence the rate of salt accumulation in Taylor Valley soils. This makes it difficult to determine a detailed lake level history of Lake Washburn by looking at soil salt contents. A sharp increase is observed in the soil soluble salt content above 300 meters above sea level, but this may simply reflect an older till deposit rather than a previous lake shoreline. Further consideration of sample locations and controls on eolian input is in progress to refine the relative age estimates.
C11D-0534
ERT of seasonal changes in steep permafrost rocks with laboratory-calibrated temperature-resistivity behavior, an incorporated resistance error model, and temperature validation (Zugspitze, German/Austrian Alps)
Rockwall instability at the Zugspitze North Face is evident due to the 3.5(±0.5)*108 m3 rockslide that occurred in 3700 B.P from the permafrost-affected Zugspitze (2962 m a.s.l.) North Face and was interpreted as a delayed permafrost response to the Holocene Climatic Optimum (Gude and Barsch, 2005; Jerz and Poschinger, 1995). Thawing of rock permafrost is a key promoting and triggering factor for rock instability, rock creep, and rockfalls. Rock instability in permafrost-affected bedrock is governed by shear stress, hydrostatic pressures, and friction along one or several planes of weaknesses in the rock mass. Friction along planes of weaknesses (without fine-grained cleft fillings) is susceptible to subzero temperature changes due to alterations in intact rock mass strength and ice-mechanical properties along planes of weaknesses. We argue that rock permafrost geophysical surveys have the potential to provide spatially resolved information on subzero temperatures that could serve as highly-important input for stability models. Here we present the first monthly ERT (electrical resistance tomography) results in steep permafrost rocks over the whole summer thaw period. Temperature-resistivity behavior of limestone from the study site was measured in the laboratory during repeated cycles of freezing at 0.2-0.4°C increments and exhibited a two-fold linear relationship divided by a freezing point at about -0.5°C. Field measurements were taken from February to October 2007 in a gallery close to the Zugspitze North Face at 2800 m a.s.l. and display thawing and refreezing behavior up to 30 m depth from the north face. We applied high-resolution, time-lapse 2D ERT with 140 electrodes and ca. 2000 resistance measurements per time frame. Local topographic settings and electrode positions were accommodated in an 8600 finite-element mesh with three different resolutions and varying boundary conditions according to the field setup. Resistivity images were computed using a smoothness-constrained inversion routine. In the inversion, we applied an empirically-derived resistance error model based on normal and reciprocal measurements to obtain adequately and consistently fitted resistivity images at the different time steps. Five temperature loggers were installed at 2.5m, 5m, 10m, 15m, and 20m from the north face to validate whether imaged resistivity values match laboratory-calibrated temperature-resistivity paths. Here we discuss whether high-resolution ERT in homogeneous rock walls with (i) laboratory-calibrated temperature-resistivity behavior, (ii) an incorporated resistance error model to ensure optimum and consistent fitting of the data, and (iii) field validation with temperature logger data provides reliable, spatially- resolved subzero temperature-information in permafrost rocks.
C11D-0535
Time-series Analysis of Icequakes and Ice Motion, Bering Glacier, AK
We have acquired a year-long GPS record and a 16-month seismic data set from an array of co-located seismic and GPS stations which were established in spring 2007 on Bering Glacier, North America's largest mountain glacier (located in Southcentral AK). Dual-frequency GPS data was recorded continuously at five stations on the glacier (15s recording interval). Four stations were co-located with short-period L-22 seismometers (sample frequency of 200Hz, corner frequency of 2Hz). A fifth L-22 seismic station was located at the center of the 4-km aperture array. An additional solitary GPS station was located about 20 km up glacier from the array. The GPS strain diamond and seismic array were installed roughly halfway between the equilibrium line and terminus. The year-long GPS motion record reveals ice velocities upwards of 3 m/day. The motion is smooth and steady (to temporal data resolution) and shows no evidence for sudden motion events. There is a slow and steady increase in velocity in early May (likely an annual spring speed up) that is followed by the onset of diurnal velocity fluctuations. An on-site temperature logger located at the snow-ice interface gives us the date that the sites become snow-free shown by an onset of diurnal temperature fluctuations. This date coincides with the diurnal onset in the motion record. Using several Short Term Averaging / Long Term Averaging (STA/LTA) triggered detectors, we have identified up to 500 icequakes per hour. We have produced a 16-month time-series analysis investigating the occurrence and signal strength of events which shows an onset of activity during spring speed up and a distinct shift to diurnal behavior by mid-summer. This pattern is reflected in the GPS motion record. The events can be grouped into two dominant discrete frequency spectra including low frequency emergent events and impulsive high frequency events. We attribute these groups to two separate mechanisms, namely fluid movement and brittle failure. There is a temporal variation in the relative abundance of the low and high frequency events. The frequency content, waveform, and particle motion of a number of located events has been examined to provided better insight to these mechanisms. The correlation between event occurrence and local weather data from a station near the terminus will be investigated to provide further explanation of fluid-induced sources.
C11D-0536
Velocity observations at Jakobshavn Isbrae 2006-2008
After nearly doubling its speed concurrently with substantial thinning in the late 1990s, Jakobshavn Isbrae is currently one of the fastest flowing glaciers in Greenland. Since summer 2006 we have used GPS to record velocities at four points along the main trunk of the glacier and at several points on the adjacent ice sheet. Velocities in the lower 20-30 km have mostly remained steady at post-speedup values. But the speed up continues to propagate to the upper reaches of the glacier. Measurements from the adjacent ice sheet indicate that since the mid-1980's, the velocity magnitude has increased and that flow has become more convergent towards the main channel. Velocity variations on diurnal and multiday timescales were also observed, but never exceeded 10% of the background velocity. In summer 2007 a speed-up event that may have been caused by a hydraulic event led to an eventual slow down of the ice stream. The slow down was accompanied by a 0.1 m drop in surface elevation, possibly due to the evacuation of basal water and increased efficiency of the subglacial drainage system. In both 2007 and 2008, diurnal velocity variations started appearing after a pronounced speed-up event. Diurnal velocity variations on the ice stream tend to be smaller (in a relative sense) than the variations on the adjacent ice sheet and they tend to lag those variations by 1-2 hours. This is an indication that runoff driven velocity variations propagate from the ice sheet down into the ice stream, while it has been shown that the seasonal velocity variation of the main trunk is strongly coupled to the position of the terminus.
C11D-0537
Classification of Cryostructures of Basal Glacier Ice Using Tomodensitometry
In permafrost areas, the extent of former glaciations can be inferred from the occurrence of buried glacier ice (Murton et al. 2005). Buried basal glacier ice can only be distinguished from other type of massive ice if the properties and structure of contemporary basal ice are well-known. Several classifications of basal ice have been developed based on the geological facies analysis (Lawson 1979, Knight 1994, Hubbard and Sharp 1995). However, none of them have addressed basal ice descriptions in terms of cryostructures. We used conventional field descriptions and microcomputed tomodensitometry to evaluate structure and properties of sediments including basal ice. The scan images were processed using a thresholding technique which assigns gray scale values to the components of basal ice based on their density. These values were used to determine the volumetric content of each component (ice, sediment, gas). We have used the Mathworks MATLAB R2007a software to produce three-dimensional images of basal ice cryostructures. The main advantage of these is to obtain in a non-destructive manner precise average ice and sediment contents for a given volume of a particular cryostructure. These models allow for a detailed 3D visualization of the cryostructures which is a powerful tool to study their architecture and the geometric relations between ice and sediments. The technical details of these operations have been presented by Dillon et al. (2008). Our cryostructures classification of basal glacier ice and sediments includes seven types of cryostructures observed in the basal ice of the Matanuska Glacier: (1) Porous; (2) Crustal; (3) Reticulate; (4) Lenticular; (5) Layered; (6) Suspended (includes four sub-types: Suspended dispersed, Micro-suspended, Suspended- intergranular, Suspended inter-crystalline); (7) Massive ice. The classification is illustrated by detailed photographs, sketches and bi-dimensional (2D) scan images of the various cryostructures along with their typical volumetric water contents and sedimentological properties. The 2D scan images are useful for permafrost studies because they can be compared to traditional detailed exposure descriptions.
C11D-0538
Greenland ice sheet outlet glacier front changes: comparison of year 2008 with past years
NASA's Moderate Resolution Imaging Spectroradiometer (MODIS) imagery are used to calculate inter-annual,
end of summer, glacier front area changes at 10 major Greenland ice sheet outlets over the 2000-2008
period. To put the recent 8 end of summer net annual changes into a longer perspective, glacier front
position information from the past century are also incorporated.
The largest MODIS-era area changes are losses/retreats; found at the relatively large Petermann Gletscher,
Zachariae Isstrom, and Jakobshavn Isbrae. The 2007-2008 net ice area losses were 63.4 sq. km, 21.5 sq.
km, and 10.9 sq. km, respectively. Of the 10 largest Greenland glaciers surveyed, the total net cumulative
area change from end of summer 2000 to 2008 is -536.6 sq km, that is, an area loss equivalent with 6.1
times the area of Manhattan Is. (87.5 sq km) in New York, USA.
Ice front advances are evident in 2008; also at relatively large and productive (in terms of ice discharge)
glaciers of Helheim (5.7 sq km), Store Gletscher (4.9 sq km), and Kangerdlugssuaq (3.4 sq km).
The largest retreat in the 2000-2008 period was 54.2 sq km at Jakobshavn Isbrae between 2002 and 2003;
associated with a floating tongue disintegration following a retreat that began in 2001 and has been
associated with thinning until floatation is reached; followed by irreversible collapse.
The Zachariae Isstrom pro-glacial floating ice shelf loss in 2008 appears to be part of an average ~20 sq km
per year disintegration trend; with the exception of the year 2006 (6.2 sq km) advance. If the Zachariae
Isstrom retreat continues, we are concerned the largest ice sheet ice stream that empties into Zachariae
Isstrom will accelerate, the ice stream front freed of damming back stress, increasing the ice sheet mass
budget deficit in ways that are poorly understood and could be surprisingly large.
By approximating the width of the surveyed glacier frontal zones, we determine and present effective glacier
normalized length (L') changes that also will be presented at the meeting. The narrow Ingia Isbrae advanced
in L' the most in 2006-2007 by 9.2 km. Jakobshavn decreased in L' the most in 2002-2003 by 8.0 km.
Petermann decreased in length the most in 2000-2001, that is, L' = -5.3 km and again by L' = -3.9 km in
2007-2008. Helheim Gl. retreated in 2004-2005 by L' = -4.6 km and advanced 2005-2006 by L' = 4.4 km.
The 10 glacier average L' change from end of summer 2000 end of summer 2008 was 0.6 km.
Results from a growing list of glaciers will be presented.
We attempt to interpret the observed glacier changes using glaciological theory and regional climate
observations.
http://bprc.osu.edu/MODIS/
C11D-0539
Underground thermo-erosion of ice wedges: numerical simulation of tunnel freeze- back
On Bylot Island in the Eastern Canadian Arctic archipelago, Fortier et al. (2007) observed and characterized the formation and development of tunnels initiated by the process of underground thermo-erosion of ice wedges networks. These tunnels often collapsed during the course of one or two summers and developed into gullies. However, observations of such tunnels in permafrost exposures indicate that they can be preserved in the permafrost record. The objective of this study is to estimate the freeze-back time of tunnels filled with water and slurry in cold and warm permafrost conditions. Ultimately, the goal is to evaluate time the tunnels remain "open" for groundwater flow. We used numerical thermal modeling to conduct simple simulations of the conductive heat transfer during freeze-back of the tunnels. The thermal analyses were performed using the GeoslopeTM unsteady finite element heat conduction model TEMP/W. We used Bylot Island, Nunavut, Canada (Mean air temperature around -15 C) as a cold permafrost study case and Beaver Creek, Yukon Territory, Canada (Mean annual air temperature around -5.5C) as a warm permafrost study case. The air temperature was converted to ground surface temperature by the n-factor method. Zero heat flux was applied at the vertical and bottom boundaries due to the permafrost which is several tens to hundreds of meters thick. Based on previous studies, we simulated tunnels partly cut in ice-wedges and in the adjacent permafrost. The syngenetic permafrost of the case studies was assumed to be fully saturated with 110% gravimetric water content. The geometry of the tunnels was based on field measurements on Bylot Island. We considered three scenarios for the slurry filling the tunnels: 1) 100% water; 2) fully saturated sand with 30% gravimetric water content; and 3) an air layer at the top of the tunnel with water and saturated sands partly filling the bottom of the tunnel. We used three water/slurry temperatures: 1) 0.5C which simulates the water temperature of early snowmelt run-off, a period of active underground thermo- erosion; 2) 2C corresponding to the water temperature of run-off over a partly frozen active layer, which is typical for early summer undergrouns thermo-erosion ; 3) 5C and 15C corresponding to the water temperature lakes at the end of August on Bylot Island and at Beaver Creek respectively. This scenario simulates underground thermo-erosion due to lake drainage through ice wedges. The volumetric heat capacity of the ground was calculated as the sum of the volumetric heat capacities of the three phases multiplied by their volumetric fractions. The thermal conductivity of the permafrost and slurry was calculated by the geometric mean model. The apparent heat capacity method was applied to deal with latent heat generation. The initial permafrost temperature was -12C for the cold permafrost case and -2.5C for the warm permafrost case. The results suggest that thermo-erosion of ice wedges creates underground water flow paths that remain unfrozen for significant period of time, particularly in warm permafrost conditions. Tunnels entirely filled with water took about 1.5 to 2 times longer to freeze back. The presence of air pockets in the tunnel significantly delayed the freeze back time. In the cold permafrost cases, the tunnels froze back in about 3 to 6 months whereas they required between 8 months to about 3 years to freeze back in the warm permafrost cases.
C11D-0540
Evidence of Seismic Anisotropy Detected on Jakobshavn Glacier
Reflected seismic wave slowness variability as a function of angle of incidence can be used to detect anisotropy in ice crystal orientation. Multi-offset seismic reflection data were acquired on Jakobshavn Glacier in Greenland to investigate ice basal conditions. Data were recorded from a stationary array of eight 100 Hz vertical geophones while 0.5 kg explosive sources buried 10 m below surface were deployed at offsets ranging from -1760 to +1600 m. Source spacing was 160 m yielding 22 variable angle recordings. The ice- bed interface was imaged at approximately 1.85 km depth and multiple internal ice layers at 1.5 to 1.7 km depth. Incidence angles ranged from -24 to +20 degrees. Analysis of the measured ray velocities reveals slowness change with angle of incidence. This variation in slowness is strong evidence of change of ice crystal orientation in the ice column near the base of Jakobshavn glacier. These observations are consistent with laterally extensive complex ice fabric development reported over the same region of Jakobshavn Glacier.
C11D-0541
Geophysical investigation of permafrost, Shishmaref Alaska
A ground-penetrating radar (GPR) investigation was done in April 2008 to investigate permafrost and stratigraphy at Shishmaref, Alaska. Shishmaref is located on Sarichef Island, a barrier island on the coast of western Alaska, experiencing marine transgression and coastal erosion. The soils are composed of frozen fine sands containing permafrost. GPR data collected in April 2008 using 100-, 200-, and 400-MHz antennas show an intermittent strong reflector at approximately 50-200 ns, assumed to be the top of permafrost. The source of this reflector will be confirmed in September 2008 by drilling and soil sampling. Resistivity and electromagnetic data will also be collected to determine active layer thickness and to detect changes in permafrost and soil properties across the island. The results of this study will be used for assessing shoreline stability at Shishmaref.
C11D-0542
Debris-Rich Basal Ice Layer Effects on Polar Glacier Dynamics
Many cold-based glaciers have a debris-rich basal ice layer, which may deform more easily than the overlying clean ice. We present results from a study exploring the importance of these debris-rich basal ice layers to the flow dynamics in the terminus region of Taylor Glacier, a cold-based outlet glacier of the East Antarctic Ice Sheet. Observed surface velocities are 20 times greater than that predicted by laminar flow for clean, homogenous, isotropic ice at the field site. We hypothesize that a soft basal-ice layer near the bed is accommodating the large strain rates. A two-layer two-dimensional flowband model is used to investigate the relative contributions from the debris- rich basal ice and the overlying clean ice to the ice flow rates. The model unknowns are the softness parameters for the clean (EC) and debris-rich ice (EB), and the thickness of the basal layer, which is spatially variable along the flowband. We constrain the model inputs and outputs using 2 years of field observations and previous tunnel studies into the glacier margin (Samyn and others, 2005), which include ground penetrating radar, ablation, ice temperature, basal ice thickness, and surface velocity measurements. The thickness of the debris-rich basal ice is solved for in the model for each chosen softness parameter combination. Analysis of two shallow ice cores in the clean ice indicate that the ice crystals are small and have a distinct preferred orientation; they have a strong, vertical, single-maximum fabric. The clean ice fabric data provides the best ice rheology constraint, and allows four plausible softness parameter combinations in the range 6 ≤ EC ≤ 9 and 38 ≤ EB ≤ 50 to be identified. The relative contribution of the basal ice to the total flow rates in the terminus region of Taylor for these four solutions ranges from 63-76%. This weak layer, therefore, significantly affects the glacier dynamics and suggests the importance of incorporating rheologically distinct layers into models of polar glacier dynamics. Samyn, D., S.J. Fitzsimons and R.D. Lorrain. 2005a. Strain-induced phase changes within cold basal ice from Taylor Glacier, Antarctica, indicated by textural and gas analyses. Journal of Glaciology, 51(175). 611-619.
C11D-0543
Global Seismic Event Detection Using Surface Waves: 15 Possible Antarctic Glacial Sliding Events
To identify overlooked or anomalous seismic events not listed in standard catalogs, we have developed an algorithm to detect and locate global seismic events using intermediate-period (35-70s) surface waves. We apply our method to continuous vertical-component seismograms from the global seismic networks as archived in the IRIS UV FARM database from 1997 to 2007. We first bandpass filter the seismograms, apply automatic gain control, and compute envelope functions. We then examine 1654 target event locations defined at 5 degree intervals and stack the seismogram envelopes along the predicted Rayleigh-wave travel times. The resulting function has spatial and temporal peaks that indicate possible seismic events. We visually check these peaks using a graphical user interface to eliminate artifacts and assign an overall reliability grade (A, B or C) to the new events. We detect 78% of events in the Global Centroid Moment Tensor (CMT) catalog. However, we also find 840 new events not listed in the PDE, ISC and REB catalogs. Many of these new events were previously identified by Ekstrom (2006) using a different Rayleigh-wave detection scheme. Most of these new events are located along oceanic ridges and transform faults. Some new events can be associated with volcanic eruptions such as the 2000 Miyakejima sequence near Japan and others with apparent glacial sliding events in Greenland (Ekstrom et al., 2003). We focus our attention on 15 events detected from near the Antarctic coastline and relocate them using a cross-correlation approach. The events occur in 3 groups which are well-separated from areas of cataloged earthquake activity. We speculate that these are iceberg calving and/or glacial sliding events, and hope to test this by inverting for their source mechanisms and examining remote sensing data from their source regions.
C11D-0544
Glaciation of the Coastal Plain of Northern Alaska
Our 15 years of studies of permafrost soils on the coastal plain of northern Alaska show that it was affected by a continental ice sheet during the last glacial maximum. Evidence for this includes: occurrence of buried glacial basal ice at Barter Island; widespread sandy diamicton from Demarcation Bay to Barrow of late Pleistocene age; orientation of surficial deposits; poorly integrated drainage and gentle ridge and swale topography; the continuity of glacial-related deposits from the coast to the Brooks Foothills; and perennially frozen sediments unlike those of unglaciated Arctic regions. We documented a 10-m-high exposure ~1 km long at Barter Island that had abundant basal glacier ice with large-scale deformation structures, complex ice-contact deposits, and highly deformed bedded silt, sand, and gravel inclusions within the basal ice. Similar ice structures were observed at Prudhoe Bay and Cape Halkett. The glacial till is highly unusual in that it is comprised of massive, non-fossiliferous, brackish, slightly pebbly loamy sand with occasional gravel to cobble-sized clasts. In most areas the till is only 25 m thick, although at Barter Island the till was up to 10- m thick. Gravel particles, which comprise 15% of the deposits, usually are 0.52 cm long, mostly durable chert, highly polished, and frequently cracked off at one end, with the broken face faceted and polished. We believe the material mostly originated from marine deposits on the continental shelf, although rocks of Canadian provenance also occur. Prevalent, large (15 m) deformation features of discontinuous yellow oxidized and gray reduced sediment suggest deformation of sediment during collapse of the ice sheet. The sandy till is found along most of the Beaufort coast with the exception of deltas and lagoons and is found inland as much as 80 km. The sandy till is easily eroded, causing the morainal margin to be indistinct and the topography subdued. Previous thermoluminescence dating by Rawlinson (1993) of the sediment indicates it was deposited 2030 ka BP, and our radiocarbon dating of organic matter at the surface of the sediment indicates that deglaciation and surface stabilization occurred 1017 ka BP. The glacial deposit is synchronous with the extensive late Pleistocene sand sea in the western Beaufort coastal plain that dates to 9.611.4 ka BP (Carter 1981) and the large loess belt along the lower Brooks Foothills that was deposited 2030 ka BP. The sand sea and loess belt probably were deposited in unglaciated areas by catabatic winds associated with the ice sheet. The presence of an ice sheet that covered the Beaufort Sea and adjacent coast helps explain the lack of typical ice-rich Pleistocene permafrost (yedoma) along the Beaufort coast, the reported lack of permafrost on the outer continental shelf of the Beaufort Sea, and the lack of thaw-lake depressions on the continental shelf. In contrast, these features are abundant on unglaciated portions of the continental shelf in the Russian Arctic and northern Seward Peninsula and adjacent Chukchi Sea. This evidence from terrestrial deposits is consistent with recent evidence of glacial scouring at depths up to 700 m along the continental margin of the Beaufort Sea documented by Engels et al. (2007). We name this glaciation the Kuukpik Glaciation and the deposit the Kuukpik member of the Gubik Formation.
C11D-0545
Non-linear visco-elastic effects of oceanic tides on ice stream motion
Oceanic tides are known to significantly affect tidal motion. A particular example is Rutford Ice Stream, West Antarctica, where tidal modulation changes the horizontal velocity by up to about 20% from its mean value. One of the peculiarities of the tides on Rutford is that the strongest modulation is at the Msf tidal frequency, or 14.76 days. This is despite the fact that the Msf tidal amplitude is almost absent in the vertical oceanic tides, and much smaller than the semi-diurnal and the diurnal tidal amplitudes. Previously, a simple conceptual model has been proposed suggesting that the fortnightly tidal motion on Rutford Ice Stream arises through a strongly non-linear interaction between the main tidal components. According to this mode, non-linear viscous till rheology is the source of the non-linearity in the system. This idea is tested using a visco-elastic flow model. Both the ice and the till is modelled as a non-linear visco-elastic medium. All terms of the momentum equations are kept in the momentum balance. In full agreement with the much simpler conceptual model it is found that the fortnightly tidal motion can be generated through strong non-linear interaction between two main semi-diurnal oceanic tidal components for non-linear till rheology. The implication is that the rheology of till underneath Rutford Ice Stream (as experienced by the overlying ice), must be non-linear and viscous.
C11D-0546
Pine Island Glacier Inland Flow Increases: Causes and Longevity From new Ground- Based GPS and Radar Results
Pine Island Glacier has one of the highest outflows of any ice stream in the world. Over the last three decades it has changed significantly. Satellite measurements have revealed acceleration in the lower 100 km of the ice stream along with thinning that stretches inland over the whole drainage basin. For the first major ground-based study of Pine Island Glacier a team spent the austral summers of 2006 to 2007 and 2007 to 2008 on the ice stream. Measurements utilizing various techniques were made between 55 km and 200 km upstream from the grounding line, including along four of the main tributaries. These techniques included seismic reflection profiling, passive seismic surveys, deep and shallow radar profiles, GPS surveys and shallow ice coring. Here we focus mainly on the results of the GPS and radar surveys along the central flowline of the glacier. The GPS measurements demonstrate that the current levels of acceleration are high compared to those over the previous three decades. This confirms recent satellite measurements that the downstream acceleration during 2007 alone reached 7 %. The GPS measurements also showed that the acceleration at distances of 111 and 171 km upstream from the grounding line were 5 and 4 % respectively in 2007. This suggests that the acceleration reaches far inland, probably in excess of 200 km. The rates of thinning during each season were measured by GPS and found to be greater than past estimates from satellite measurements at 3.5, 2.5 and 1.0 m per annum 55, 111 and 171 km from the grounding line respectively. There is a strong correlation between the rate of change in slope, hence driving stress, and the acceleration. The radar measurements show accumulation features that can be tracked from the surface by simple velocity models. At the upstream end of the surveys these features can be tracked back in time, up to 200 years B.P. These features are revealing aspects of the flow history of Pine Island Glacier and the extent to which the ice stream flow could have changed in the last two centuries. This puts the current high accelerations into context, which may enhance predictions of the future stability of Pine Island Glacier.
C11D-0547
Thermodynamics of Polythermal Glaciers and Ice Sheets
Polythermal conditions in ice sheets are found in both Greenland and Antarctica. Current ice sheet models can be divided into two groups: (1) "Cold-ice method" ice sheet models; and (2) truly polythermal ice sheet models. The only member of the second group uses a front-tracking method to handle the internal free cold- temperate transition surface. Front-tracking methods require either deforming grids or transformed coordinate systems and therefore are somewhat cumbersome to implement. Enthalpy formulations, on the other hand, are known for their ease of implementation. Here we apply an enthalpy formulation to solve the enthalpy equation in both cold and temperate ice, thus avoiding front-tracking. We assume enthalpy to be the fundamental quantity which uniquely determines temperature below and above the pressure melting point and water content at the pressure melting point. The enthalpy formulation is more physical than the "cold- ice method" because the effects of the phase change at the pressure melting point are adequately accounted for. Results are presented to demonstrate that the enthalpy formulation is able to simulate the relevant thermal structures found in polythermal glaciers and ice sheets. The proposed enthalpy formulation is a suitable replacement for the temperature equation used in "cold-ice method" ice sheet models and renders these models polythermal. Substantial improvement is expected if applied to small ice caps and ice sheets, where significant parts of temperate ice exist.
C11D-0548
Thermal State of Permafrost in the Northern Yakutya: modern state and dynamic during the last 50 years
One of the fundamental tasks of permafrost science is investigation of permafrost responce on the climate changes. Geothermal observation in the boreholes on the Yakutian coastal lowlands carried out since 80th years of the last century. Existant network allows us to estimate both spatial and temporal changes of the geothermal field. Investigated region covers the area from the Lena delta to Kolyma and characterized by cold continental climate (mean annual air temperature -13.5 to -14°C) and continuous permafrost distribution. Active layer 0.3-1 m. Recently, the network for continuous geothermal observation was established. It includes 10 boreholes located on the different latitudes, natural zones and landscapes. Modern thermal state of permafrost in this region is following: Mean annual ground temperature on the top of rest of Late Pleistocene accumulative planes varies in the range from -12.3°С on the latitude 72°50' north to -9.9°С on the latitude 69°30' north. Latitudinal zonality here is about 1°С on the degree of latitude. Within the Alas depression mean annual ground temperature is a little bit warmer (-10°С on the 71°40' north and - 7°С on the 68°50' north). Comparison of modern observations and published data shows that most significant changes of the geothermal field take place on the Kolyma lowland. Since the 1980th permafrost temperature increase here on 1.5-2°C. At the same time thermal state of permafrost in the western part of region is more stable. Also there are some sites where modern landscape changes (vegetation succession) leads to the stabilization of permafrost temperature. Current research supported by NSF grants ARC-0520578 and ARC-0632400.
C11D-0549
Temperate Ice Depth-Sounding Radar
Glaciers in several parts of the world are reported to be retreating and thinning rapidly over the last decade. Radar instruments can be used to provide a wealth of information regarding the internal and basal conditions of large and small ice masses. These instruments typically operate in the VHF and UHF regions of the electromagnetic spectrum. For temperate-ice sounding, however, the high water content produces scattering and attenuation in propagating radar waves at VHF and UHF frequencies, which significantly reduce the penetration depths. Radars operating in the HF band are better suited for systematic surveys of the thickness and sub-glacial topography of temperate-ice regions. We are developing a dual-frequency Temperate-Ice-Depth Sounding Radar (TIDSoR) that can penetrate through water pockets, thus providing more accurate measurements of temperate ice properties such as thickness and basal conditions. The radar is a light-weight, low power consumption portable system for surface-based observations in mountainous terrain or aerial surveys. TIDSoR operates at two different center frequencies: 7.7 MHz and 14 MHz, with a maximum output peak power of 20 W. The transmit waveform is a digitally generated linear frequency-modulated chirp with 1 MHz bandwidth. The radar can be installed on aircrafts such as the CReSIS UAV [1], DCH-6 (Twin Otter), or P-3 Orion for aerial surveys, where it could be supported by the airplane power system. For surface based experiments, TIDSoR can operate in a backpack configuration powered by a compact battery system. The system can also be installed on a sled towed by a motorized vehicle, in which case the power supply can be replaced by a diesel generator. The radar consists of three functional blocks: the digital section, the radio-frequency (RF) section, and the antenna, and is designed to weigh less than 2 kg, excluding the power supply. The digital section generates the transmit waveforms as well as timing and control signals. It also digitizes the output signal from the receiver and stores the data in binary format using a portable computer. The RF-section consists of a high- power transmitter and a low-noise receiver with digitally controlled variable gain. The antenna is time-shared between the transmitter and receiver by means of a transmit/receive (T/R) switch. In regards to the antenna, we have made a survey study of various electrically small antennas (ESA) to choose the most suitable radiating structure for this application. Among the different alternatives that provide a good trade-off between electrical performance and small size, we have adopted an ESA dipole configuration for airborne platforms and a half-wavelength radiator for the surface-based version. The airborne antenna solution is given after studying the geometry of the aerial vehicle and its fuselage contribution to the antenna radiation pattern. Dipoles are made of 11.6 mm diameter cables (AWG 0000) or printed patches embedded into the aircraft fuselage, wings, or both. The system is currently being integrated and tested. TIDSoR is expected to be deployed during the spring 2008 either in Alaska or Greenland for surface based observations. In this paper, we will discuss our design considerations and current progress towards the development of this radar system. [1] Center for Remote Sensing of Ice Sheets (Cresis), Sept 2008, [Online]. Available: http://www.cresis.ku.edu