C43A-0218 1340h
A revised inventory of Antarctic subglacial lakes
The locations and details of 145 Antarctic subglacial lakes are presented. Subglacial lakes can be identified on radio-echo sounding (RES) data from the recognition of three characteristics. First, the radio reflections from subglacial lakes are strong and typically 10-20 dB greater than from the ice-bedrock boundary. Second, the echoes have constant strength along the record track, which is indicative of an interface that is smooth on the scale of the radio wavelength. Third the reflections from a lake are very flat compared with the surrounding topography and have a slope which is around ten times, and in opposite direction to, the ice surface slope (which is required if the lake is in hydrostatic equilibrium). The revised inventory is based on a former catalogue of lake-type features (Siegert et al., 1996), which has been subsequently reanalysed, and three additional datasets. The first is from Italian RES of the Dome C region of East Antarctica, from which fourteen new lakes are identified. These data also show that, in a number of occasions, multiple lake-type reflectors thought previously to be individual lakes are in fact reflections from the same relatively large lake. This reduces the former total of lake-type reflectors by six, but also adds a significant level of information to these particular lakes. The second dataset is from a Russian RES survey of the Dome A and Dome F regions of East Antarctica, which provides evidence of eighteen new lakes and extends the coverage of the inventory considerably. The third dataset comprises three airborne RES surveys undertaken by the US in East Antarctica over the last five years, from which forty three new lakes have been identified.
http://www.ggy.bris.ac.uk/ellsworth
C43A-0219 1340h
Preliminary Results Using C-band Ground Penetrating Radar to Determine Backscatter Sources Within Glaciers
Ground Penetrating Radar (GPR) operating at 5.3GHz (C-band), the same frequency as the Advanced Synthetic Aperture Radar (ASAR), is used to resolve the depth resolution not retrievable from the ASAR data. We present results from the first field season, April 2004. Two GPR instruments, one at 5.3GHz (C-Band) and the other at 800MHz, were used to take profiles along the centre line of Kongsvegen, Svalbard. ASAR images were acquired for this same time period. In addition cores and snow pits were taken along the centre line. Data from the two GPR instruments is very comparable. Both see the same internal layering and transitions in glacial facies. In order to make a comparison between the C-band GPR and the ASAR data, the radar equation is used to convert the GPR signal to backscatter cross section as a function of depth. Plotting this value reveals the backscatter contribution over depth. The depth at which most backscattering occurs varies up glacier. Integration of the sigma value over depth gives the single backscatter coefficient value obtained by ASAR. We see that the ASAR signal is in fact the result of all the backscatter sources identified with the C-band GPR. The core and snow pit data will provide the means by which to quantify the physical nature of the backscatters seen in the GPR data. Next year a fully polarimetric GPR will be used allowing greater flexibility and a more accurate comparison with the ASAR data.
C43A-0220 1340h
GPS and GPR Profiles of Snow Megadunes in East Antarctica
Global Positioning System (GPS) and ground-penetrating radar (GPR) profile data were acquired over a series of snow megadunes about 400 km southeasst of Vostok Station (80.78 S, 124.5 E) in November 2002 and January 2004. GPS profiles (110 km total) acquired in the mean wind direction (bearing 225) show the topographic characteristics of the dunes: 3 to 7 meters relief, with short, steep windward faces and longer, shallower lee surfaces relative to regional along-wind slope. Regional absolute slope in this area is nearly perpendicular to wind flow, i.e. nearly parallel to dune crests. GPS-derived mean ice velocity is 4 +/- 0.2 m/yr at bearing 130, with little regional gradient. GPR profiles (total of 60 km) were acquired at two frequencies, 100 MHz and 250 Mhz, providing information on snow layering to about 80 meters. Megadune regions are composed of stacked sequences of fore-set, windward-accreted snow layers, 8 to 15 meters thick, separated by lee-side layer sections showing near-zero accumulation (and possibly some erosion). Buried lee faces of the dunes, characterized by mono-layer ice 'glazes' on the surface are represented by 1- to 4meter thick zones of reduced reflection amplitude. Dune sequence layering can be detected at up to 70 meters depth, implying that dunes have been present for several millennia. Comparison with a shallow ice core drilled at the Megadunes camp indicates that interpreted buried glaze/lee surfaces coincide with anomalous low density regions in the firn. A separate profile of a region showing a transition from megadune terrain to a non-undulating, low slope plateau surface shows that megadunes became lower in amplitude, and intervening glazed-surface regions showed more evidence of accumulation as slope decreased to near-zero. At depth (20 - 40 meters), dunes structures underlay the smooth plateau surface, possibly due to advection. Details of the shape of the accretionary section of the windward face of the dunes and the arrangement of stacked dune sequences at depth will be used to constrain models of dune formation.
C43A-0221 1340h
The Effect of Horizontal Stress Gradients on Layer Architecture
This paper expands upon previously known results to demonstrate that as the wavelength of bedrock obstacles changes, the flow routes around and over them change. This means that the response of isochronic layers to bedrock variability is wavelength dependent. The effects of basal topography and variations in basal sliding on layer architecture are investigated using finite element calculations and perturbation methods. Perturbation methods are valid for small amplitude bed features while being computationally undemanding, while finite element methods can deal with large amplitude perturbations, but are restricted to plane flows in this study.
C43A-0222 1340h
Shallow (400-MHz) and deep (3-MHz) radar used to image wind erosion features in East Antarctica
Two radar systems were used to record data simultaneously along the 2001 and 2002 US-ITASE traverses. A 400-MHz short-pulse radar recorded firn stratigraphy to depths up to 120 m while a 3-MHz radar recorded ice internal stratigraphy and bedrock topography to more than 3 km depth. Because the radars were designed for different purposes (shallow versus deep), the characteristics of each system limited regions of overlap to depths of about 100 m. Nevertheless, data from the two radars have been used to better understand the dynamics of particular sites. We report on a site of wind erosion found over a large bedrock mountain in East Antarctica (87$\deg$S, 108$\deg$W). By geolocating both radar data sets we are able to correlate the internal reflectors recorded by each system. The deep radar shows a large-scale unconformity over the basal mountain and subsequent deposition that extends to a depth of 600 m. We use the shallow radar data to show that currently active wind erosion has removed up to 30 m of snow and firn at this site. The erosion results from increased surface topography as the ice flows over the mountain. Using regional accumulation rates, we estimate the age of the exposed firn surface at less than 200 years. This estimate will be refined when data from nearby US-ITASE shallow ice cores become available.
C43A-0223 1340h
Radar Sounding Studies of Cracks at the Base of the Ross Ice Shelf, Antarctica
Ice shelf systems are important links between ice sheets and global climate change. Through iceberg calving and basal melting and freezing, these ice shelves are vital contributors of fresh water to the thermohaline circulation of the world's oceans. In addition, as ice shelves evolve in concert with ocean circulation, the back stress that they provide to their parent ice sheet has been shown to modulate its evolution. The distribution and character of cracks beneath these ice shelves are important indicators of the evolving state of stress associated both with iceberg calving and grounding line migration. We will present results from airborne radar sounding surveys of the basal crack system from two locations on the Ross Ice Shelf. The first of these is iceberg B15a, representing a largely intact sample of 140 km of the calving front of the central Ross Ice Shelf. The second is a transect across the grounding zone where the now stagnant Kamb ice stream (ice stream C) abuts Siple Dome. Radar returns observed from basal cracks at these two locations are combined with dielectric models of crack evolution to classify cracks as either insipient or healed. The distributions of these classes of cracks at each site are interpreted from the perspective of the hypothesized stress states for ice shelf calving and grounding line retreat.
C43A-0224 1340h
Detection of subglacial lakes in airborne radar sounding data from East Antarctica.
Airborne ice penetrating radar is an essential tool for the identification of subglacial lakes. With it, we can measure the ice thickness, the amplitude of the reflected signal from the base of the ice, the depth to isochronous surfaces and, with high quality GPS, the elevation of the ice surface. These four measurements allow us to calculate the reflection coefficient from the base of the ice, the hydrostatic head, the surface slope and basal temperature. A subglacial lake will be characterized by: a consistently high reflection coefficient from the base of the ice, a nearly flat hydraulic gradient at a relative minimum in the hydraulic potential, an exceptionally smooth ice surface, and an estimated basal temperature that is at or near the pressure melting point of ice. We have developed a computerized algorithm to identify concurrences of the above-mentioned criteria in the radar data sets for East Antarctica collected by the University of Texas (UT). This algorithm is henceforth referred to as the "lake detector". Regions which meet three or more of the above mentioned criteria are identified as subglacial lakes, contingent upon a visual inspection by the human operator. This lake detector has added over 40 lakes to the most recent inventory of subglacial lakes for Antarctica. In locations where the UT flight lines approach or intersect flight lines from other airborne radar surveys, there is generally good agreement between the "lake detector" lakes and lakes identified in these data sets. In locations where the "lake detector" fails to identify a lake which is present in another survey, the most common failing is the estimated basal temperature. However, in some regions where a bright, smooth basal reflector is shown to exist, the lake detector may be failing due to a persistent slope in the hydraulic gradient. The nature of these "frozen" and "sloping" lakes is an additional focus of this presentation.
C43A-0225 1340h
A Study of Radar Echo Sounding Response and Clutter Using Data From Jakobshavn, Greenland
Since outlet glacier responses to climate change may occur much more rapidly than those of the ice sheet interior, the behavior over time of the outlet glaciers is likely a key indicator to observe. Thus, a better understanding of the linkage between the outlet responses and the rest of the ice sheet may be desirable for refining the models that try to predict changes in ice sheet mass balance. Under the NASA PARCA program, the Remote Sensing Laboratory of the University of Kansas has collected extensive airborne radar depth sounding data over Greenland's outlet glaciers as well as its interior. The radar (150 MHz center frequency) has sounded the ice thickness (i.e. from surface and bedrock returns) of the vast majority of its flight lines. Around the coastal regions in a few locations, strong clutter can and does mask the bedrock topographical information. Portions of the radar data from the Jakobshavn glacier exhibit noticeable levels of clutter. This study examines the clutter in the radar data from the Jakobshavn region. We modeled the clutter by an effective surface electromagnetic (EM) model and using the available data. The results are presented for averaged Jakobshavn radar data for an effective rough surface and an effective smooth surface. This first-order approach produced reasonable agreement between the radar data and the EM model. Also, typical samples from the same radar from outside the clutter regions will be examined for comparison.