Investigation of the spatial variations in synthetic aperture radar backscatter in western Dronning Maud Land, Antarctica

Abstract

C-band SAR observations show that backscatter varies significantly across small scales (tens of kilometers) in western Dronning Maud Land. Generally, backscatter was found to diminish with altitude reflecting lower accumulation and reduced ice inclusions in the firn of the percolation zone at higher elevations. Reference to (incomplete) mass balance data suggests an anticorrelation between backscatter and net balance, although more data are needed to confirm the trend. Even within the percolation zone, areas of low backscatter (< −12 dB) exist. These are uncorrelated with altitude above sea level. Using a Rayleigh backscatter model, we show that backscatter over such regions may represent differences in grain sizes. The application of a buried-layers model did not accurately estimate backscatter from these regions. We suggest that these regions occupy exposed positions subject to increased wind sublimation which, in turn, results in smaller grain sizes in the firn and therefore reduced backscatter. The Radarsat Antarctica Mapping Mission SAR mosaic indicates such regions occur in Coates Land, near the edge of the Antarctic Plateau, and on exposed promontories elsewhere in East Antarctica. More information regarding the structure of the firn and the mass balance of the region is needed before we can definitively explain controls on backscatter and understand the climatology of the low backscatter zones.

C-Band SAR Imagery for Snow-Cover Monitoring at Treeline, Churchill, Manitoba, Canada

“…Many models developed over the last 15 years have been validated in various environments, including mountainous terrain and glaciers [2][3][4], sea ice and ice sheets [5][6][7][8], agricultural fields [9,10], boreal forest [11][12][13][14][15][16], and forest-tundra ecotones [17]. The ability of SAR to discriminate snow-free from wet or dry snow-covered surfaces has been investigated substantially more than its potential for estimating the snow mass of dry snow cover; this is particularly true near the Arctic treeline.…”