Derivation of spatial distribution of snow precipitation using interferometric SAR technique

Abstract

Accurate knowledge of spatial variation of snow precipitation in an area is important because it provides information of an initial condition that has significant impact on subsequent processes. However, such information is difficult to obtain because (1) snow gauges are sparsely distributed, and (2) snow gauges are notoriously inaccurate for snow measurement. The sparsity of snow gauges is especially severe in Alaska because of its remoteness. We applied satellite interferometric SAR (InSAR) technique to investigate spatial variation of snow precipitation on the Alaskan North Slope. SAR images acquired during the ERS-1 second ice phase in early 1994 were processed to form interferograms at a study site near Franklin Bluffs. Because new snow accumulation would increase radar path length, a pattern of spatial variation in snow precipitation might be derived from the phase pattern seen on the interferograms. A new snow precipitation of 7 cm on February 5, 1994 was reported for the area. An interferogram constructed from ERS-1 SAR images acquired on February 4 and 7 spanned this snow event. The terrain effects were removed using another interferogram that was constructed from a pair of SAR images acquired before the snow event. The resulting interferometric phase pattern is interpreted as a thickness variation of 8 cm in the new snow precipitation.