Automatic Ice Thickness Estimation from Polar Subsurface Radar Imagery
Mentor: Christopher Gifford
CReSIS

This work focuses on automating the tedious task of estimating ice thickness from airborne radar data acquired over Greenland and Antarctica. This process involves the identification and accurate selection of the ice sheet's surface location and interface between the ice sheet and underlying bedrock for each measurement. Knowing the surface and bedrock locations in the radar imagery allows us to compute ice sheet thickness, which is very important for the study of ice sheets, their volume, and how they may contribute to climate change issues. The previous time-consuming manual approach required sparse hand-selection of surface and bedrock interfaces by several human experts, and interpolating between selections to save time. Two primary methods have been studied: edge-based, and active contour. Results are compared and presented in terms of time requirements, error, and advantages which each method offers. Automatic ice thickness estimation results from 2006 and 2007 Greenland field campaigns show that the edge-based approach offers faster processing (seconds compared to minutes), but suffers from a lack of continuity and smoothness aspects that active contours provide. The active contour approach is more accurate when compared to ground truth selections from human experts, and has proven to be more robust to image artifacts.