Automated detection of near surface martian ice layers in orbital radar data

Abstract

An algorithm is presented to automatically detect near surface ice layers in images from the Shallow Subsurface Radar (SHARAD) on NASA's Mars Reconnaissance Orbiter. Mars' ice-rich Northern Polar Layered Deposits (NPLD) represents an extensive geologic record of climate history. Identifying ice layers in cross-sectional images leads to understanding the three-dimensional structure of ice layers. Scientists have manually identified layers in large data volumes, but the automated algorithm will allow studying more images from over a thousand orbital crossings. A unique coordinate transformation, based upon the surface reflection, makes subsequent filtering and detection more effective on near surface layers. Results show promising capabilities for automatically detecting ice layers on Mars. © 2010 IEEE.