Machine Learning for Characterizing Shallow Subsurface Ice via Radar-Thermal Data Fusion: Validation at Lake Vostok, East Antarctica
Abstract
The forthcoming NASA Clipper and ESA JUICE missions to Jupiter's moon Europa will both carry a radar sounder and a thermal imager (REASON/E-THEMIS; RIME/MAJIS respectively). Our work creates an initial understanding of how thermal and radar imagers can be used together to significantly enhance remote surface characterization and inform follow-up landing operations onto icy worlds. We combine machine learning and Bayesian statistics to develop a joint geophysical radar and thermal model for ice. We validate the method by performing the joint inversion on remote sensing data of the ice sheet above Lake Vostok, East Antarctica.
- Publication:
-
EPSC-DPS Joint Meeting 2019
- Pub Date:
- September 2019
- Bibcode:
- 2019EPSC...13..192C