Using Deep Learning to Model Elevation Differences between Radar and Laser Altimetry

Satellite and airborne observations of surface elevation are critical in understanding climatic glaciological processes quantifying their impact on changes ice masses sea level contribution. With the growing number dedicated campaigns experimental operational satellite missions, science community has access to unprecedented ever-increasing data. Combining datasets allows potentially greater spatial-temporal coverage improved accuracy; however, combining data from different sensor types acquisition modes is difficult by differences intrinsic properties processing methods. This study focuses combination measurements derived ICESat-2 Operation IceBridge LIDAR instruments CryoSat-2’s novel interferometric radar altimeter over Greenland. We develop a deep neural network based sub-waveform information CryoSat-2, between LIDAR, additional inputs representing local geophysical information. A time series maps created showing observed LIDAR-radar model predictions. Mean vs. adjustments broad spatial temporal trends thereof recreated network. The also predicts radar-LIDAR with respect waveform parameters better than simple linear model; point magnitudes underestimated.