Characterizing Arctic sea-ice floe elevation distribution using NASA's IceBridge ATM data

Abstract:

In this study, we aim to characterize sea-ice surface roughness by using NASA's IceBridge Airborne Topographic Mapper (ATM) L1B data and the coincident Digital Mapping System (DMS) imagery over the Arctic. The ATM elevation and the ratio of received pulse power to transmitted pulse power (RTratio) show clear bias patterns as a function of laser azimuth angle for some campaigns. These biases are empirically corrected in this study. From the bias-corrected ATM elevation we subtract the Danish Technical University (DTU) mean sea surface (DTU18 MSS) to obtain the sea-ice surface height. Cluster analysis is applied to sea-ice surface height and corrected RTratio to separate sea-ice floes and leads. The sea-ice surface height and RTratio bias corrections and the cluster analysis results are validated using coincident DMS imagery. The probability density function (PDF) of the floe height is calculated from ATM elevation and modeled using an exponentially modified normal distribution (exGaussian). The spatial and temporal variation of the parameters derived from exGaussian model fitting to the floe height PDF will be discussed. The impact of the ATM footprint distribution on the floe height PDF due to ATM's conical scan pattern are studied. The footprints are gridded with different grid sizes and results from gridded and not gridded data are shown. In addition, we characterize the surface roughness correlation characteristics by computing the autocorrelation function of ATM measurements along the conical scan. All of these results will help us inform the retracking of radar altimeter waveforms from sea-ice surfaces.