Study on sea ice thickness estimation using the latest space-borne polarimetric SAR sensors

Variation of sea ice is one of the most evident indicators of climate change on our planet. Thinning of sea ice has been regarded as the major contributing factor that leads to the massive sea ice loss in Arctic sea. Direct measurement of sea ice can be made available by drilling a hole through the ice and/or by using Electromagnetic Induction system. Although these methods are accurate, they are time consuming and limited in space and time. We investigated the usability of polarimetric parameters (backscattering coefficient, VV-to-HH backscattering coefficient ratio, depolarization factors, and target decomposition theorems) of X-, C- and L-band space-borne SAR data in estimating sea ice thickness in Arctic sea. Acquisitions of the latest high-resolution SAR data (i.e. TerraSAR-X, RADARSAT-2, and ALOS PALSAR) were coordinated with sea ice field campaigns at the coast of Greenland. We found that the backscattering coefficients and VV-to-HH backscattering coefficient ratio (BCR) of C- and X-band SAR data had a weak and insignificant correlation with sea ice thickness, while the BCR of L-band SAR data had a perceptible correlation with sea ice thickness. We also found that a strong correlation between sea ice thickness and depolarization factors (co-polarized correlation and cross-polarized ratio). This suggests that target depolarization factors can be effective parameters in estimating sea ice thickness in Arctic sea. It is known that the target depolarization is strongly related to changes in surface roughness. Sensitivity study has shown that the observed ice thickness to depolarization relationship was partly explained by surface roughness effects.