Sea-ice thickness retrieval in the Sea of Okhotsk using dual-polarization SAR data

We investigated the feasibility of using multi-polarization synthetic aperture radar (SAR) data to estimate the thickness of undeformed first-year ice. Analysis of the radar signatures for the C- and L-bands showed that the correlation between the ice thickness and VV-to-HH backscattering ratio is larger than the correlation between the ice thicknesses and the backscattering coefficients. This is in part because the ice surface salinity and hence the surface reflection coefficient decreases as the ice thickens. The backscattering ratio had low sensitivity to the small-scale ice surface roughness for the C-band and is almost independent of roughness at L-band. Given that the ratio is most sensitive to ice surface dielectric constants, which depend on salinity, we developed an algorithm for retrieving the ice thickness that is based on the backscattering ratio and on the integral-equation-method (IEM) surface scattering model. Comparison of the observed and estimated ice thicknesses showed that the correlation was much better when the thicknesses were estimated using the backscattering ratio than when the backscattering coefficient was used directly. The algorithm also performed better than previous retrievals using an empirical technique.