Abstract
New methods for ice concentration estimation from satellite based passive microwave data based on inversion are proposed. The proposed methods are superior to the existing methods, the NASA Team algorithm and the Comiso's Bootstrap algorithm with up to 45% of improvement on ice concentration estimation accuracy based on the simulation study. Also 1.5 to 2.1% of improvement was achieved for the proposed method compared to the NASA Team and Comiso's Bootstrap algorithms for the actual SSM/I data of Okhotsk uisng JERS-1/SAR data as a truth data for estimating ice concentration.
