Sea Ice Concentration Retrieval Using Composite ScanSAR Features in a SAR Data Assimilation Process

Abstract

Sensor and target originated variations in the scanning synthetic aperture radar (ScanSAR) images cause increase in variances of observations that affect the performance of data assimilation systems for sea ice analysis. Incidence angle (IA) dependence of radar backscatter along the range direction is one of the largest contributors to this variation. This important aspect can be taken into account both in a forward model and in a feature selection process. In this letter, an incidence-angle-dependent forward model is employed to reduce this variation. Moreover, instead of assimilation of ScanSAR features for the entire swath, individual features for specific IA intervals are used in order to decrease analysis bias and to achieve more consistent results.