Sensitivity of multisensor estimates of sea ice type concentration

Synthetic aperture radar (SAR) and multichannel passive microwave imagers are extensively for polar remote sensing. Individual sensor exploitation methods are maturing as evidenced by the exploitation processing development of both ERS-1 and RADARSAT, led by the Jet Propulsion Laboratory. Recently, methods have been developed for exploitation of multiple sensors for sea ice concentration estimates. The authors examine the sensitivity of errors in the underlying measurements on the joint estimates of sea ice concentration using a multisensor algorithm. The algorithm is a hybrid fusion method originally developed in 1995 and modified over the last two years to take advantage of advances in processing developed for hyperspectral exploitation. In this paper they examine the sensitivity of estimates of sea ice type concentration to variability of measurements in each of the individual sensors. The hybrid fusion technique is shown to have reduced sensitivity to sensor measurement error from the passive microwave sensor when compared to single-sensor methods.