A numerical study on the SVD-based retrieval of radiometric data

A numerical study on the singular value decomposition (SVD) approach for enhancing the intrinsic spatial radiometric resolution is presented and discussed for a linear scanning configuration. The SVD approach is novel and attractive in terms of time efficiency due to its structure. The resolution enhancement study is of special relevance for application purposes within the field of environmental monitoring and multisensor data fusion. The results that we present show that, for a SSM/I-like configuration, the novel resolution enhancement procedure is capable of obtaining an 18 km resolution from data of 70 km intrinsic resolution in the presence of 1/spl deg/K root mean square (RMS) additive noise.