A Multichannel Data Fusion Method to Enhance the Spatial Resolution of Microwave Radiometer Measurements

In this study, a method to improve the reconstruction performance of antenna-pattern deconvolution based on the gradient iterative regularization scheme is proposed. The method exploits microwave measurements acquired by a multichannel radiometer to enhance their native spatial resolution. The proposed rationale consists of using the information carried on a high-frequency (finer spatial resolution) channel to ameliorate the spatial resolution of the lowest resolution radiometer channel. Experiments performed using both synthetic and real special sensor microwave/imager (SSM/I) radiometer data demonstrate that an enhanced spatial resolution 19.35-GHz channel can be obtained by ingesting in the algorithm information coming from 37.0-GHz channel. This multichannel spatial resolution method is also shown to outperform the conventional gradient-like regularization scheme in terms of both observation of smaller targets and reduction of ringings and fluctuations.