Brightness Temperature Reconstruction Using BGI

This paper departs from the popular usage of the Backus-Gilbert inversion (BGI) method as a tool for inversion of antenna temperature measurements in microwave radiometry. The BGI method is applied in this paper to enhance the information content of an existing set of oversampled brightness-temperature (T
_B
) data. The purpose is to isolate the inversion process from its resolution enhancement counterpart. The advantage gained is that the resolution enhancement can be performed in a simplified way and in a different level of processing that starts with the scan-mode T
_B
data product and simply requires with it the knowledge of the antenna gain pattern and the sensor's scan geometry. The technique is demonstrated with the 19.35-GHz Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI) channel, which provides oversampled T
_B
data. The radiometric resemblance of this channel with that of the 37 GHz and geocollocation of their T
_B
footprints facilitate validation of the enhancement of features. The significance of oversampling the low-frequency (LF) radiometer channels is underscored in the process, which gives the authors the confidence to propose oversampling of the LF data for the forthcoming sensor Microwave Analysis and Detection of Rain and Atmospheric Structures (MADRAS) onboard the Megha_Tropiques mission, which is a joint ISRO-CNES collaboration (due for launch in 2009).