Spatial Resolution Enhancement And Interpolation Using Backus-Gilbert Inversion: Demonstration With TMI Brightness Temperature Data

摘要

The Backus-Gilbert Inversion technique is a proven method of temperature inversion in microwave radiometry with a scope of trade-off between desired resolution and tolerable noise. In this study it is used to super-resolve brightness temperature data independent of the inversion process. The BGI-method is applied on scan-mode TMI (TRMM Microwave Imager) data. The 19.35GHz TMI channel provides brightness temperature data oversampled by a factor of two and geo-collocated with the 37GHz channel data. Consequently, the enhancement of features in the 19.35GHz image after BG-resolving is validated by the identification of such features in the 37GHz image. The potential of the BGI-method as a non-degrading interpolation technique is also tested.