Retrieval of Atmospheric Temperature Profiles from Hyperspectral Microwave Radiative Data Based on the Neural Network

Abstract

In this study, the back propagation neural network (BPNN) is applied to retrieve atmospheric temperature profiles, using 50~70 GHz passive microwave brightness temperature values. This paper makes a comparison of the retrieval of atmospheric temperature using the statistical inversion method and the BPNN inversion method. The results of the retrieval experiment show that the accuracy of the BPNN is better than that of the statistical inversion method in the upper atmosphere. It is of great application value to study the retrieve of atmospheric temperature profiles.