Research on the selection method of FY-3D/MWHTS clear sky observation data based on neural network

Abstract

The selection of clear sky data in space-borne remote sensing data is very important for its data application. For FY-3D satellite microwave humidity and temperature sounder (MWHTS), an inversion system of atmospheric cloud water content by MWHTS is established based on neural network. The cloud water content inversion value is used to select clear sky data from MWHTS observation data. The experimental results show that FY-3D/MWHTS clear sky data selection method based on neural network can effectively select MWHTS observation data, thus improving the simulation brightness temperatures accuracy of MWHTS by radiative transfer model. This method can be used to select clear sky data by using space-borne observation data itself. It is easy to operate and has important practical value for climate change research, numerical weather forecast, etc., based on space-borne observation data.