Microwave radiometer temperature and humidity inversion algorithm based on second sounding data

The invention discloses a second sounding data-based microwave radiometer temperature and humidity inversion algorithm, and the algorithm comprises the steps: constructing a BPNN-based neural network algorithm through historical second sounding meteorological observation data, ground meteorological observation data and brightness temperature data of each channel of a microwave radiometer; according to the algorithm, real-time ground meteorological observation data and real-time brightness temperature data of each channel of the microwave radiometer are used as input quantities to inversely calculate vertical distribution of real-time atmospheric temperature and humidity; the method mainly comprises the following steps of effective data selection of second sounding data, average value and extreme value calculation of the second sounding data, extreme value calculation of radiation brightness temperature data of each channel of the microwave radiometer, data normalization, inversion calculation based on a BPNN