Short-term forecasting of atmospheric meteorological parameters based on the results of the neural network of a three-band microwave radiometric system

Abstract

The article considers a variant of inclusion of a neural network into the microwave radiometric system of remote sensing of the atmosphere to perform short-term forecasting of meteorological parameters, based on current measurements of power of own radio-thermal emission of the atmosphere in three frequency bands. The initial data for the training of the neural network- the results of daily microwave radiometric measurements in three frequency bands and data on atmospheric meteorological parameters in the near ground layer obtained using a mobile meteorological station- are indicated. The results of forecasting the temperature and humidity of the near ground layer of the atmosphere based on the current results of microwave radiometric measurements for the current moment and with a two-hour lead time are presented.