Surface solar radiation calculation method based on deep learning

Abstract

The invention discloses a surface solar radiation calculation method based on deep learning which solves the problems that an existing method is long in calculation time and poor in universality and cannot be applied to cloud conditions. The method comprises the following steps of constructing an earth surface solar short wave radiation data set based on an atmospheric radiation transmission model; taking the surface solar short-wave radiation data set as a training sample; training through the neural networks to obtain a deep learning model, wherein the solar zenith angle, the aerosol opticalthickness, the atmospheric water vapor content, the cloud phase state, the cloud optical thickness, the cloud particle effective radius and the surface albedo are used as the neural network input layers, and the surface albedo, the surface downlink total solar radiation, the surface downlink direct solar radiation, the surface downlink diffused solar radiation and the atmospheric layer top upwardradiation are used as the neural network output layers. According to the method, the comprehensive, rapid and accurate calculation of the surface solar radiation is realized.

Description

technical field [0001] The invention relates to the field of satellite remote sensing inversion, in particular to a calculation method for surface solar radiation based on deep learning. Background technique [0002] The existing methods commonly used to calculate the surface solar short-wave radiation include traditional radiative transfer equation method and empirical model. The traditional radiative transfer equation method has a complex physical model and a large amount of calculation, which leads to a long calculation time, so it will be carefully considered during commercial operation; the empirical model mainly relies on a large number of surface measured data to establish meteorological parameters (such as temperature, humidity, etc.) , precipitation, cloud parameters, aerosol content, etc.) and the measured value of surface solar short-wave radiation, the model depends on the actual local geographical environment, so the universality is not high. At the same time, ...

AI Technical Summary

Problems solved by technology

[0003] The invention provides a method for calculating surface solar radiation based on deep learning, which solves the problems of long calculation time, poor universality and inability to apply to cloudy conditions in existing methods

Images