A method for air temperature retrieval from thermal infrared remote sensing data of polar orbiting meteorological satellites

Abstract

The invention provides a method for retrieving air temperature from thermal infrared remote sensing data of polar-orbiting meteorological satellites, comprising the following steps: Step 1: Obtaining polar-orbiting satellite observation data and matching to form a data set; Step 2, dividing the data set into cloudy data and clear sky data; Step 3: use stepwise linear regression method to carry out variable analysis on the variables in the cloudy data and clear sky data, so as to screen out the important variables that affect the temperature model construction as the input variables of the temperature inversion model ; Step 4: Using the extreme gradient boosting algorithm, use the selected cloudy data as the cloudy data set to build a cloudy sky inversion model; use the filtered clear sky data as a clear sky data set to build a clear sky inversion model. The model constructed by the invention can quickly estimate high-precision near-surface atmospheric temperature information under cloudy and clear sky conditions, and has high spatial resolution.

Description

technical field [0001] The invention relates to the technical field of satellite meteorology, in particular to a method for retrieving temperature from thermal infrared remote sensing data of polar-orbiting meteorological satellites. Background technique [0002] At present, remote sensing inversion of temperature based on thermal infrared data from polar-orbiting satellites is mainly carried out under clear sky conditions. According to relevant literature, the global coverage of cirrus clouds is about 20%-30%, and the coverage of cirrus clouds in tropical regions is even higher than 70%. Under cloudy conditions, the absorption, scattering and particle microphysical properties of the cloud will affect the radiation observed by satellites. Therefore, the influence of clouds must be considered in the physical process of atmospheric radiative transfer to reduce the error brought to the results of the physical retrieval of temperature by satellite remote sensing. [0003] The ...

AI Technical Summary

Problems solved by technology

However, since polar-orbiting satellites usually only observe the same location twice a day and the observation area is not fixed, and the high-resolution data of polar-orbiting satellite data is large

These problems will bring some difficulties to the temperature inversion

Images