Method for decomposing passive microwave mixed pixel of forest land accumulated snow based on multi-frequency and dual-polarization

Abstract

The invention discloses a method for decomposing a passive microwave mixed pixel of forest land accumulated snow based on multi-frequency and dual-polarization, belonging to the technical field of remote sensing image processing. The method disclosed by the invention aims at the relatively big error problem when snow depth inversion of the forest land accumulated snow is carried out by utilizing passive microwave data in the prior art. The method disclosed by the invention comprises the following steps: firstly, obtaining and re-classifying land classification data of a forest land underlying surface, then, establishing a model for decomposing the passive microwave mixed pixel of the forest land accumulated snow based on multi-frequency and dual-polarization on the basis of a classification result, and finally, solving based on an underdetermined equation set of a dynamic window data selection strategy so as to obtain component brightness and temperature data and error data corresponding to various underlying surfaces after decomposing. In the process, the spatial correlation of the microwave pixel is considered; two kinds of data input schemes including 8 neighbouring window data input and 4 neighbouring window data input are provided; and four solutions are provided for optimizing a solving result according to the two kinds of input schemes.

Description

technical field [0001] The invention belongs to the technical field of remote sensing image processing, in particular to a multi-frequency dual-polarization woodland snow passive microwave hybrid pixel decomposition method. Background technique [0002] Snow cover is one of the most active natural factors on the earth's surface. 3 / 4 of the land's fresh water resources exist in the form of ice and snow, and 1 / 3 of the surface of the earth has seasonal snow cover. Eurasia and North America have at least 80% of the land surface is covered by snow. Snow cover is an important factor determining the radiation balance. It not only has a strong climate effect, but also has an extremely important impact on energy and water cycle processes. Because passive microwave remote sensing has high temporal resolution and can quickly cover the whole world, it plays a particularly prominent role in monitoring the temporal and spatial changes of snow cover at the global and continental scales. ...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the problem that the heterogeneity in the woodland pixel has a large influence, which leads to a large error when using the snow depth inversion method, the present invention combines the passive microwave antenna with the underlying surface data in the woodland passive microwave observation pixel Gain function, providing a multi-frequency dual-polarization woodland snow passive microwave hybrid pixel decomposition method

Images