Soil moisture inversion model training method, soil moisture inversion method and device

By performing consistency analysis and bias correction on a series of satellite remote sensing images, the inversion physical model was revised. By combining the physical model with deep learning, the problems of terrain factors and model instability in existing technologies were solved, and high-precision soil moisture inversion in complex terrain areas was achieved.

CN122021362BActive Publication Date: 2026-07-03AEROSPACE INFORMATION RES INST CAS
View PDF 3 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
AEROSPACE INFORMATION RES INST CAS
Filing Date
2026-04-14
Publication Date
2026-07-03

Smart Images

  • Figure CN122021362B_ABST
    Figure CN122021362B_ABST
Patent Text Reader

Abstract

This invention provides a method for training a soil moisture inversion model, a soil moisture inversion method, and an apparatus, belonging to the field of remote sensing information processing technology. The method includes: determining a target correction factor using consistency analysis results of synthetic aperture radar (SAR) remote sensing images from a series of satellites, and correcting the target inversion physical model to obtain a corrected inversion physical model; using the corrected inversion physical model to generate a theoretical simulation data sample set; using the measured backscattering coefficients determined based on SAR remote sensing images to correct the deviation of the theoretical backscattering coefficients; and finally, using the corrected backscattering coefficients and the parameter values ​​of each physical variable in the theoretical simulation data sample to construct a corrected simulation training sample set for pre-training the soil moisture inversion model. This invention constructs a multi-satellite collaborative SAR inversion system and a semi-empirical deep learning fusion model combining "physical guidance + data learning," significantly improving the accuracy and stability of soil moisture inversion.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Multi-objective function change detection method for inverting soil moisture

    CN115840224A

  • Multi-source remote sensing soil moisture inversion method based on Stacking algorithm

    CN116579148A

  • Soil water content inversion method and system based on L-band dual-polarization data and medium

    CN121141591A