Research On Soil Parameter Inversion Based On Polarimetric SAR Data

Soil's surface is an essential part of our ecosystem and a huge number of parameters is required for its detailed description and monitoring.Soil moisture content is an important parameter in the area of ecology,hydrology,agriculture and climate change.It is an important indicator of drought monitoring and estimation of crops,'especially as a basic condition for plant growth.Synthetic aperture radars are sensitive to changes in dielectric constant,making them the primary means of monitoring soil moisture content.The backscatter coefficient of SAR is mainly affected by many surface parameters,such as soil dielectric constant,surface roughness and vegetation coverage.Therefore,it is always a difficult point to inverse the surface water content under the dense vegetation cover.In this study,the soil moisture retrieval algorithm under vegetation cover was developed to estimate the soil moisture content under crop cover based on fully polarimetric spaceborme SAR data such,as Radarsat-2.The main contents of this paper include:1.Setting the parameter interval in AIEM model common range to build the simulation data set to analyze the influence of the parameters of the surface and the parameters of the radar system on the backscatter Coefficient by using AIEM model,the soil moisture content and surface roughness of the surface parameters are the main parameters of the study.2.The empirical relationship between the combination of backscatter coefficients of radar and each surface parameter is established to retrieve the surface Water content by combining the Dubois model that describing the surface backscatter coefficient of bare soil region and the water cloud model that describing the backscattering mechanism of vegetation coverage.Finally,the inversion algorithm of soil water content under vegetation cover based on the full polarization data of Radarsat-2 is proposed.close-range photogrammetry,which can reduces the errors caused by the calibration parameters in the inversion algorithm.The rms height and the correlation length measurement error of the close-range photogrammetric method are reduced by 12%-35.1%and 19.6%-62.4%respectively compared to the pin-profiler method,which improves the accuracy and efficiency of surface roughness measurement.4.Consistency analysis between the Gaofen-3 fully polarimetric SAR backscattering coefficient and other SAR backscattering coefficient shows that there is some error in the radiometric calibration of Gaofen-3 SAR data.However,the correlation between the backscattering coefficient and soil water content is studied by the method of eigenvector decomposition,which shows that it still has the conditions for the retrieval of soil water content.