Synergistic Inversion Of Soil Salinity Multi-source Remote Sensing Data In Hetao Irrigation District

Hetao irrigation area in Inner Mongolia Autonomous Region,as an important agricultural area in China,also faces serious soil salinization which hinders the sustainable and rapid development of agriculture in Hetao irrigation area.Remote sensing technology can monitor and master the soil salinization situation in a real-time and dynamic manner,thereby developing targeted soil salinization control program.Through image acquisition,soil sampling and spectral acquisition in the experimental area,this paper explored the response characteristics of sensing image surface reflectance to surface salinized soil analyzed the response characteristics of hyperspectral reflectance and hyperspectral data after processing to surface salinized soil,and studied the response characteristics of the imaginary part of dielectric constant and radar image backscattering coefficient to surface salinized soil.The single factor soil salinity inversion model,synergistic soil salinity inversion model and soil salinity simulated value and measured value model were established,and the corresponding accuracy of each model was evaluated.The results showed that the correlation coefficients of the two synergistic inversion models were higher than 0.79,the maximum error of root mean square was 0.8934,and the maximum score of the average absolute error and average relative error was 0.7188 and 0.1256 respectively.The accuracy assessment showed that the accuracy of the synergistic inversion model was better than that of single factor inversion model and had better predictability for soil salinity.