Application Potential Of Polarimetric SAR Data For Estimation Surface Parameter In Arid Region

Soil moisture and surface roughness are important parameters that reflect surface conditions and play an important role in hydrology,meteorology,and agronomy.In previous studies,optical remote sensing was the main method for studying parameters such as soil moisture and surface roughness.Although significant results have been achieved,optical remote sensing is subject to certain limitations due to limitations of data acquisition and other conditions.Sex.In recent decades,with the rapid development of remote sensing technology,especially radar remote sensing technology in theory and application,Polarized SAR(PolSAR)system has gradually emerged and matured,due to its full-time,all-weather,distinctive characteristics,and a certain degree of wearability.Transparency and other advantages,polarimetric radar images have caused great interest in the remote sensing community.As a typical region of arid-semi-arid regions in China,the Juyanze region is typical and representative in the study of global climate change and environmental evolution.This paper uses RADARSAT-2 PolSAR data.First,we use the Advance Integral Equation Model(AIEM)model and the Oh2004 model to invert the soil moisture and surface roughness in the Juyanze area;second,we use polarimetric and phase processing and polarimetric decomposition.A total of 16 polarimetric parameters such as entropy,anti-entropy,α-angle,ERD(Eigenvalue Relative Difference),and RVI(Radar Vegetation Index)are extracted.The single-polarimetric parameters and multipoles are analyzed by linear and nonlinear fitting methods.The response relationship between the parameters and combined polarimetric parameters and soil moisture and surface roughness;Finally,the application potential of C-band PolSAR data in surface parameter estimation in arid area was explored.The main conclusions are as follows:(1)There is a good correlation between the results of the inversion of surface parameters and the measured results of the AIEM model and the Oh2004 model.After verifying,the correlation coefficient of the inversion of soil moisture by the AIEM model is 0.880,and the root mean square height of the AIEM model is inverted.The correlation coefficients with the correlation length L were 0.551 and 0.670,respectively.The correlation coefficient of the root mean square height S after inversion of the Oh2004 model was 0.628,indicating that the inversion of surface parameters using the above model had higher accuracy in the study area.(2)Correlation analysis of polarimetric parameters and soil moisture showed that the correlation between single-polarimetric parameters and soil moisture was low,and the correlation coefficient between anti-entropy and soil moisture was the highest 0.287,SERD and soil The correlation coefficient between moisture was the lowest at 0.023;however,the coefficient of correlation increased with the non-linear fitting method,and the correlation coefficients between entropy and SERD and soil moisture were 0.348 and 0.401,respectively;the multipolar parameter was linear with the soil moisture.During the fitting,with the increase of the number of polarimetric parameters,the correlation between the polarimetric parameters and the soil moisture becomes higher and higher,RMSE and MAE gradually decrease,and the correlation coefficient between soil moisture and polarimetric parameters is the highest.0.680;In the non-linear fitting,the correlation between soil moisture and multipolarimetric parameters also increased,the highest correlation coefficient was 0.501;the correlation between combined polarimetric parameters and soil moisture and surface roughness was not satisfactory,and the correlation coefficient was The highest is 0.269.(3)Correlation analysis of polarimetric parameters and surface roughness shows that the correlation between single-polarimetric parameters and surface roughness is low,and the correlation coefficient is less than 0.30;however,there are correlation coefficients when using non-linear fitting method.The increase,in which the correlation coefficient between the entropy and the root mean square height S increases from 0.269 to 0.298,and the correlation coefficient with the correlation length L increases from 0.154 to 0.235;the multipolarimetric parameter is linear with the surface roughness During the fitting,as the number of polarimetric parameters increases,the correlation between polarimetric parameters and surface roughness gradually increases,RMSE and MAE gradually decrease,and the correlation coefficient between S and L and polarimetric parameters is the highest.At 0.655 and 0.560,the correlations between S and L and multi-polarimetric parameters are significantly improved by the nonlinear fitting.The highest correlations are 0.767 and 0.674,respectively.The correlation between the combined polarimetric parameters and the soil moisture and surface roughness is also significant.Not ideal,the highest correlation coefficient is 0.263.(4)Based on the analysis results of single-parameter,multi-parameter,and combined polarimetric parameters,we believe that there may be a certain correlation between the polarimetric parameters and the surface parameters,although the nonlinear fitting method can improve the degree to a certain extent.The correlation between the parameters and the surface parameters,however,the current results have not yet reached the accuracy of being able to use only the polarimetric parameters to invert the surface parameters.Applying the polarimetric parameters to the inversion of surface parameters fully exploits the potential of full-polarimetric radar data and explores a more in-depth quantitative relationship between polarimetric parameters and surface parameters.This is a work that needs to be further developed in the future.