盐渍化土壤水分微波遥感反演方法研究与解析

Study of retrieval method of salinized soil moisture using microwave remote sensing

土壤水分是农田粮食产量和质量的关键影响因素,高精度大面积的土壤含水量的反演对于评价不同地区的土壤质量、估产、灌溉等具有重要意义。合成孔径雷达(SAR)是当前进行土壤含水率反演的重要手段之一,但盐渍化农田的土壤含水率SAR反演方法目前还未明确。本文利用Radarsat-2全极化影像结合内蒙古河套灌区农田实地监测数据,构建了盐渍化土壤的含水率的SAR反演模型。首先提取了四种极化方式的后向散射系数,之后通过建立不同后向散射系数组合与土壤水分的线性回归模型的方式,将反演数据与实测数据进行对比,发现HH-VV极化方式取得了较高的精度,最终利用同极化比构建的模型反演了土壤含水率,反演值的均方根误差达到2.23%,其精度能够满足农田土壤含水率反演精度要求。

Soil moisture is a key factor affecting grain yield and quality in farmland.The inversion of soil moisture with high precision and large area is of great significance for evaluating soil quality,yield estimation and irrigation in different regions.Synthetic aperture radar(SAR)is one of the most important methods for soil moisture retrieval at present,but the method for soil moisture retrieval in salinized farmland is not clear at present.In this paper,a SAR inversion model of salinized soil moisture was constructed by using Radarsat-2 fully polarized image combined with field monitoring data in Hetao irrigation area of Inner Mongolia.First,the back-scattering coefficients of the four polarization modes were extracted,and then the inversion data was compared with the measured data by establishing a linear regression model of the combination of different backscattering coefficients and soil moisture.It was found that the HH-VV polarization mode achieved high accuracy,and finally the soil water content was retrieved by using the model constructed by the same polarization ratio.The RMSE of the inversion value is 2.23%,which can meet the requirement of the accuracy of soil moisture inversion.