利用全极化SAR数据反演地表土壤水分方法研究

项目名称： 利用全极化SAR数据反演地表土壤水分方法研究

项目编号： No.41471299

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 天文学、地球科学

项目作者： 余凡

作者单位： 中国测绘科学研究院

项目金额： 85万元

中文摘要： 微波遥感以其自身的特点与优势，成为目前土壤水分观测中最有潜力的手段之一。全极化数据拥有较多的观测波段，且极化分解技术能解析主散射机制，对提高地表土壤水分反演精度有着重要的意义。本研究利用全极化SAR数据，分别对裸露地表和植被覆盖区建立土壤水分反演算法。对于裸露地表，首先利用极化分解对地形效应引起的极化偏转角进行修正，然后选择相关性较低的不同极化SAR数据组合做为因变量建立土壤水分反演的半经验模型，并对极化偏转角补偿的效果进行验证；在植被覆盖区，提出一种基于全极化SAR数据的植被微波辐射模型参数化方法，在简化MIMICS模型的基础上，首先利用遗传规划搜索算法建立简化模型中的关键参数与植被参数间的非线性关系，然后对植被的重叠效应进行校正，并建立极化分解散射分量与简化模型中散射分量的函数关系，得到植被覆盖区的反演模型。最后利用地面实测土壤水分对遥感反演结果进行真实性检验。

中文关键词： 土壤水分；主动微波遥感；全极化；极化分解

英文摘要： Microwave remote sensing has become one of the most potential methods in soil moisture observations, based on its own characteristics and advantages. In this research, two soil moisture inversion algorithm which are used for bare soil surface and vegetated area, are presented using fully polarimetric SAR data. In the model for bare soil surface, the polarimetric decomposition technique is used to correct the polarimetric orientation angle which caused by topographic effect. Then, a semi-empirical model for soil moisture inversion is established by combination of different polarimetric SAR bands which are less relevant. And the effect of polarimetric orientation angle correction is validated by the semi-empirical model. When it comes to vegetated area, we proposed a radiation transfer model based on fully polarimetric SAR data. Derivation of the algorithm is based on simplification of the Michigan Microwave Canopy Scattering Model (MIMICS). Empirical relationships simulated among key parameters of simplified MIMICS and vegetation parameters. Then, A method is introduced to correct for the radar-shadow effect caused by over-laying vegetation, and the empirical function among the scattering component produced by polarimetric decomposition and scattering component of simplified MIMICS is built to inverse the soil moisture. At last the soil moisture inversed by remote sensing data is validated by the field ground measurements.

英文关键词： soil moisture;active microwave remote sensing;full polarization;polarimetric decomposition