土壤水分多源卫星遥感联合反演研究进展

Progress in research on the joint inversion for soil moisture using multi-source satellite remote sensing data

土壤水分与全球气候变化、碳循环、水循环等过程,以及农业生产、生态保护修复等环节密切相关。土壤水分的探测经历了从地面测量到遥感探测的发展过程,实现了全球及区域尺度的调查监测。由于数据谱段不一、辐射传输机制不同、反演算法多样,需从算法机理、优势、局限等角度进行全面分析,为精度提升、算法改进提供基础。为此,该文从光学遥感、微波遥感、光学微波协同3个方面,系统梳理了光学遥感温度-植被指数空间特征、温度-地表短波净辐射时间特征反演土壤水分,被动、主动微波反演和主、被动微波遥感联合反演,以及基于精度改善和时空尺度转化的光学微波协同反演等技术方法的特点、存在的问题。目前,多源遥感数据联合反演土壤水分主要存在如下问题:①数据存在缺失及时空尺度不匹配问题;②不同数据源对地表穿透性不一致;③联合反演模型依赖于经验参数和大量辅助参数。随着卫星监测网络的完善、数据源对地表探测深度研究的深入,以及联合反演物理机理的明确、辅助参数时空连续数据集的建立,上述问题会得到有效解决。

Soil moisture is closely associated with global climate change,the carbon cycle,and the water cycle,as well as agricultural production and ecological conservation and restoration.The detection of soil moisture has shifted from ground survey to remote sensing detection,achieving global-and regional-scale survey and monitoring.Given differences in data spectrum segments,radiative transfer mechanisms,and inversion algorithms,it is necessary to comprehensively analyze the mechanisms,advantages,and limitations of algorithms,with the purpose of laying a foundation for accuracy and algorithm improvement.From the aspects of optical remote sensing,microwave remote sensing,and optic-microwave cooperation,this study systematically analyzed the features and challenges of the following inversion techniques:inversion based on the T s-VI spatial and T s-NSSR temporal characteristics of optical remote sensing data,inversion using passive and active microwave data,joint inversion using active and passive microwave data and remote sensing data,and optical-microwave cooperative inversion based on accuracy improvement and spatio-temporal transformation.At present,the joint inversion of soil moisture using multi-source remote sensing data faces the following challenges:①The data suffer missing and spatio-temporal mismatching;②Different data sources exhibit varying degrees of surface penetration;③The joint inversion model relies on empirical parameters and numerous auxiliary parameters.These challenges can be addressed with the improvement in the satellite monitoring network,the increase in the surface detection depths of data sources,the clarification of the physical mechanisms of joint inversion,and the establishment of spatio-temporal continuous datasets of auxiliary parameters.