被动微波遥感反演地表发射率研究进展  

Advances in the Study of Land Surface Emissivity Retrieval from Passive Microwave Remote Sensing

作　　者：吴莹[1,2] 王振会[1,2] 

机构地区：[1]南京信息工程大学气象灾害省部共建教育部重点实验室,南京210044 [2]南京信息工程大学大气物理学院,南京210044

出　　处：《国土资源遥感》2012年第4期1-7,共7页Remote Sensing for Land & Resources

基　　金：国家自然科学基金项目(编号:41175035;40975019和40875016);南京信息工程大学校科研启动项目(编号:20100372)共同资助

摘　　要：微波地表发射率是表征地表特征的重要参数,也是反演地表、大气参数的重要条件。相比较物理模型,其模拟计算需要若干输入参数,且相当一部分地表、植被特征参数很难从常规资料中获取,应用星载被动微波辐射计资料可以在更大空间和时间尺度范围内直接反演地表发射率。从目前常用的几种被动微波遥感反演方法(包括经验统计方法、辐射传输方程方法、指数分析方法、神经网络方法、一维变分方法等等)回顾了微波地表发射率反演的国内外研究进展及其研究中存在的问题,并对这些方法的优、缺点进行了评价。最后指出,今后应开发识别和订正直接影响卫星观测值的无线电频率干扰(RFI)算法,改善云、雨检测算法,并且加强微波波段大气辐射传输等过程的机理研究。The microwave land surface emissivity（MLSE）is a very important parameter for describing the characteristics of the lands,and it is also a key factor for retrieving the parameters of land and atmosphere.Space-borne passive microwave radiometers provide direct retrieved land surface emissivity spectra with larger temporal and spatial scales compared with physical modeling simulation in that the physical modeling simulation needs plenty of parameters,but quite a few of these parameters,such as parameters of land surface and vegetation,are not available from traditional measurements.This paper systematically reviews MLSE retrieving algorithms for passive microwave remote sensing data,which include statistical approach,atmospheric radiation transfer model approach,index analysis approach,neural network approach and one-dimensionally variational analysis approach.The main advantages and limitations of these five methods are also discussed.Finally,the development tendencies of estimating MLSE by remote sensing are pointed out,such as developing algorithms of Radio Frequency Interference（RFI） detection and correction,improving algorithms of detection of clouds and rain-affected radiances,and intensive research on microwave atmospheric radiation transfer process.

关 键 词：被动微波遥感 地表发射率 反演方法 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置] TP722.6[自动化与计算机技术—控制科学与工程]