摘要:Currently, application - oriented researches on the data of Medium Resolution Spectral Imager (MERSI) , which is on board China ' s new generation polar orbit meteorological satellite FY - 3 , are very insufficient, due to the reason that the data as a new source have been delivered only since 2008. With the normal operation of FY - 3 satellite system, it is necessary to develop an operational module for FY - 3/MERSI regional land surface temperature (LST) retrieval and its post - processing, since LST is required for a wide variety of scientific studies but FY - 3/MERSI' s operational LST products have not yet been provided by National Satellite Meteorological Center (NSMC ). Rased on an analysis of FY - 3/MERSI LI data' s HDF5 format and its channel characteristics, the authors selected the generalized single - channel algorithm developed by Jimenez - Munoz & Sobrino to directly realize the LST retrieval at 250 m spatial resolution with MATLAR programming and the thematic mapping of LST derivative products. This paper describes the parametric processes of LST retrieval algorithm in detail, which include radiometric calibration, cloud detection, estimation of two intermediate parameters - surface emissivity and atmospheric water vapor, and calculation of thermal indexes from LST. On these bases, an automatic flowchart for FY - 3/MERSI LST retrieval and thematic mapping was established. Experimental results of this flowchart applied in Shanghai thermal environmental monitoring show that it can process FY - 3/MERSI LI data in a fast, real - time and automatic way, thus suitable for operational products producing and sharing, with the saving of human resources. It is also proved that FY - 3/MERSI data and various forms of LST products can reveal the spatial pattern of Shanghai thermal field and the urban heat island effect more finely and intuitively.%我国新型自主研发的风云三号卫星MERSI(FY-3/MERSI)数据目前多见于试验研究,国家卫星气象中心(NSMC)发布的业务产品中尚缺乏地表温度(land surface temperature,LST)产品.为此,在分析FY-3/MERSI L1级数据HDF5格式及通道特点的基础上,采用Jiménez-Mu(n)oz和Sobrino普适性单通道地表温度反演算法,借助MATLAB编程,实现了直接利用MERSI数据进行250 m空间分辨率的LST反演与专题制图.详细介绍了LST反演算法的参数化流程,包括辐射定标、云检测处理以及地表比辐射率和大气水汽含量的估计等,并在反演LST的基础上,计算其他衍生指标,建立了FY-3/MERSI数据反演LST及制图流程.对上海市空间热环境监测分析的实验结果表明:该方法能根据设定范围对FY-3/MERSI L1数据进行快速、实时和自动处理,适用于业务化产品制作与共享,节省人力资源;形式丰富的MERSI专题制图产品可更直观、精细地揭示出上海城市热场空间格局与热岛效应特征.