英文论文研究An RM-NN algorithm for retrieving land surface temperature.pdf

1、本文档共17页，可阅读全部内容。
2、原创力文档（book118）网站文档一经付费（服务费），不意味着购买了该文档的版权，仅供个人/单位学习、研究之用，不得用于商业用途，未经授权，严禁复制、发行、汇编、翻译或者网络传播等，侵权必究。
3、本站所有内容均由合作方或网友上传，本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺！文档内容仅供研究参考，付费前请自行鉴别。如您付费，意味着您自己接受本站规则且自行承担风险，本站不退款、不进行额外附加服务；查看《如何避免下载的几个坑》。如果您已付费下载过本站文档，您可以点击这里二次下载。
4、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等，请点击“版权申诉”（推荐），也可以打举报电话：400-050-0827(电话支持时间：9:00-18:30)。

JOURNALOFGEOPHYSICALRESEARCH,VOL.112,D21102,doi:10.1029/2007JD008428,2007

Click

Here

for

Full

Article

AnRM-NNalgorithmforretrievinglandsurfacetemperatureand

emissivityfromEOS/MODISdata

KebiaoMao,1,2,3JianchengShi,4,5Zhao-LiangLi,6,7andHuajunTang8

Received18January2007;revised27April2007;accepted28June2007;published1November2007.

[1]ThreeradiativetransferequationsarebuiltforMODISbands29,31,and32,which

involvesixunknownparameters(averageatmospherictemperature,landsurface

temperature(LST),threebandemissivities,andwatervaporcontent).Therelationships

betweengeophysicalparametershavebeenanalyzedindetail,whichindicatesthatneural

networkisoneofthebestmethodstoresolvetheseill-posedproblems(LSTand

emissivity).Retrievalanalysisindicatesthatthecombinedradiativetransfermodel(RM)

withneuralnetwork(NN)algorithmcanbeusedtosimultaneouslyretrieveland

surfacetemperatureandemissivityfromModerate-ResolutionImagingSpectroradiometer

(MODIS)data.SimulationdataanalysisindicatesthattheaverageerrorofLSTis

under0.4Kandtheaverageerrorofemissivityisunder0.008,0.006,and0.006for

bands29,31,and32,respectively.Thecomparisonanalysisbetweenretrievalresultby

RM-NNandMODISproductalgorithmindicatesthatthegeneralizedsplitwindow

LSToverestimatestheemissivityandunderestimateslandsurfacetemperature.The

retrievalresultsbyRM-NNliebetweenthetwoproductsprovidedbyNASAandcloserto

day/nightLSTalgorithmafterstatisticsanalysis.Theaverageerroris0.36Krelative

toMODISLSTproduct(MOD11_L2)retrievedbygeneralizedsplitwindowalgorithmif

wemakearegressionrevision.Thecomparisonofretrievalresultswithground

measurementdatainXiaotangshanalsoindicatesthattheRM-NNcanbeusedtoretrieve

accuratelylandsurfacetemperatureandemissivi

该用户很懒，什么也没介绍

咨询Ta 进入空间

更多 >