資源描述:
《基于TM8衛(wèi)星熱紅外數(shù)據(jù)地表溫度反演及其應(yīng)用-論文.pdf》由會員上傳分享,免費(fèi)在線閱讀,更多相關(guān)內(nèi)容在行業(yè)資料-天天文庫。
1、146廣東農(nóng)業(yè)科學(xué)2015年第6期基于TM8衛(wèi)星熱紅外數(shù)據(jù)地表溫度反演及其應(yīng)用何光聰,胡月明,-,肖志峰(1.華南農(nóng)業(yè)大學(xué)信息學(xué)院/廣東省土地利用與整治重點(diǎn)實(shí)驗(yàn)室,國土資源部建設(shè)用地再開發(fā)重點(diǎn)實(shí)驗(yàn)室,廣東廣州510642;2.廣州市測繪地理信息行業(yè)工程中心,廣東廣州510642)摘要:地表溫度(I.ST)是農(nóng)業(yè)旱災(zāi)監(jiān)測模型和農(nóng)作物估產(chǎn)模型的關(guān)鍵因子,在干旱遙感監(jiān)測中有著廣泛應(yīng)用。針對TM8衛(wèi)星運(yùn)行陸地成像儀(OLI)和熱紅外傳感器(TIRS)數(shù)據(jù)波段特點(diǎn),提出新的劈窗算法流程圖,推導(dǎo)新的劈窗算法系數(shù),對地表比輻射率
2、和大氣透過率這兩個(gè)基本參數(shù)進(jìn)行了分析。選取同一時(shí)刻珠三角區(qū)域的MODIS溫度產(chǎn)品和氣象站點(diǎn)觀測值作為基準(zhǔn),對該算法進(jìn)行精度分析。結(jié)果表明:LST散點(diǎn)圖得出RMSE為0.3845;從LsT差值直方圖得出反演LsT和產(chǎn)品LST誤差范圍主要集中在一0.6~0.6℃之間;對觀測的氣象站點(diǎn)值進(jìn)行統(tǒng)計(jì),產(chǎn)品站點(diǎn)溫度均值、算法反演溫度均值和站點(diǎn)觀測溫度均值分別為21.61、21.19、21_38℃,對高溫區(qū)域的氣象站點(diǎn)溫度數(shù)據(jù)統(tǒng)計(jì)得出該算法的RMSD為0.742oC,誤差小于ICE。從實(shí)例應(yīng)用來看,該算法能快速有效地反演農(nóng)業(yè)旱災(zāi)
3、監(jiān)測中所需要的LST參數(shù),并獲得較好的反演效果,能提高農(nóng)業(yè)旱災(zāi)監(jiān)測模型的反演精度。關(guān)鍵詞:地表溫度;遙感;傳感器;農(nóng)業(yè)旱災(zāi)監(jiān)測;反演中圖分類號:S127文獻(xiàn)標(biāo)識碼:A文章編號:1004—874X(2015)06—0146—08LandsurfacetemperatureretrievalbasedonTM8thermalinfrareddataanditsapplicationHEGuang-cong,HUYue-ming,XIAOZhi-feng(1.CollegeofInformatics,SouthChina
4、AgriculturalUniversity~KeyLaboratoryofLandUseandRemediationofGuangdongProvince~KeyLaboratoryofConstructionLandRedevelopment,MinistryofLandandResources,Guangzhou510642,China;2.EngineeringCenterofSurveyingandMappingGeographicInformationIndustryinGuangzhou,Gunagz
5、hou510642,China)Abstract:Thelandsurfacetemperature(I_ST)isakeyfactorwidelyappliedinagriculturaldroughtmonitoring,cropyieldestimationmodelandithasanimportantapplicationindroughtremotesensingmonitoring.Inthispaper,weproposedanewsplit-windowalgorithmflowcha~andde
6、rivednewparameterstoretrieveI_STfromLandsat8satelliteaccordingtooperationlandimager(OLI)andthethermalinfraredsensor(TIRS)bandcharacteristics.Theatmospherictransmittanceandsurfaceemissivitywerecomputedandanalyzedindetail.SelectingMODIStemperatureproductandmeteo
7、rologicalstationsinthesameregionasreference,thispaperanalyzedtheresultandevaluatedtheaccuracy.TheresultsshowedthatRMSEwas0.3845fromLSTscatterandLSTerrorrangedfrom一0.6℃to0.6oCbetweenretrievalandproduct.Analyzingthestatisticalvaluesfrommeteorologicalstations,the
8、meantemperaturesofproductstationsandretrievedbyalgorithmandobservationsfrommeteorologicalstationswere21.61℃.21.19%and21I38ccrespectively.RMSDonthehightemperatureregionfrommeteorolo