简介：以海南后海浅水区域为研究区,分析了Landsat8OLI多光谱数据和实测水深值的关系,建立了OLI数据单波段反射率和反射率比值的浅海水深反演方法.研究表明：OLI遥感影像的第1、第2和第3波段反射率和水深有较好的相关性;研究区17米水深以浅,第2和第3波段的反射率比值与水深具有最佳的线性关系,决定系数达到0.93,平均相对误差仅为6.42%.利用OLI数据结合少量的实测水深数据可有效的获取海岸带浅水区域的水下地形信息.

简介：Weusedgeographicinformationsystemapplicationsandstatisticalanalysestoclassifyyoung,prematureforestareasinsoutheasternGeorgiausingcombineddatafromLandsatTM5satelliteimageryandgroundinventorydata.Wedefinedprematurestandsasforestswithtreesupto15yearsold.Weestimatedtheprematureforestareasusingthreemethods:maximumlikelihoodclassification(MLC),regressionanalysis,andk-nearestneighbor(kNN)modeling.Overallaccuracy(OA)ofclassifyingtheprematureforestusingMLCwas82%andtheKappacoefficientofagreementwas0.63,whichwasthehighestamongthemethodsthatwehavetested.ThekNNapproachrankedsecondinaccuracywithOAof61%andaKappacoefficientofagreementof0.22.RegressionanalysisyieldedanOAof57%andaKappacoefficientof0.14.WeconcludethatLandsatimagerycanbeeffectivelyusedforestimatingprematureforestareasincombinationwithimageprocessingclassifierssuchasMLC.