基于遥感影像和NDVI阈值法的银川市植被覆盖度反演与监测.docx
文本预览下载声明
基于遥感影像和NDVI阈值法的银川市植被覆盖度反演与监测方杰诗1,2,周晓霖1,2(1.中山大学地理科学与规划学院,广州510275;2.香港理工大学建设与环境学院,香港)摘要:充分考虑由于大气、时间、土壤和树种组成、郁闭度等因素造成的干扰,通过对遥感影像进行精确的大气校正,运用混合像元法中的像元二分模型思想和归一化差值指数结合的NDVI阈值法模型,反演出多时相的区域内的植被覆盖度空间分布信息;最后将植被覆盖度进行分级,方便了定性比较分析区域内植被覆盖度的情况和变化,为监测提供了便利。以宁夏回族自治区银川市为研究区,采用了2002、2006和2010年的LandsatTM影像,结合区域内权威土地利用数据,对比分析出模型结果精度,最后从单时相和多时相两方面针对银川市进行了植被覆盖分析,并指出城市化建设和河流两方面对植物覆盖变化产生的影响。研究结果证明,在宁夏银川应用该模型是有效的。关键词:像元二分法;NDVI阈值法;植被覆盖度;LandsatTM影像中图分类号:TP7文献标识码:AInversionandMonitoringofVegetationCoverageinYinchuanCityBasedonRemoteSensingImagesandNDVIThresholdMethodFANGJie-shi1,2,ZHOUXiao-lin1,2(1.SchoolofGeographicScienceandPlanning,SunYat-senUniversity,Guangzhou510275,China;2.DepartmentofLandSurveyingandGeo-Informatics,HongKongPolytechnicUniversity,HongKong,China)Abstract:Inthisstudy,preciseatmosphericcorrectionsonremotesensingimagesarefirstlydoneinordertoreducetheinterferenceduetosoil,treespeciescomposition,canopyandotherfactors.WiththeapplicationoftheNDVIthresholdmethodcombinedwiththedimidiatepixelmodelandnormalizeddifferencevegetationindex,spatialdistributioninformationofvegetationcoverageinmulti-phaseregionsisinversed.Finally,thevegetationcoverageresultsareclassified,whichisconvenientforqualitativecomparingandanalyzingtheconditionandchangeintheresearchareaandmonitoring.Inthispaper,YinchuancityofNingxiaHuiAutonomousRe-gionisselectedastheresearchareaandtheLandsatTMimagesof2002,2006and2010areused.Withthelandusedatafromtheauthority,theaccuracyofthemodelresultsiscomparedandanalyzed.Finally,vegetationcoverageinYinchuanisanalyzedfromtwoaspects,i.e.single-phaseandmulti-phase,andtheinfluencefromtheurbanizationconstructionandriversispointedout.There-searchresultsshowthattheNDVIthresholdmethodcombinedwithdimidiatepixelmodelandnormalizeddifferencevegetationindexiseffective.Keywords:dimidiatepixelmodel;NDVIthresholdmethod;vegetationcoverage;LandsatTMimages植被覆盖度是区域内植被垂直投影面积占地表面积的百分比[1],是重要的生态环境参数之一,其对衡量一个地区的生态环境和绿化程度有重要意义。地表实测是测量草地植被盖度的传统方法。按原理可将地表实测的方法分为3类,即采样法、仪器法和目视估测法[2]。由于地表实测的方法在地表植被调查中起着非
显示全部