摘要
采用2011年8月获取的黄丰桥林场SPOT5数据为信息源,并同步开展现地样地调查,依据典型抽样,以不同的海拔、坡度、坡向进行选样,利用手持GPS和LAI-2000植物冠层分析仪,分别对选取的60块样地进行定位和叶面积指数测量。结合遥感数据和实地调查数据,对地理因子和遥感因子变量进行主成分分析,采用逐步回归法筛选出2个主成分建立多元回归方程,对该研究区域的植被叶面积指数进行模拟,精度达到84.17%。结果表明:RVI,NDVI,MSAVI,MCAVI和DVI与LAI之间存在较好的相关性。
By using the SPOT5 image in Huangfengqiao forest farm which has been get in August 2011, we sur- veyed the sample plots at the same time, chose plots through different elevations, slopes and aspects based on typical sampling, then located the 60 sample plots to measure the leaf area index by using the GPS and LAI - 2000 plant canopy analyzer. Based on the data of remote sensing and the sample plots, first to analyze eleven geographic and remote sensing factors which as input variables by using principal component analysis, and then a multiple regression equation was established by two principal components which were selected by stepwise re- gression, predicted the leaf area index of study area and the accuracy was 84.17%. Study results show that there has been a good correlation among RVI,NDVI, MSAVI, MCAVI, DVI and LAI.
出处
《中南林业调查规划》
2012年第2期24-27,38,共5页
Central South Forest Inventory and Planning
基金
国家林业局林业公益项目专题:林分结构与生长模拟技术研究(201104028)
湖南省高等学校科学研究项目"高分辨率遥感影像森林结构参数反演研究"(11C1313)
关键词
遥感
多元回归模型
参数反演
叶面积指数
SPOT5
remote sensing
multivariable regression model
parameter inversion
leaf area index
SPOT5
