森林冠层叶面积指数遥感反演——以小兴安岭五营林区为例  

Retrieving forest canopy LAI from remote sensing data: A case study over Wuying forest in the Lesser Khingan

作　　者：刘振波[1,2] 刘杰[1,2] 

机构地区：[1]南京信息工程大学气象灾害省部共建教育部重点实验室,南京210044 [2]南京信息工程大学地理与遥感学院,南京210044

出　　处：《生态学杂志》2015年第7期1930-1936,共7页Chinese Journal of Ecology

基　　金：江苏省基础研究计划(自然科学基金)项目(BK20130992);国家重点基础研究发展计划项目(2010CB950701);江苏省高校优势学科建设工程项目资助

摘　　要：以中国东北小兴安岭五营林区为研究区，基于MODISBRDF遥感模型参数产品数据，首先利用4-Scale模型建立查找表计算像元尺度上各组分比例，估算研究区森林乔木冠层反射率，然后利用冠层反射率数据，获取研究区3种常用森林冠层植被指数，最后基于植被指数与实测叶面积指数构建研究区冠层叶面积指数反演模型，并选取最优模型实现研究区森林冠层叶面积指数反演。结果表明：研究区冠层LAI遥感反演模型中，基于比值植被指数sR（simpleratio，SR）构建的二次多项式反演模型精度最高，且反演精度比未考虑背景反射影响的SR反演模型精度有较大幅度提高，模型决定系数由0．58提高至0．54；反演获取的研究区冠层LAI在2．38～12．67，平均值6．52，LAI值在阔叶林区域相对较高。In this study, approaches of retrieving torest canopy LAI （leaf area index） were inves- tigated using remote sensing data over Wuying forest in the Lesser Khingan as a case study. We firstly calculated the forest canopy reflectivity using 4-Scale model in combination of MODIS （Moderate Resolution Imaging Spectroradiom6ter） BRDF （bidirectional reflectance distribution function） product. And then three well.known vegetation indices （VIs） were obtained from the forest canopy reflectivity. Finally, four models were examined in the forest canopy LAI retrieving using the canopy VIs and on-site canopy LAI measurements. The results showed that the model using quadratic polynomial and simple ratio （SR） VI was the best LAI retrieving model among the four models. Furthermore, the accuracy of LAI estimate could be improved using canopy reflectivity instead of the whole surface reflectivity, with the coefficient of determination （R2） increasing from 0.38 to 0.54. The mean canopy LAI in the study area ranges from 2.38 to 12.67, with an average of 6.52, and the relatively higher LAI values were found in the deciduous forest area.

关 键 词：叶面积指数 MODIS BRDF 4-Scale模型 冠层反射率 

分 类 号：S771.8[农业科学—森林工程]