中文摘要：

叶面积指数（Leaf area index,LAI）是陆地生态系统的一个十分重要的结构参数。随着空间精细化模型的发展和基于过程的分布式模拟技术的应用,对LAI的区域估算显得越来越重要,但目前尚缺乏有效的估算手段。该项研究以青海云杉（Picea crassifolia）林为研究对象,利用LAI-2000冠层分析仪、鱼眼镜头法和经验公式法对林冠层LAI进行了测定,观测值分别为1.03～3.70、0.48～2.26和2.27～8.20,显然,仪器测定值偏低。针对针叶的集聚效应导致仪器测定值偏低的现象,利用跟踪辐射与冠层结构测量仪（TRAC）测定的青海云杉林聚集系数计算调整系数,对鱼眼镜头法获取的LAI值进行订正。根据高分辨率的遥感数据反演青海云杉林的植被指数与LAI的关系,最后获得了较合理的该地区林冠层LAI的空间分布图。

英文摘要：

Aims There is increasing need for regional estimates of leaf area index （LAI） because it is an essential input for many eco-hydrological processes models; however,there has been a lack of effec-tive methods for its estimation. Picea crassifolia is the dominant species in the forest ecosystem of Qi-lian Mountains and is critical to the eco-hydrological processes of the ecosystem. Our objective is to compare methods for determining canopy LAI of this P. crassifolia forest. Methods We investigated canopy LAI with an LAI-2000 canopy analyzer, hemispherical photography and allometric regression on tree height and diameter at breast height （DBH）. The value was underesti- mated by the two instruments because of clumping in the conifer forest. In order to adjust the LAI measured and obtain the spatial distribution of LAI, we first measured the clumping index by Tracing Radiation and Architecture of Canopies （TRAC）. Then we calculated the adjusting coefficient by the clumping index, which was used to adjust the LAI value measured by hemispherical photography. Then, we determined the relationship between adjusted LAI and vegetation indexes retrieved by high resolution remote sensing data （Quickl3ird） and estimated the spatial distribution of canopy LAL Important findings The values of canopy LAI were 1.03-3.70 by LAI-2000 canopy analyzer, 0.48-2.26 by hemispherical photography, and 2.27-8.20 by allometric regression. LAI value by al- lometric regression on tree height and DBH was used for assessing the measurement accuracy by the other two indirect measurement techniques. We found the two instruments （LAI-2000 canopy analyzer and hemispherical photography） under estimated the canopy LAI of the forest by about 3.14-3.86 times. We built statistical models between adjusted LA1 and vegetation indexes and selected the optimal model, i.e., correlation between normalized difference vegetation index and LAI, through validating.