本文研究了LAI产品真实性检验的指标和方法，建立了LAI产品真实性检验的流程，将遥感产品真实性检验误差分解为模型误差、数据定量化差异和尺度效应3个方面。以内蒙古锡林浩特草原为研究区，结合实测数据和Landsat TM数据建立NDVI-LAI模型，得到LAI验证参考“真值”，据此“真值”按照本文的流程对MODIS LAI产品进行验证，分析了研究区MODIS LAI产品真实性检验的误差来源。研究表明，该研究区的MODIS LAI（MOD15A2）产品相对高估约25%。各个误差因素中，LAI遥感模型差异对于结果影响最大，MODIS LAI模型高估了该区域草地LAI（高估约44.2%）；数据定量差异的影响也比较大，MODIS地表反射率数据与Landsat TM地表反射率数据的差异造成了约16.2%的低估；尺度效应的影响较小，造成约3.1%的低估，其中NDVI-LAI模型的尺度效应带来2.4%的低估，NDVI数据的尺度效应造成约0.7%的低估。

The Leaf Area Index (LAI) is essential in the remote sensing of vegetation, and is widely used in quantitative remote sensing products. LAI products should be validated to enable their effective applications in various fields. In this study, the indicators and methods of LAI product validation were investigated, and a process for LAI product validation was established. The grassland in Xilinhot, Inner Mongolia of China was selected as the study area. Landsat TM data and ground measurements were used to build the Normalized Difference Vegetation Index (NDVI)-LAI model. Then, the "true" LAI reference for validation was established. The MODIS LAI product was also validated in this study according to the reference LAI and the proposed model. In addition, the sources of error in the MODIS LAI product were analyzed. Results showed that the MODIS LAI product (i.e., MOD15A2) in the study area was overestimated by approximately 25%. Among the factors that affected the error, the difference in LAI remote sensing models was the most dominant, leading to an overestimation of 44.2% in the MODIS LAI model. Meanwhile, image quantitative difference exerted a significant influence, causing an underestimation of approximately 16.2%. In addition to these two factors, the scale effect produced an underestimation of 3.1%, including an underestimation of 2.4% from the scale effect of the NDVI-LAI model, and an underestimation of 0.7% from the scale effect of NDVI.