中文摘要：

利用PROSAIL模型和地面实测数据分别探讨了冠层株型特征对建立VIs—LAI拟合模型的影响，并基于Beer定律分析了消光系数KVI随株型的变化特征，其中共考虑了3种株型品种（披散型、中间型、紧凑型），VIs则选用了增强型植被指数（EVI）、归一化植被指数（NDW）以及比值植被指数（RVI）。结果表明，不同VIs与LAI间的相关性程度会受到株型特征的影响，而同一株型内部VIs与LAI的相关性较混合株型好。不同植被指数比较，EVI-LAI相关关系受株型的影响较大，划分株型后其复相关系数R^2可提高30％以上；而NDVI—LAI相关关系受株型的影响较小，并且这种影响仅在LAI〈3时略有表现，可能与NDVI在较高LAI时的饱和有关；RVI—LAI相关关系对株型变化的反映则稍强于NDVI，且当株型趋于紧凑型时，其拟合模型有从指数型向直线型转化的趋势。此外消光系数KVI会随株型趋于紧凑而降低，3种植被指数的消光系数符合：KNDVI〉KEVI〉KRVI。

英文摘要：

The variation of empirical LAI-VI relationship and the variation of the extinction coefficient （KVI） derived from Beer＇s law were analyzed by using PROSAIL model and canopy spectrum data across different canopy geometry varieties （erective varieties, middle varieties and horizontal varieties）. Reflectance of blue （B） at 450 nm, red （R） at 670 nm, near infrared （NIR） at 830 nm were chosen to calculate 3 kinds of VIs （Normalized Difference Vegetation Index-NDVI, Ratio Vegetation Index-RVI, Enhanced Vegetation Index- EVI）. The results showed that the relationships between VIs and LAI were affected by canopy geometries, and they should be developed separately for different geometry types. However, the influencing degrees of canopy geometries were different among NDVI, RVI and EVI. In particular, the EVI - LAI relationship could be improved significantly （R^2 increased over 30 % ） when canopy geometries were considered. The difference of NDVI - LAI relationship among different wheat geometries was slim and was displayed only when LAI was lower than 3. The RVI - LAI relationship was more sensitive than NDVI - LAI relationship. When the canopy geometry was erective, the estimated model of RVI and LAI could be changed from exponential form to linear form, and the value of extinction coefficient （KVI） would be reduced. It was also showed that the sequence of KVI was KNDVI 〉 KEv1 〉 KRVI. Therefore, the influence of canopy geometry structures should not be ignored in studying the relationship between LAI and VIs for different crop geometry varieties.