中文摘要：

在深入探讨目前广泛使用的提取叶绿素含量的植被指数的光谱响应机制基础上，利用PROSPECT＋SAIL模型模拟的作物冠层反射率样本数据对比分析了这些植被指数对叶绿素含量变化的敏感性差异，包括PSSRa、PSSRb、PSNDa、PSNDb、NPCI、PRI、MCARI和TVI等。结果表明，上述植被指数或对土壤背景变化敏感，或受高值LAI影响趋于饱和，对作物叶绿素含量反演效果均不理想。提出了4种基于TVI和MCARI的改进植被指数MTVI1、MTVI2、MCARI1和MCARI2，揭示了它们对土壤背景和LAI不敏感，对叶绿素含量变化更为敏感的光谱机制，并根据实验数据对其进行验证。实验表明，改进的植被指数MTVI2和MCARI2是作物冠层叶绿素含量较好的预测器，可据此建立作物冠层叶绿素含量反演模型。

英文摘要：

Based on the thorough analysis of the spectral response mechanism of vegetation indices, which are used widely to retrieve chlorophyll content at present, taking advantage of sample reflectance data of crop canopy, simulated by models PROSPECT ＋ SAIL, the sensitivity differences of these vegetation indices for chlorophyll content are compared and discussed, including Pigment-Specific Simple Ratio a （PSSRa）, Pigment-Specific Simple Ratio b （PSSRb） , Pigment-Specific Normalized Difference a （PSNDa） , Pigment-Specific Normalized Difference b （PSNDb）, Normalized Pigments Chlorophyll ratio Index （NPCI）, Pigment Ratio Index （PRI）, Modified Chlorophyll Absorption Ratio Index （MCARI） and Triangular Vegetation Index （TVI）. The analysis results indicated that the above vegetation indices were either sensitive to changes of soil background or affected by saturation at high LAI levels. Therefore, it is not an ideal way to retrieve chlorophyll content by these vegetation indices. Then, four new kinds of modified spectral indices based on TVI and MCARI are proposed, including Modified Triangular Vegetation Index 1 （ MTVI1 ）, Modified Triangular Vegetation Index 2 （ MTVI2）, Modified Chlorophyll Absorption Ratio Index 1 （MCARI1） and Modified Chlorophyll Absorption Ratio Index 2 （MCARI2）. The spectral response mechanism of four modified spectral indices was disclosed that they more sensitive to the variation of chlorophyll content, validated with the experimental data. The experiment are both less sensitive to soil background and green LAI and and their performances of chlorophyll content retrieval were demonstrated that two modified spectral indices, MTVI2 and MCARI2, are proved to be the better predictors for chlorophyll content, and the retrieval model of chlorophyll con tent of crop canopy can be established based on them.