中文摘要：

利用高光谱仪，对新疆棉花（2个品种4种配制方式）生物量和高光谱数据进行分析，经过多元统计分析与光谱微分处理，建立了基于比值植被指数（RVI）和归一化植被指数（NDVI）的5种函数形式的棉花干物质积累估测模型，相关系数均达到了极显著水平（a=1％，n=16）。基于RVI与NDVI构建的估测模型，后者比前者有更高的估测精度，指数函数、对数函数和双曲线函数的模型可以产生较高的估测精度；一阶微分光谱数据与棉花干物质积累量的逐步回归相关分析表明，相关系数的最高值发生在756nm（r=-0．6749，n=16），由756nm波段处的微分数值建立的回归模型，估测精度较高，具有实际应用的潜力。

英文摘要：

This paper presented an estimation model of cotton dry matter accumulation which was based on five functions of RVI and NDVI by using the hyper-speetrometer, analysing the biomass and hypersepectral data for cotton of Xinjiang and processing the spectral differential. The correlation coefficient of this model arrived extremely significant level（a=1%,n=16）. Analysing of correlation between the first derivative spectra data and dry matter yield, it showed the maximum value of correlation coeffi- cient（r=-0. 6749 , n= 16） in waveband 756 nm. The model built by using derivative spectra data at waveband 756 nm could provide higher accuracy, so it showed potential capacity in application.