冬小麦叶面积指数的高光谱估算模型研究

本文以山东禹城为研究区,利用地面实测光谱数据,探讨不同植被指数和红边参数建立高光谱模型反演冬小麦叶面积指数的精度。通过逐波段分析计算了4种植被指数(NDVI、RVI、SAVI、EVI),结合同步观测LAI数据,确定反演叶面积指数的最优波段;计算了5种常用的高光谱植被指数MCARI、MCARI2、OSAVI、MTVI2、MSAVI2,同时利用4种常用方法计算红边位置和红谷,与实测LAI进行回归分析,比较植被指数和红边参数模型对冬小麦LAI的估测精度。结果表明各因子与LAI均具有较高的相关性,整个研究区归一化植被指数具有最高的反演精度,确定了估算冬小麦LAI的最优模型,并使用独立的LAI观测数据对模型进行了验证。

Hyperspectral estimation models for leaf area index of winter wheat

Based on the hyperspectral ground survey data from Yucheng study area in Shandong province,the winter wheat LAI estimation precision was evaluated with hyperspectral model using different vegetation index and red edge.The optimum estimation band was confirmed after the four vegetation indexes(NDVI,RVI,SAVI,and EVI) were analyzed and calculated band by band with in-phase LAI.Five hyperspectral vegetation indexes(MCARI,MCARI2,OSAVI,MTVI2,and MSAVI2) were calculated,and the positions of red edge and red trough were detected by four common methods.All of them were regressed with the LAI,and the precision was evaluated.The result indicated the high correlation between those factors and LAI,and NDVI had good estimation precision in the region.The optimum model equation was put forward,which was validated by the independent LAI observation data.