基于CHRIS数据的新型植被指数的LAI估算研究

The Study of LAI Estimation Using a New Vegetation Index Based on CHRIS Data

叶面积指数(LAI)作为重要的植被冠层结构参数,对其进行正确估算一直是遥感应用研究的重点。CHRIS/PROBA是目前具有较高分辨率(17m)的高光谱多角度数据,该数据在反演LAI方面有着重要的应用价值。本次研究应用辐射传输ACRM模型来模拟一系列LAI在不同观测天顶角(-80°～+80°)情况下的植被光谱数据,在此基础上利用红波段和近红外波段构建了一个新型高光谱多角度植被指数HDVI,并成功地应用于CHRIS/PROBA数据对LAI的估算。结果表明:(1)相比光谱指数NDVI和多角度指数HDS,新指数能更好地利用光谱和多角度双重信息,与研究区LAI有着更好的相关性,决定系数R2高达0.734 7。(2)利用LAI-HDVI最优拟合方程关系来估测LAI值,得到了研究区的LAI分布图,LAI估算精度均方根误差RMSE为0.619 8。

Leaf area index（LAI） is an important structural parameter of vegetation canopy,the correct estimation of which has been the focus in the remote sensing community.As a kind of hyperspectral and multi-angle remote sensing data with higher resolution（17 m）,PROBA/CHRIS has significant application value in LAI inversion.In the present paper,the analytical two-layer canopy reflectance model（ACRM） was used to simulate a series of reflectances with different LAI values.Based on this,a new vegetation index was built and successfully applied to LAI inversion of PROBA/CHRIS image data.Our results indicated that： compared with the spectral index NDVI and multi-angle index HDS,the new index could make better use of spectrum and multi-angle messages and have a better correlation with LAI of the study area;moreover,the correlation coefficient R2 reached up to 0.734 7.And in order to obtain the figure of LAI distribution of the study area,we used the optimal fit equation between LAI and HDVI to estimate LAI,and the accuracy of the RMSE was 0.619 8.