Optimization of Spectral Indices for the Estimation of Leaf Area Index Based on Sentinel-2 Multispectral Imagery

摘要

Spectral vegetation indices are powerful tools in statistically estimating leaf area index (LAI) with remotely sensed imagery. However, the band selection in some generic vegetation indices influenced their performance to a great extent due to the rapid development of new sensors. As the latest launched satellite carried with multispectral sensors, Sentinel-2 provides 3 extra red-edge bands and 1 extra SWIR band. For the purpose of statistical LAI retrieval based on Sentinel-2 data, the optimal bands combination of the DVI, SR and NDVI-formed spectral indices were selected on the basis of correlation analysis. The experiment results demonstrated that band 7 (red-edge) and 8 (near-infrared) of Sentinel-2 MSI data were the optimal bands combination of LAI retrieval. And finally the optimal spectral index were validate with the in-situ LAI observations, which performed satisfied estimated accuracy.