Application conditions and impact factors for various vegetation indices in constructing the LAI seasonal trajectory over different vegetation types

Abstract

Abstract Leaf area index (LAI) is a required input for various ecological and crop models. To investigate the application conditions of various vegetation indices (VIs), especially the VIs constructed by red-edge band (VIRE) for estimating LAI, six VIs derived from Medium Resolution Imaging Spectrometer (MERIS) data were used to construct LAI seasonal trajectory for different vegetation types at 15 sites. The PROSAIL model combined with the Extended Fourier Amplitude Sensitivity Test (EFAST) method was adopted to explore the influences and physical basis of canopy biophysical and non-canopy variables on the construction of LAI seasonal trajectory using VIs. For deciduous forests, the normalized difference vegetation index (NDVI) had the highest sensitivity to LAI when LAI 2. For evergreen forests, there were no obvious differences among the sensitivities of six VIs to LAI when LAI 5. For crops, all the VIs had the similar sensitivities at LAI 3. For all three types of vegetation, the VIRE maintained relatively high sensitivity to LAI over the whole range of LAI, especially at high LAI values. The VIs were most affected by chlorophyll content (Cab) and average leaf inclination angle (ALA); their total contribution was about 85%. However, the influence of ALA on VIRE was relatively weak, implying that the VIRE had the potential to establish a universal model for LAI estimation among different vegetation types. Therefore, the optimal VIs over different ranges of LAI were suggested to estimate LAI. In addition, the VIRE should be a preferred choice for estimating LAI to reduce the simulation errors of seasonal LAI, if the RE band is available.