Leaf area index retrieval from remotely sensed hyperspectral data

An experimental leaf area index (LAI) retrieval model was proposed with the aid of a leaf- radiative transfer model (PROSPECT) and a canopy bidirectional reflectance model (SAILH) to simulate the canopy reflectance in this paper. Then, the vegetation indices (VIs) were introduced, and the sensitivities were analyzed between LAI and VIs, soil background. Based on the sensitivity analysis, a modified chlorophyll ratio index II (MCARI2) was proposed by Haboudane et al. (2004) was used to build the LAI retrieval model, because it is rather sensitive to the LAI and insensitive to soil background. Finally, the retrieval model proposed was performed to estimate LAI from the hyperspectral data. Compared with the ground-measured LAI, the LAI retrieved from hyperspectral data underestimate approximately 0.42.