Rangeland monitoring using hyperspectral remote sensing data and spectral mixture analysis

Abstract

The abundance mapping of rangeland plant communities using hyperspectral remote sensing data is investigated. The cover fraction of five rangeland components (green grass, yellow grass, litter, shrubs and soil) was estimated using constrained Spectral Mixture Analysis (SMA). Three sets of endmembers were assessed. The first set labeled the laboratory endmember refers to the laboratory reflectance measurements of different samples of leaves. The second set called field endmember is based on field spectra, while the third identifiled as modeled endmember was simulated using a canopy reflectance model. The three sets were applied to CHRIS/PROBA data using SMA to estimate the rangeland component cover fractions. Performances of these endmember sets were evaluated based on the field knowledge of the area of interest.