High-Resolution Worldview-2 dune systems sparse vegetation monitoring, by modeling leaf optical properties and directional reflectance of the vegetation

Abstract

Vegetation mapping is a priority element for the management of natural protected areas. In this context, very high-resolution satellite remote sensing data can be fundamental to provide accurate vegetation cartography at species level. Specifically, the analysis has been carried out using WorldView-2 (WV-2) imagery, which offers high spatial and spectral resolutions. In this paper, the SVM classification is properly applied to biophysical and structural vegetation indicators retrieved after an analysis of the most sensitive PROSAIL parameters to the WV-2 bands and applying the inversion of the model. The classification with these parameters have evidence robust results to discriminate different species with environmental interest. A field radiometer is used to characterize the different plant species and validate the results.