Integration of remote sensing and GIS for mapping and spatial modelling vegetation in the Supramonte mountain area (Central East Sardinia)

An application of an integrated approach for mapping the vegetation of Supramonte di Orgosolo, a Sardinian mountain area on limestone and siliceous substrata, is presented herewith. The spectral unmixing on satellite data has given reliable quantitative information over the vegetation cover and distribution pattern over the whole study area. The digital image processing on color air photographs has been a good quantitative measure of soil cover density, ranging from values 5% to 100%. The main result has been the semi-automatic production of the physiognomic-structural vegetation map of the study area, distinguishing nine classes over the 15 assessed in the field surveys. This is due to the limitations and constrains of the used dataset and to the reasons discussed in the paper, with some suggestions about future improvement of the methodology. © 2005, Universita degli Studi di Roma 'La Sapienza'. All rights reserved.