DYNAMIC LAND COVER MAPPING EXPLOITING HYPERSPECTRAL PRISMA DATA

Earth Observation derived land cover products can providean advance insight into coastal ecosystem monitoring tocapture the fragmentation and changes of land surface. Thework addresses the testing of consolidated classificationalgorithms on hyperspectral data in synergy with SAR andfield data for the definition of biological and geophysicalcharacteristics of surface coverage components to produce adetailed Dynamic Land Cover map. We obtain reliable landcover and land use information identifying different types ofvegetation and soils condition and indicators ofbiogeophysical surface processes in the testing costal area ofthe Gulf of Oristano (Italy). The results offer furtherclassification of specific land cover targets representing acontribution for the production of land cover products atdifferent levels of thematic resolution to improvedownstream application services and to support land policiesand management.