ISPRS International Journal of Geo-Information · 2017 · 39 citations · 41 references
Natural SciencesEarth SciencesForestryForest EcologyForest InventoryRemote SensingLand Cover ClassificationLand Cover MappingRemote Sensing For Forest InventoryRemote Sensing Image AnalysisRemote Sensing ImageryVegetation MonitoringEcosystem ServicesForest Cover ClassificationForest Cover Type MappingForest Type MappingRemote Sensing ClassificationRemote Sensing-based Species DetectionRemote Sensing-based Species MappingTree Species Classification
The studies of forest ecosystems from remotely-sensed data are of great interest to researchers because of ecosystem services provided by this ecosystem, including protection of soils and vegetation, climate stabilization, and regulation of the hydrological cycle. In this context, our study focused on the use of a spectral angle mapper (SAM) classification method for mapping species in the Azrou Forest, Central Middle Atlas, Morocco. A Sentinel-2A image combined with ground reference data were used in this research. Four classes (holm oak, cedar forest, bare soil, and others-unclassified) were selected; they represent, respectively, 27, 11, 24, and 38% of the study area. The overall accuracy of classification was estimated to be around 99.72%. This work explored the potential of the SAM classification combined with Sentinel-2A data for mapping land cover in the Azrou Forest ecosystem.
41