The dataset consists of 5 land cover maps rasters of Mont Avic watershed (5500 ha) for the past period and of 2 land cover simulations for the future.
The 7 time steps are: 1965, 1975, 1988, 2006, 2017 (past); 2050, 2100 (future).
Simulations were obtained through an integration of Markov Chain and Multi-Layer Perceptron models by comparing 1965 and 2017 maps.
Mont Avic Park represents a subalpine forest landscape in Northwestern Italy with an intense history of forest exploitation for mining activities.
Metadata
Cell size (X,Y): 10, 10 meters
Coordinate System: WGS84 UTM Zone 32N (EPSG: 32632)
NoData Value: 255
Data type: 8 bit unsigned integer
Format: GeoTIFF
Compression: LZW
Land cover classification: object-oriented segmentation and video photointerpretation (Garbarino et al. 2013)
Land class values: 1 = dense forest; 2 = sparse forest and shrubs; 3 = grassland; 4 = human; 5 = unvegetated (Future projections lack in human land cover class)