A multivariate clustering approach for a gis based territorial characterization of the montepulciano d’abruzzo DOCG “Colline Teramane”

The aim of the project was to characterize the Premium Denomination of Guaranteed Origin (DOCG) “Colline Teramane” wine-growing region and to delineate and define homogeneous zones (terroir units) within it, by applying a multivariate clustering approach combined with geomatics. The inventory, characterization and classification of the land resources included components of climate (temperature and rainfall from meteorological stations), landform (Digital Elevation Model) and lithology (geolithologic map). Managing of environmental variables was performed using a GIS. From the environmental variables, vine-related derived indices (bioclimatic: Huglin index, cool night index, Riberau-Gayon-Peynaud index; and morphologic: Aspect, Topographic Wetness Index, Curvature, Slope, Incoming Solar Radiation) were calculated, spatialized and implemented to the GIS. Then, normalized variable values for each raster cell were use in a PCA followed by a multivariate clustering algorithm (Isodata) to obtain a continuous morpho-climatic map, in which each cluster represented a unit or zone. Finally, the morpho-climatic map obtained was overlaid with the geolithologic map. The result shows different morpho-climatic conditions located over different lithotypes.