APPLICATION OF REMOTE-SENSING AND MULTIVARIATE-ANALYSIS TO THE AGROCLIMATIC CHARACTERIZATION OF A MEDITERRANEAN REGION

This paper presents a climatological application of the combination of remote sensing data and multivariate analysis. It proposes a methodol ogy to perform an agroclimatic characterization of a region with the a im of determining homogeneous areas. This methodology is applied to Ca talonia, in the northwest of the Mediterranean Sea, in the Iberian. Pe ninsula. To this end various multivariate analysis techniques have bee n applied to data from 107 meteorological stations and to digitally pr ocessed AVHRR images from a NOAA satellite (ground surface temperature and thermal inertia). Values of emissivity and albedo for different l and uses and phenomenological states, in addition to the most suitable algorithms for the soil surface temperature, have been obtained. Comb ining this information with orographic characteristics and the Papadak is classification, a final map is obtained with 68 homogeneous zones f or the period between April and October, the most important for the ag riculture of the area studied.