Time modelling and spatial clustering of daily ambient temperature: an application in Southern Italy

Knowledge of the spatial and temporal patterns of meteoclimatic variables is one of the crucial points when dealing with environmental problems. For this purpose we analysed the temporal and spatial behaviour of ambient daily temperature values in order to single out the time dynamics of this variable and its correlation with landscape parameters. A method based on two techniques is implemented; the first describes the time dynamics using stochastic models and the second characterizes the homogeneous areas using a multivariate algorithm for spatial clustering. This approach allows us to use a powerful and parsimonious description of temporal and spatial patterns of daily ambient temperatures. Our analysis relates to temperatures but the approach is completely general and can be applied to other meteoclimatic variables. We checked the proposed method in the study of daily ambient temperature patterns in Southern Italy. The data, collected during the period from 1973 to 1988 in 50 thermometric stations located in two regions of Southern Italy (Basilicata and Calabria), are analysed.