CR 62:173-188 (2015)  -  DOI: https://doi.org/10.3354/cr01256

ABSTRACT: We compare the output of a high-resolution regional climate model (RCM) with 40 yr of weather data from 64 weather stations from the Italian National Air Force network. Climatic zones were identified using the Ward’s method for cluster analysis for minimum and maximum temperature and rainfall. The model is able to generate realistic spatial patterns of the observed clusters, although with different skills depending on the considered variable. The closest match between model and observations is for daily minimum temperature. The maximum temperature shows an unrealistic summer peak for most of the clusters. The model also consistently produces too many strong warm events during summer, and shows a tendency to overestimate total rainfall, especially during spring and early summer. The frequency and intensity of extreme events are well captured only for minimum temperature for Alpine and mountain weather stations. In spite of these biases and mismatches, this study shows, by means of a direct comparison with weather station data, that even in an area of complex orography and land-sea contrast such as the Mediterranean area, RCMs are able to produce realistic climate patterns at spatial scales that the global driver cannot capture.

KEY WORDS: Downscaling · Regional climate model · Climate model validation