The relation between circulation types and regional Alpine climate. Part II: the dependence of the predictive skill on the vertical level of the classification for Trentino

This study investigates the effect of varying the vertical level of a synoptic circulation classification on its predictive skill with respect to surface temperature, rainfall, solar radiation and wind in Trentino, a mountainous region in the South-Eastern Alps. A synoptic climatology based on the same data set and classification method presented in part I of the present article, in fact, showed that seasonal anomalies of mean daily temperature, daily rainfall, daily solar irradiation and mean daily wind intensity vary not only among weather types and seasons but also within the same type and season for different vertical levels. This analysis quantifies the ability of the method to classify synoptic circulation in classes associated with distributions of atmospheric variables different from climatology and to identify the occurrence of extreme events. The statistical metrics presented in the article demonstrate that the differences in predictive skill between classifications applied to distinct levels are comparable in magnitude to those between different atmospheric variables and seasons for the same level. The level of 500 hPa generally provides the largest predictive skill with respect to total daily rainfall, daily solar irradiation and daily temperature range in all seasons. On the other hand, the largest skill with respect to wind intensity is found for sea level pressure and 850 hPa circulation types. Large seasonal variations are also evident, with the colder seasons providing the largest predictive skill. Combinations of circulation types derived at two vertical levels increase the skill, although an optimal combination of levels could not be found even for the same atmospheric variable and season.