Precipitation data downscaling at local scale and their verification over the Valle d'Aosta

The first part of the study was focused on the analysis of the global climate model projections included in the IV IPCC assessment report and regional climate model projections for Europe from specific projects such as PRUDENCE and ENSAMBLES. A feature common to all projections is a global increase of the average temperature, larger over the Alpine region, and a variation of the precipitation quantity and spatial distribution. Furthermore, analyses performed by means of regional climate models (RCM) for the Alps sectors outline a double pattern for precipitation levels: in the northern sector, models forecast an increase in precipitation levels, whereas for the southern sectors precipitation will decrease. The limit between these sectors is expected to move during the year: over the Alps, in winter, and over the central Europe, in summer. Such pattern seems to indicate that, for the future climate at local scale, the Italian Alps will experience a general increase in temperature and a small increase (drastic reduction) of precipitation in winter (in summer). These changes, combined together, result in an increase in rainfall (also during the cold season) and evapotranspiration. Also, melting of the snow cover is expected to occur one month earlier than at present. As a consequence, it is possible to expect in the future climate an increment of flooding risk, due to the approaching of melting snow period with the peak of spring precipitation, and of drought risk, because of the diminution of precipitation and the evapotranspiration increase. Therefore, the study about extreme events and related hydro-geological phenomena, whose frequency and magnitude are projected in increase, deserve a paramount importance. To optimize this kind of analysis, it is necessary to consider that the definition of precipitation change shows less agreement than temperature predictions, due to its higher interannual variability, and the same consideration can be extended to extreme events versus average values. This fact does not surprise, especially considering the difficulty to represent phenomena that occur at smaller scales than those used in global climate models (GCM), such as those which are responsible for intense rainfall.