Small‐scale precipitation variability in the Alps: Climatology in comparison with semi‐idealized numerical simulations

This study examines small‐scale precipitation patterns in a north‐Alpine region, and their dependence on the freezing level and on the crest‐level (700 hPa) wind direction and speed. On the one hand, measurements from a uniquely dense operational rain‐gauge network are analyzed for a period of 15 years (1991‐‐2005). Information on the ambient atmospheric fields was extracted from climate‐mode MM5 simulations driven with ECMWF (re‐analysis data. On the other hand, high‐resolution semi‐idealized MM5 simulations have been conducted, combining realistic topography with idealized atmospheric fields. The atmospheric flow parameters have been chosen to be representative of those used to classify the observational data, focusing on atmospheric conditions conducive to stratiform, orographically enhanced precipitation in the region under consideration. The results of the data analysis indicate a pronounced tendency for local precipitation maxima in the lee of individual mountain ridges, whereas the variability between stations in the centre of wider valleys and stations on the windward foot of individual ridges is comparatively small. This points towards a strong contribution of local precipitation enhancement due to the seeder‐‐feeder mechanism, combined with downstream advection of the precipitating hydrometeors by the ambient winds. The data analysis also reveals that strong winds and high temperatures tend to shift the precipitation field towards the interior of the Alps, whereas low temperatures and weak winds favour precipitation maxima near the northern edge of the Alps. The semi‐idealized simulations are consistent with these findings, but their quantitative agreement with the observed precipitation patterns depends on the ambient flow conditions. The closest agreement is found for atmospheric conditions conducive to strong orographic lifting, for which our present idealized flow fields were designed. Lower skill is obtained for conditions not dominated by orographic lifting, which implies that future work should include a generalization of the idealized flow fields. Nevertheless, precipitation patterns generated with semi‐idealized simulations seem to be very promising to support the spatial interpolation of point measurements (such as are needed for precipitation climatologies), which currently is usually based on statistical methods rather than physically motivated structures.