A multi-model view of future extreme precipitation in the Greater Alpine Region

The Greater Alpine Region (GAR) stands out as a hotspot for hydroclimatic extremes, where the interplay between complex orography and atmospheric circulation enhances the occurrence of intense precipitation events. As climate change progresses, understanding how these extremes are represented in climate models — and how they may evolve — is crucial for anticipating future hydrological risks in mountainous areas and downstream lowlands, as well as for managing water resources. This work explores daily precipitation extremes using a multi-model, multi-dataset approach. Alongside ERA5, we incorporate high-resolution gridded datasets to benchmark simulations from 29 CMIP6 global models and 18 EURO-CORDEX regional models. Extreme precipitation is characterized through the suite of ten indices defined by the ETCCDI, applied consistently across datasets. Our evaluation focuses on both the statistical properties of precipitation extremes and their spatial patterns across the GAR, with particular emphasis on how model resolution and orographic complexity influence skill. Regional simulations with higher resolution are further analyzed over the Piedmont domain, where terrain-driven variability is especially pronounced. We apply a clustering technique to identify coherent model behaviors and select those that best reproduce observed extremes. The analysis reveals a subset of models that outperforms the broader ensemble, especially in capturing spatial signals shaped by orography. In addition, a comparison is made between regional results in the Piedmont domain and projections derived from a trend-based scaling of ERA5, allowing us to assess how model projections perform in relation to observed extreme precipitation. This refined selection provides a more credible basis for projecting future changes in extreme precipitation over the Alpine region.