Potential and limitations of convection‐permitting <scp>CNRM‐AROME</scp> climate modelling in the French Alps

Convection-permitting climate modelling is a promising avenue for change research and services especially in mountainous regions. Work required to evaluate the results of high-resolution simulations against relevant observations, put them broader context coarser resolution frameworks. Here we numerical with convection-permitting regional model CNRM-AROME ran at 2.5 km horizontal over large pan-Alpine domain European Alps, using either ERA-Interim or output as boundary conditions. This study analyses annual seasonal characteristics 2 m temperature, total precipitation, solid fraction precipitation snow depth scale French Alps under past future The are compared local reanalysis S2M, raw adjusted, ADAMONT method, CNRM-ALADIN driven by CNRM-CM5 global model. highlights generally similar differences between datasets, well obstacles use some outputs they stand. These consist excessive accumulation on ground above 1,800 a.s.l., lower temperature values same elevations than S2M ADAMONT-adjusted outputs. Besides these obstacles, present several advantages Among them, significantly smaller cold bias, more realistic accumulated precipitations, better representation spatial variability different variables investigated, which always stand closer reference data As suggested many studies, could even produce precipitations high elevation taken our consequently projections, but lack reliable set observations remains an obstacle their evaluation.