Assessment and ranking of CMIP6-global climate models over the Calabria region (southern Italy)

General Circulation Models (GCMs) can be considered the most accurate numerical models for simulating climate variables across past and future scenarios. The reliability of GCMs projections can vary markedly across different geographical regions but, despite the importance of selecting the most suitable GCM for a specific study area, there has been limited exploration into identifying the optimal model, particularly concerning its implications at regional level. In this context, the present study proposes a framework for the classification of 34 different Coupled Model Intercomparison Project Phase 6 (CMIP6) GCMs, included in the NASA Earth Exchange Global Daily Downscaled Projections NEX-GDDP dataset, based on their ability to replicate observed precipitation in the Calabria region (southern Italy). With this aim, first daily precipitation data from 1990 to 2014 were considered to create a gridded dataset with a spatial resolution matching that of the climate models (0.25°). Then, to evaluate the performance of NEX-GDDP-CMIP6 precipitation data across the study area, two distinct sets of statistical metrics have been applied, one at daily scale and one at annual scale, respectively. Finally, the models performances have been evaluated by means of four different ranking models and the best one has been chosen by applying a meta-ranking approach. Different results for the different timescales have emerged, with the NESM3 model resulting the most reliable in reproducing daily precipitation (ranked as first in 3 ranking models), while the KACE-1‐0-G model has been identified the best for the analysis of precipitation data at annual scale in the study area (ranked as first in 2 ranking models and second in another one). At any rate, in order to reproduce both daily and annual data, the INM-CM4-8 model can be considered the best one given its good performance for both the timescales. In fact, it has been ranked as second overall considering the meta-ranking approach and, in particular, it was ranked as second in 1 ranking model and fourth in the remaining ones at daily scale and second in 2 ranking models at annual data.