Skill of CMIP6 decadal climate predictions at the subregional scale

Effective adaptation and mitigation strategies rely on climate model projections capable of exploring future scenarios across multiple timescales. While long-term climate projections are well established, there is growing demand for reliable short-term decadal simulations. This study aims to assess the skill of a CMIP6 climate model from the Decadal Climate Prediction Project, focusing on its ability to simulate subregional climate conditions, an essential requirement for supporting local decision-making in sectors such as agriculture, water management, and infrastructure planning. The analysis targets the Emilia-Romagna region in northern Italy, an area characterized by complex topography and diverse climate conditions, and examines precipitation, minimum, and maximum temperatures. Coarse-resolution model data are first regridded using second-order conservative interpolation to enhance spatial detail and match the resolution of the available observational dataset. Model drift is then assessed and corrected to improve predictive accuracy. Results reveal significant temporal and spatial variability in model performance across the region, with the uncorrected model generally exhibiting limited skill in reproducing observed climate conditions. While drift correction improves performance, substantial uncertainty and challenges remain, particularly in capturing precipitation patterns. These findings highlight the need for caution when using global climate models for decadal predictions intended to inform impact assessments at the subregional scale.