Assessing Multidomain Overlaps and Grand Ensemble Generation in CORDEX Regional Projections

The Coordinated Regional Climate Downscaling Experiment (CORDEX) initiative has made available an enormous amount of regional climate projections in different domains worldwide. This information is crucial for the development of adaptation strategies and policy‐making. A relevant open issue in this context is assessing the potential multidomain conflicts that may result in overlapping regions and developing appropriate ensemble methods trying to make the most of all available information. This work addresses this timely topic by focusing on precipitation over the Mediterranean region, a first illustrative case study that is encompassed by both the Euro‐ and Africa‐CORDEX domains. We focus on several mean, extreme, and temporal indices and use variance decomposition to assess the separate contribution of the domain and models to the climate change signal, concluding that the contribution of the domain alone is nearly negligible (below 5% in all cases). Nevertheless, for some cases, the combined model/domain effect triggers up to 40% of the total variance. Plain Language Summary The Coordinated Regional Climate Downscaling Experiment (CORDEX) provides spatially detailed climate change projections for different regions across the world. These projections are obtained through numerical models that solve the governing equations of the atmosphere over spatial domains, which typically cover continental areas and encompass several regions. The regional climate change information generated by these models presents various sources of uncertainties. This work addresses the uncertainty related to the choice of domain, which has not been properly assessed to date, despite it can potentially affect vast regions of the world for which model simulations coming from different CORDEX domains are available. We focus on precipitation over the Mediterranean region, which is encompassed by both the EURO‐ and AFR‐CORDEX domains, and quantify the separate contribution of the model and domain alone to the total uncertainty for the climate change signals. Our results indicate that this uncertainty comes mostly determined by the choice of model, with little variability coming from the domain. This would allow for combining different model simulations corresponding to overlapping domains since conflicting signals are very unlikely to occur. These findings may ease the decision‐making process in regions for which multimodel and multidomain heterogeneous climate change information is a