• English 
  • Spanish 

Multi-domain intercomparison and potential conflicts from CORDEX simulations: An example for the Mediterranean region

The Coordinated Regional Climate Downscaling Experiment (CORDEX) has made available an enormous amount of regional climate simulations and projections in different domains across the world. Despite being crucial for the development of better adaptation strategies and policy making, it is difficult for a typical user of this kind of information to understand the differences that may be expected across the different data sources available for a given region (different domains, GCM/RCM ensembles, resolutions, etc.). This is the case for a number of regions of particular interest which are covered by more than one CORDEX domain. A relevant question in those cases is the consistency and potential conflicts which may arise for the projections provided by different domains. This work addresses this hot topic by focusing on the Mediterranean region, a first illustrative case-study which is encompassed by the EURO-, AFR- and MED-CORDEX domains.
To do so, we first assess the performance of a common multi-domain set of RCMs to simulate a number of precipitation-related indicators (accounting for spells and high-order moments) when driven with ERA-Interim data. Preliminary results from this “perfect” validation experiment indicate that the effect of the choice of domain is in general small. Second, to assess whether or not these small differences become relevant when the RCMs are driven by GCMs, we also inter-compare the climate change signals projected by the full ensemble of GCM/RCM combinations available for the EURO-, MED- and AFR-CORDEX domains, as well as for those obtained for the common set of simulations available for the three domains. This allows to better characterize the uncertainty that is present in CORDEX climate change simulations, covering not only the main factors which have been traditionally analyzed (choice of GCM, RCM, scenario) but also a new one which has not been taken into account before (the choice of domain). The results from this study will help different users’ communities to better frame their decisions for impact applications.
Moreover, we plan to broaden this study by applying the same methodology to other regions of the world where RCM-based simulations coming from different CORDEX domains are also available.