EURO‐CORDEX regional climate model simulation of precipitation on Scottish islands (1971–2000): model performance and implications for decision‐making in topographically complex regions

Due to their scale and complex topography, islands such as the Hebrides and Shetland Islands are not completely resolved by global climate models, which may impact the quality of data that can be provided about future climate in such locations. In principle, dynamical downscaling may provide helpful additional detail about future local climate. However, there is also the potential for error and uncertainty to cascade through to the regional simulation. Here, we evaluate the simulative skill of the EURO‐CORDEX regional climate model ensemble on regional and local scales in the Hebrides and Shetland Islands, and consider the potential for such models to aid decision‐making in island settings, and other locations characterized by complex topography. Several precipitation indices (accumulated precipitation amount, mean daily precipitation amount, max 1‐day and 5‐day precipitation amounts, simple daily intensity, number of heavy and very heavy precipitation days) are used to assess model performance and identify bias relative to observations. Models are compared regionally, and at specific locations, namely Stornoway in the Hebrides and Lerwick in Shetland, for the period 1971–2000. Regional evaluation utilizes the UKCP09 gridded observational data set and local evaluation at Stornoway Airport and Lerwick utilizes observed mean precipitation and extreme indices from the European Climate Assessment & Dataset project. While no models perform skilfully across all the metrics studied, some models capture aspects of the precipitation climate at each location particularly well. Differences in model performance between the two case study sites highlight the value of evaluating models on multiple spatial scales. The implications of model uncertainty for decision‐making are also discussed.