Observed and modelled influence of atmospheric circulation on central England temperature extremes

The regional atmospheric circulation is a major driver of climate for western Europe and so the ability of climate models to accurately simulate its characteristics forms an important part of the rigorous testing of their performance. In this paper we examine the skills of seven regional climate models (RCMs) to reproduce mean daily temperatures for the central England region. Their ability to reproduce observed characteristics of the circulation is then tested using three airflow indices. This is achieved by comparing: (1) the frequency distribution of each index, (2) the relationships between the daily airflow indices and temperature, including daily extremes and (3) the ability of models to reproduce the observed persistence of specific flow regimes and the temperature response to this persistence. It is demonstrated that RCM selection introduces uncertainty into temperature simulations and that there is no single model which outperforms the others. Although the models qualitatively reproduce the observed distributions of the airflow indices reasonably well, most models produce distributions that are statistically significantly different from the observations. In particular, all models overestimate the frequency of winter westerly flow, and biases in the relationships between circulation and temperature are also noted for extreme values of the airflow indices. The persistence of flow regimes is shown to have a significant effect on the observed temperature, though models have difficulty in reproducing the magnitude of the response to this persistence for some flow types. The results presented here not only form a useful model validation exercise but also further highlight the need for the use of multi-model ensembles in the generation of future climate scenarios. Furthermore, they suggest that the use of atmospheric circulation in statistical downscaling methods might be enhanced by the inclusion of persistence, particularly when producing scenarios of temperature extremes.