Projected climate change in Fennoscandia – and its relation to ensemble spread and global trends

Abstract. The need for information about climate change is great. This information is usually based on climate model data, which often have systematic biases. Furthermore, climate information is based on ensembles of climate models, which raises the question about how such ensembles are affected by the choice of models and emission scenarios. Here, we aim to describe climate change in Sweden and neighbouring countries and discuss how local changes relate to global warming. We present climate change projections based on bias adjusted Euro-CORDEX (Coordinated Regional Downscaling Experiment) regional climate model data centred over Sweden. Global warming results in higher temperature, more warm days, and fewer cold days in Sweden. The regional climate models replicate the signal of the driving global models. Yet, the model spread is smaller than in the full CMIP5 ensemble, which means that the RCMs do not fully represent the potential model spread. The choice of emission scenario has minimal effect on the calculation of mean climate change at a global warming level of 2 degrees. This implies that it would be safe to mix emission scenarios in calculations of global warming levels, at least up to +2 °C, and as long as mean values are concerned. Moreover, the differences in local and global warming rates seem to decrease with time, suggesting that climate change in Sweden may currently be at its fastest.