Atmospheric Chemistry and Physics Discussions www.atmos-chem-phys-discuss.net 1680-7367 1680-7375 5 4 2005 10.5194/acpd-5-7415-2005 http://www.atmos-chem-phys-discuss.net/5/7415/2005/ http://www.atmos-chem-phys-discuss.net/5/7415/2005/acpd-5-7415-2005.html http://www.atmos-chem-phys-discuss.net/5/7415/2005/acpd-5-7415-2005.pdf 7415 7455 2005-08-25 Large-scale atmospheric circulation biases and changes in global climate model simulations and their importance for regional climate scenarios: a case study for West-Central Europe A. P. van Ulden G. J. van Oldenborgh Koninklijk Nederlands Meteorologisch Institutuut (KNMI), PO Box 201, 3730 AE, De Bilt, Netherlands The credibility of regional climate change predictions for the 21st century depends on the ability of climate models to simulate global and regional circulations in a realistic manner. To investigate this issue, a large set of global coupled climate model experiments prepared for the Fourth Assessment Report of the Intergovernmental Panel on Climate Change has been studied. First we compared 20th century model simulations of longterm mean monthly sea level pressure patterns with ERA-40. We found a wide range in performance. Many models performed well on a global scale. For northern midlatitudes and Europe many models showed large errors, while other models simulated realistic pressure fields. <br><br> Next we focused on the monthly mean climate of West-Central Europe in the 20th century. In this region the climate depends strongly on the circulation. Westerlies bring temperate weather from the Atlantic Ocean, while easterlies bring cold spells in winter and hot weather in summer. In order to be credible for this region, a climate model has to show realistic circulation statistics in the current climate, and a response of temperature and precipitation variations to circulation variations that agrees with observations. We found that even models with a realistic mean pressure pattern over Europe still showed pronounced deviations from the observed circulation distributions. In particular, the frequency distributions of the strength of westerlies appears to be difficult to simulate well. This contributes substantially to biases in simulated temperatures and precipitation, which have to be accounted for when comparing model simulations with observations. <br><br> Finally we considered changes in climate simulations between the end of the 20th century and the end of the 21st century. Here we found that changes in simulated circulation statistics play an important role in climate scenarios. For temperature, the warm extremes in summer and cold extremes in winter are most sensitive to changes in circulation, because these extremes depend strongly on the simulated frequency of eastery flow. For precipitation, we found that circulation changes have a substantial influence, both on mean changes and on changes in the probability of wet extremes and of long dry spells. Because we do not know how reliable climate models are in their predictions of circulation changes, climate change predictions for Europe are as yet uncertain in many aspects.