21st century changes in the European climate as simulated in an ensemble of 16 transient regional climate model simulations with the Rossby Centre model Erik Kjellström and Grigory Nikulin, SMHI Model and data Area The ENSEMBLES RT3 common domain (50km res.) Variables Seasonal and annual average MSLP, T 2m , Precipitation and W 10m Time periods 1961-1990, 2011-2040, 2041-2070, 2071-2100 Observations The gridded ENSEMBLES data set (E-OBS, Haylock et al. (2008)) Model RCA3 (Kjellström et al. 2005; Samuelsson et al., 2009) driven by ERA40 and different AOGCMs. Emission scenarios A2 (3), A1B (10), B2 (2), B1 (1) Forcing AOGCMs BCM (1), CNRM (1), ECHAM4 (2), ECHAM5 (5), HadCM3 (3 perturbed physics), IPSL (1), CCSM3 (3) Natural variability ECHAM5 A1B run three times differing only in initial conditions Background Uncertainties in future climate change are related to: (1) how will the external forcing of the climate system change in the future? (2) how will these changes in external forcing factors influence climate? (3) to what degree is the future climate change signal masked/amplified by natural variability of the climate system? Existing ensembles like those derived in the PRUDENCE (Christensen and Christensen, 2007) and ENSEMBLES projects do not take into account all three sources of uncertainty. For example; The PRUDENCE project only used 2(3) GCMs, ENSEMBLES is restricted to the A1B scenario and does not address natural variability. At the Rossby Centre an ensemble existing of 16 RCM simulations at the same resolution and domain differing in choice of global climate model, emission scenario and intitial conditions has been developed. References • Christensen, J. H. and Christensen, O. B., 2007. A summary of the PRUDENCE model projections of changes in European climate by the end of the century. Climatic Change 81, 7-30. doi:10.1007/s10584-006-9210-7. • Haylock, M. R., Hofstra, N., Klein Tank, A. M. G., Klok, E. J., Jones, P. D. and New, M. A. 2008. European daily high-resolution gridded data set of surface temperature and precipitation for 1950-2006. J. Geophys. Res., 113, D20119, doi:10.1029/2008JD010201 • Kjellström, E., Bärring, L., Gollvik, S., Hansson, U., Jones, C., Samuelsson, P., Rummukainen, M., Ullerstig, A., Willén U. and Wyser, K., 2005. A 140-year simulation of European climate with the new version of the Rossby Centre regional atmospheric climate model (RCA3). Reports Meteorology and Climatology, 108, SMHI, SE-60176 Norrköping, Sweden, 54 pp. • Kjellström, E., Hansson, U., Jones, C., Nikulin, G., Strandberg, G. and Ullerstig, A., 2009. 21st century changes in the European climate: uncertainties derived from an ensemble of regional climate model simulations. In preparation for Tellus. • Samuelsson, P., Gollvik, S., Hansson, U., Jones, C., Kjellström, E., Nikulin, G., Ullerstig, A., Willén, U. Wyser, K., 2009. The Rossby Centre Regional Climate Model RCA3: Model description and performance. Manuscript in preparation for Tellus. E-OBS (V1) 1012 1014 1016 0 6 12 18 24 30 o C 2m temperature Summer (JJA) 1961-1990, (SLP: 2 hPa) RCA (ERA40) RCA (6 GCMs) 2 2 RCA (ECHAM5) 2 RCA (HADCM3) -4 -4 -4 -4 -2 -2 -2 -2 RCA (BCM) -4 -4 -2 -2 2 4 4 4 6 RCA (CCSM3) 2 2 2 2 2 4 4 RCA (CNRM) 2 2 2 4 4 6 RCA (IPSL) 2 2 4 4 6 8 -6 -4 -2 0 2 4 6 o C RCA(ERA40, GCMs) - E-OBS E-OBS (V1) 1006 1010 1014 1018 1018 -20 -16 -12 -8 -4 0 4 8 12 16 o C 2m temperature Winter (DJF) 1961-1990, (SLP: 2 hPa) RCA (ERA40) 2 RCA (6 GCMs) -4 -2 -2 2 2 4 4 RCA (ECHAM5) -2 2 2 4 RCA (HADCM3) -4 -4 -4 -2 -2 -2 4 RCA (BCM) -8 -6 -4 -4 -2 4 6 RCA (CCSM3) -10 -8 -8 -8 -6 -6 -4 -4 -2 2 4 6 RCA (CNRM) 2 2 2 4 4 6 RCA (IPSL) -2 2 4 4 6 6 8 -6 -4 -2 0 2 4 6 o C RCA(ERA40, GCMs) - E-OBS RCA (6 GCMs) CTL 1002 1006 1010 1014 1018 1022 -20 -16 -12 -8 -4 0 4 8 12 16 o C 2m temperature Winter (DJF) SCN: 2071-2100 CTL: 1961-1990 (SLP: 1 hPa) RCA (6 GCMs) 1 2 2 RCA (ECHAM5) 1 RCA (HADCM3) RCA (BCM) -3 2 2 2 3 RCA (CCSM3) 2 2 RCA (CNRM) 1 2 2 3 3 RCA (IPSL) 1 2 2 0 1 2 3 4 5 6 7 8 o C SCN - CTL E-OBS (V1) 1014 1014 0 15 30 45 60 75 90 105 120 150 180 mm month Precipitation Summer (JJA) 1961-1990, (SLP: 2 hPa) RCA (ERA40) RCA (6 GCMs) 2 2 RCA (ECHAM5) 2 RCA (HADCM3) -4 -4 -4 -4 -2 -2 -2 -2 RCA (BCM) -4 -4 -2 -2 2 4 4 4 6 RCA (CCSM3) 2 2 2 2 2 4 4 RCA (CNRM) 2 2 2 4 4 6 RCA (IPSL) 2 2 4 4 6 8 -100 -60 -20 20 60 100 % [ RCA(ERA40, GCMs) - E-OBS ] E-OBS E-OBS (V1) 1006 1010 1014 1018 1018 0 15 30 45 60 75 90 105 120 150 200 300 mm month Precipitation Winter (DJF) 1961-1990, (SLP: 2 hPa) RCA (ERA40) 2 RCA (6 GCMs) -4 -2 2 2 4 4 RCA (ECHAM5) 2 2 4 RCA (HADCM3) -4 -4 -4 -2 4 RCA (BCM) -8 -6 -4 -4 -2 4 6 RCA (CCSM3) -10 -8 -8 -8 -6 -6 -4 -4 -2 2 4 6 RCA (CNRM) 2 2 2 4 4 6 RCA (IPSL) -2 2 2 4 4 6 6 8 -100 -60 -20 20 60 100 % [ RCA(ERA40, GCMs) - E-OBS ] E-OBS Biases compared to E-OBS in the control climate: • Biases generally smaller when downscaling ERA40 than when downscaling AOGCMs • Ensemble mean shows relatively small biases • Biases are strongly related to errors in the large-scale circulation in the AOGCMs • Precipitation generally larger in RCA3 than in E-OBS except in southern Europe in summer • Systematic differences between RCA3 and E-OBS in mountainous areas • Small differences between three ECHAM5 ensemble members (not shown) • Large differences between three HadCM3 perturbed physics ensemble members (not shown) • Local/regional differences up to 3-4°C (100%) or more are present in the simulations RCA (6 GCMs) CTL 1011 1014 1014 1014 1014 1017 0 5 10 15 20 25 30 35 o C 2m temperature Summer (JJA) SCN: 2071-2100 CTL: 1961-1990 (SLP: 1 hPa) RCA (6 GCMs) 1 RCA (ECHAM5) -2 -1 RCA (HADCM3) -1 -1 -1 -1 RCA (BCM) 1 1 2 2 RCA (CCSM3) -2 -2 -1 RCA (CNRM) 1 1 2 2 RCA (IPSL) -1 -1 -1 1 0 1 2 3 4 5 6 7 o C SCN - CTL RCA (6 GCMs) CTL 1002 1006 1010 1014 1018 1022 0 15 30 45 60 75 90 105 120 150 200 300 mm month Precipitation Winter (DJF) SCN: 2071-2100 CTL: 1961-1990 (SLP: 1 hPa) RCA (6 GCMs) 1 1 2 2 RCA (ECHAM5) 1 RCA (HADCM3) RCA (BCM) -3 2 2 2 3 RCA (CCSM3) 2 2 RCA (CNRM) 1 2 3 3 RCA (IPSL) 2 2 -60 -40 -20 0 20 40 60 % (SCN-CTL)/CTL RCA (6 GCMs) CTL 1014 1014 1014 1014 0 15 30 45 60 75 90 105 120 150 200 300 mm month Precipitation Summer (JJA) SCN: 2071-2100 CTL: 1961-1990 (SLP: 1 hPa) RCA (6 GCMs) 1 RCA (ECHAM5) -2 -1 RCA (HADCM3) -1 -1 -1 RCA (BCM) 1 1 2 2 RCA (CCSM3) -2 -2 -1 RCA (CNRM) 1 1 2 2 RCA (IPSL) -1 -1 -1 1 -60 -40 -20 0 20 40 60 % (SCN-CTL)/CTL • CC patterns in temperature are more coherent between simulations than bias patterns • Differences between CC patterns are partly related to differences in the large-scale circulation in the AOGCM • Temperature changes are statistically significant in all seasons • Largest increases in the northeast in winter and in the south in summer • Precipitation changes are statistically significant in southern Europe (decrease) and northern Europe (increase). The border line between these areas migrates with season • Changes in 10m-wind speed are less coherent between simulations than temperature and precipitation. Mostly decreasing wind speed in all seasons. Exceptions are the Baltic Sea and oceanic areas to the north of Europe in all seasons and local increases in the Medterranean area in summer • Large differences between different ensemble members (ECHAM5-A1B-r1/2/3) related to large-scale circulation contributes to uncertainty in earlier time periods (not shown) Climate change: Acknowledgements The work presented here is a result of the collective work of all Rossby Centre staff members. The RCA3 runs are partly funded through the EU ENSEMBLES project, partly by the Climate and Energy Systems (CES) project funded by the Nordic Energy research and Nordic Energy sector, and partly funded through the MISTRA-Swecia programme. The model simulations were performed on the climate computing resource "Tornado" operated by the National Supercomputer Centre at Linköping University. Tornado is funded with a grant from the Knut and Alice Wallenberg Foundation. We are grateful to the GCM centres that have provided us with boundary conditions for the simulations. We acknowledge the E-OBS dataset from the EU- FP6 project ENSEMBLES (http://www.ensembles-eu.org) and the data providers in the ECA&D project (http://eca.knmi.nl) Temperature (2m) Precipitation Precipitation Temperature (2m) JJA JJA DJF DJF JJA JJA DJF DJF RCA (6 GCMs) CTL 1014 1014 1014 1014 2 3 4 5 6 7 8 9 10 11 m/s Wind Summer (JJA) SCN: 2071-2100 CTL: 1961-1990 (SLP: 1 hPa) RCA (6 GCMs) 1 RCA (ECHAM5) -2 -1 RCA (HADCM3) -1 -1 -1 -1 -1 RCA (BCM) 1 1 2 2 RCA (CCSM3) -2 -2 -1 RCA (CNRM) 1 1 2 2 RCA (IPSL) -1 -1 -1 1 -1 -0.5 0 0.5 1 1.5 2 m/s SCN - CTL RCA (6 GCMs) CTL 1002 1006 1010 1014 1018 1022 2 3 4 5 6 7 8 9 10 11 m/s Wind Winter (DJF) SCN: 2071-2100 CTL: 1961-1990 (SLP: 1 hPa) RCA (6 GCMs) 1 1 2 2 RCA (ECHAM5) 1 RCA (HADCM3) RCA (BCM) -3 2 2 2 3 RCA (CCSM3) 2 2 RCA (CNRM) 1 2 3 3 RCA (IPSL) 1 2 -1 -0.5 0 0.5 1 1.5 2 m/s SCN - CTL Wind speed (10m) JJA DJF Summary A new regional climate change ensemble consisting of 16 transient simulations at 50 km horizontal resolution has been developed at the Rossby Centre. Dynamical downscaling with RCA3 of in total 7 AOGCMs (including all those used in ENSEMBLES RT2B) gives a good possibility of investigating uncertainty in European climate change due to choice of AOGCM. Results show that regional features in both biases in the recent past climate as well as in the climate change signal are to a large degree dependent on the large-scale circulation as inherited from the AOGCMs. The ensemble also includes four different emission scenarios and ensemble members with one AOGCM run three times with different initial conditions. The natural variability is an important contributor to the spread in the projections. Differences between emission scenarios become important in the latter part of the century.