European Temperature Extremes: Mechanisms and Responses to Climate Change PhD Thesis

Transcription

1 European Temperature Extremes: Mechanisms and Responses to Climate Change PhD Thesis Julien Cattiaux Supervisors: Robert Vautard & Pascal Yiou Laboratoire des Sciences du Climat et de l Environnement UMR 8 CEA-CNRS-UVSQ, IPSL Gif-sur-Yvette, France December, J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

2 Extremes? Why? Summer 3: public & scientific awareness. highest impacts on societies & ecosystems. numerous recent examples: signature of climate change? Understanding the summer C (+3.σ) (e.g., Beniston, 4; Trigo et al., 5). soil atmosphere feedback (e.g., Fischer et al., 7; Seneviratne et al., 6; Vautard et al., 7). future amplification of summer heat-waves (e.g., Fischer and Schär, 9) ISI Publi's topic: 'Temperature extreme' Years ECA&D stations (Klein-Tank et al., ). Anomalies wrt J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

3 Extremes? Why? My PhD... Recent & future extremes in European seasonal temperatures, for all seasons. Role of the internal variability? Feedbacks? ISI Publi's topic: 'Temperature extreme' Years ECA&D stations (Klein-Tank et al., ). Anomalies wrt J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

4 A burning issue... MAM European seasonal temperature anomalies Densities Tobs anomalies in degc JJA SON DJF Years J. Cattiaux (LSCE) European temperature extremes under CC December, 3 / 4

5 A burning issue... MAM European seasonal temperature anomalies Densities Tobs anomalies in degc JJA SON DJF Years J. Cattiaux (LSCE) European temperature extremes under CC December, 3 / 4

6 A burning issue... Tobs anomalies in degc MAM JJA SON DJF European seasonal temperature anomalies Densities Years J. Cattiaux (LSCE) European temperature extremes under CC December, 3 / 4

7 A concern for future years IPCC, 7. Fate of extremely warm/cold seasons in the st century? J. Cattiaux (LSCE) European temperature extremes under CC December, 4 / 4

8 A concern for future years Future changes in European summer temperatures IPCC, ??? 3 Fate of extremely warm/cold seasons in the st century? J. Cattiaux (LSCE) European temperature extremes under CC December, 4 / 4

9 Main questions Question What are the main drivers of recent extremes & trends observed in European seasonal temperatures? physical & dynamical mechanisms Question Does climate change affect the natural variability in the North-Atlantic European (NAE) area? How does it contribute to the European warming? responses to climate change J. Cattiaux (LSCE) European temperature extremes under CC December, 5 / 4

10 Outline. Drivers of recent extremes & trends Introduction: NAE dynamics as the main driver of European climate Case study of the exceptionally warm autumn of 6 J. Cattiaux et al. (9), Origins of the extremely warm European fall of 6, Geophysical Research Letters, 36 (6), pp. L673. DOI:.9/9GL37339 Generalization to other seasons J. Cattiaux et al. (b), North-Atlantic SST amplified recent wintertime European land temperature extremes and trends, Climate Dynamics, published online. DOI:.7/s Role of the natural variability in future European warming Analysis of future climate projections from IPCC-AR4 (7) J. Cattiaux et al. (a), Dynamics of future seasonal temperature trends and extremes in Europe: a multi-model analysis from IPCC-AR4, to be submitted to Climate Dynamics, in prep.. Application: case study of the cold winter 9/ J. Cattiaux et al. (c), Winter in Europe: A cold extreme in a warming climate, Geophysical Research Letters, 37, pp. L74. DOI:.9/GL4463 J. Cattiaux (LSCE) European temperature extremes under CC December, 6 / 4

11 Outline. Drivers of recent extremes & trends Introduction: NAE dynamics as the main driver of European climate Case study of the exceptionally warm autumn of 6 J. Cattiaux et al. (9), Origins of the extremely warm European fall of 6, Geophysical Research Letters, 36 (6), pp. L673. DOI:.9/9GL37339 Generalization to other seasons J. Cattiaux et al. (b), North-Atlantic SST amplified recent wintertime European land temperature extremes and trends, Climate Dynamics, published online. DOI:.7/s Role of the natural variability in future European warming Analysis of future climate projections from IPCC-AR4 (7) J. Cattiaux et al. (a), Dynamics of future seasonal temperature trends and extremes in Europe: a multi-model analysis from IPCC-AR4, to be submitted to Climate Dynamics, in prep.. Application: case study of the cold winter 9/ J. Cattiaux et al. (c), Winter in Europe: A cold extreme in a warming climate, Geophysical Research Letters, 37, pp. L74. DOI:.9/GL4463 J. Cattiaux (LSCE) European temperature extremes under CC December, 6 / 4

12 European temperatures & NAE dynamics Introduction Europe is under the influence of the disturbances in the mid-latitudes jet stream (strongest in winter). First mode of variability: the North Atlantic Oscillation (NAO). NAO influence: NAO+ (NAO ) warm (cold) weather. Lamont-Doherty Earth Observatory. J. Cattiaux (LSCE) European temperature extremes under CC December, 7 / 4

13 European temperatures & NAE dynamics Recent winter warming: increase in NAO+? NAO influence: NAO+ (NAO ) warm (cold) weather. 98, early 99s: high frequency of NAO+ winters European warming is driven by changes in the NAE dynamics (e.g., Corti et al., 999; Gillett et al., 3; Hsu and Zwiers, ; Palmer, 999). since 995: high frequency of NAO winters, while temperatures still increase (e.g., 6/7) inconsistency. J. Cattiaux (LSCE) European temperature extremes under CC December, 8 / 4

14 European temperatures & NAE dynamics Recent winter warming: increase in NAO+? NAO Index 3 3 NAO influence: NAO+ (NAO ) warm (cold) weather. 98, early 99s: high frequency of NAO+ winters European warming is driven by changes in the NAE dynamics (e.g., Corti et al., 999; Gillett et al., 3; Hsu and Zwiers, ; Palmer, 999). since 995: high frequency of NAO winters, while temperatures still increase (e.g., 6/7) inconsistency. NAO Index Years Tanom (K) 3 r =.75 Winter temperature Years NAO index: Jones et al. (998). Temperatures: ECA&D stations (Klein-Tank et al., ). J. Cattiaux (LSCE) European temperature extremes under CC December, 8 / 4

15 European temperatures & NAE dynamics Recent winter warming: increase in NAO+? NAO Index 3 3 NAO influence: NAO+ (NAO ) warm (cold) weather. 98, early 99s: high frequency of NAO+ winters European warming is driven by changes in the NAE dynamics (e.g., Corti et al., 999; Gillett et al., 3; Hsu and Zwiers, ; Palmer, 999). since 995: high frequency of NAO winters, while temperatures still increase (e.g., 6/7) inconsistency. NAO Index Years Tanom (K) 3 r =.7 Winter temperature Years NAO index: Jones et al. (998). Temperatures: ECA&D stations (Klein-Tank et al., ). J. Cattiaux (LSCE) European temperature extremes under CC December, 8 / 4

16 Inconsistency NAE dynamics / European temperatures The example of autumn winter 6/7 Flow-analogues Estimating the temperature of a given day by looking at temperatures associated with similar atmospheric circulations in the past (Lorenz, 969). Observed & analog temperatures of SONDJF 6/7 Yiou et al. (7). Temperatures: ECA&D stations (Tmin) (Klein-Tank et al., ). Z5: NCEP/NCAR reanalysis (Kistler et al., ). J. Cattiaux (LSCE) European temperature extremes under CC December, 9 / 4

17 Contribution of NAE dynamics to the mild autumn of 6 Tm V5 Tm anomaly SON.6 degc Years Cattiaux et al. (9). Tm & V5: NCEP/NCAR reanalysis (Kistler et al., ). J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

18 Contribution of NAE dynamics to the mild autumn of 6 Tm V5 Tm anomaly SON V5 anomaly SON degc m/s Years Years Cattiaux et al. (9). Tm & V5: NCEP/NCAR reanalysis (Kistler et al., ). J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

19 Contribution of NAE dynamics to the mild autumn of 6 Tm V5 degc Tm anomaly SON Years.6 Linear regression model.3 Tm and V5 highly correlated (r =.7). V5: no trend in recent years, while European Tm increase. Cattiaux et al. (9). Tm & V5: NCEP/NCAR reanalysis (Kistler et al., ). J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

20 Contribution of NAE dynamics to the mild autumn of 6 Tm V5 degc Tm anomaly SON Years.6 Linear regression model.3 Tm and V5 highly correlated (r =.7). V5: no trend in recent years, while European Tm increase. Cattiaux et al. (9). Tm & V5: NCEP/NCAR reanalysis (Kistler et al., ). J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

21 Contribution of NAE SST to the mild autumn of 6 Tm SST Tm anomaly SON SST anomaly SON.6 degc.3 degc Years Years Cattiaux et al. (9). Tm, V5 & SST: NCEP/NCAR reanalysis (Kistler et al., ). J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

22 Contribution of NAE SST to the mild autumn of 6 Tm SST degc Tm anomaly SON Years.6 Linear regression model.3 Adding SST improves the fit (r =.8). The increasing trend in Tm is reproduced by the SST warming. Cattiaux et al. (9). Tm, V5 & SST: NCEP/NCAR reanalysis (Kistler et al., ). J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

23 Contribution of NAE SST to the mild autumn of 6 Tm SST degc Tm anomaly SON Years.6 Linear regression model: limits.3 No evidence for any causality. Dynamics & SST are not independent. Cattiaux et al. (9). Tm, V5 & SST: NCEP/NCAR reanalysis (Kistler et al., ). J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

24 Contribution of NAE SST to the mild autumn of 6 MM5 sensitivity experiment MM5: regional climate model developed at PSU/NCAR. Experimental set-up: wind-nudged simulations of SON 6, only differing by SST forcing such as SST observed anomaly. Resolution.5.5, temperature & humidity not nudged. SST forcing Tm response Tm response =.8 C, consistent with statistical estimate. Cattiaux et al. (9). SST forcings: NCEP/NCAR reanalysis (Kistler et al., ). J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

25 Contribution of NAE SST to the mild autumn of 6 MM5 sensitivity experiment MM5: regional climate model developed at PSU/NCAR. Experimental set-up: wind-nudged simulations of SON 6, only differing by SST forcing such as SST observed anomaly. Resolution.5.5, temperature & humidity not nudged. SST forcing Tm response Tm response =.8 C, consistent with statistical estimate. Cattiaux et al. (9). SST forcings: NCEP/NCAR reanalysis (Kistler et al., ). J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

26 Contribution of NAE SST to the mild autumn of 6 Physical mechanisms E = LW + SW + LH + SH SST advection Limits sensible heat. water vapor, enhancing local greenhouse effect. Difficulty to interpret sea air fluxes in SST-forced experiments (Barsugli and Battisti, 998). See also: M.E. Shongwe et al. (9), Energy budget of the extreme Autumn 6 in Europe, Climate Dynamics, pp.. DOI:.7/s J. Cattiaux (LSCE) European temperature extremes under CC December, 3 / 4

27 Contribution of NAE SST to the mild autumn of 6 Physical mechanisms E = LW + SW + LH + SH SST advection Limits sensible heat. water vapor, enhancing local greenhouse effect. Difficulty to interpret sea air fluxes in SST-forced experiments (Barsugli and Battisti, 998). See also: M.E. Shongwe et al. (9), Energy budget of the extreme Autumn 6 in Europe, Climate Dynamics, pp.. DOI:.7/s J. Cattiaux (LSCE) European temperature extremes under CC December, 3 / 4

28 Contribution of NAE SST to the mild autumn of 6 Physical mechanisms E = LW + SW + LH + SH SST advection Limits sensible heat. water vapor, enhancing local greenhouse effect. Difficulty to interpret sea air fluxes in SST-forced experiments (Barsugli and Battisti, 998). See also: M.E. Shongwe et al. (9), Energy budget of the extreme Autumn 6 in Europe, Climate Dynamics, pp.. DOI:.7/s J. Cattiaux (LSCE) European temperature extremes under CC December, 3 / 4

29 Mild autumn 6 in Europe: summary An extremely warm event comparable in seasonal amplitude to the summer 3. Record-breaking southly flow: 5% of the temperature anomaly. Record-breaking warm NAE SST: 3% more, through the advection of both water vapor and sensible heat by the westerlies. % unexplained: non-linearities? Feedbacks (soil moisture, clouds, aerosols)? J. Cattiaux (LSCE) European temperature extremes under CC December, 4 / 4

30 Back to outline.... Drivers of recent extremes & trends Introduction: NAE dynamics as the main driver of European climate Case study of the exceptionally warm autumn of 6 J. Cattiaux et al. (9), Origins of the extremely warm European fall of 6, Geophysical Research Letters, 36 (6), pp. L673. DOI:.9/9GL37339 Generalization to other seasons J. Cattiaux et al. (b), North-Atlantic SST amplified recent wintertime European land temperature extremes and trends, Climate Dynamics, published online. DOI:.7/s Role of the natural variability in future European warming Analysis of future climate projections from IPCC-AR4 (7) J. Cattiaux et al. (a), Dynamics of future seasonal temperature trends and extremes in Europe: a multi-model analysis from IPCC-AR4, to be submitted to Climate Dynamics, in prep.. Application: case study of the cold winter 9/ J. Cattiaux et al. (c), Winter in Europe: A cold extreme in a warming climate, Geophysical Research Letters, 37, pp. L74. DOI:.9/GL4463 J. Cattiaux (LSCE) European temperature extremes under CC December, 5 / 4

31 Generalization: dynamics contribution (V5 SDI) Cattiaux et al. (b): SDI based on Z5 Tm correlation patterns. Tobs: E-OBS (ECA&D), Z5: NCEP/NCAR reanalysis. Flow-analogues: Vautard and Yiou (9), NAO index: Jones et al. (998). J. Cattiaux (LSCE) European temperature extremes under CC December, 6 / 4

32 Generalization: dynamics contribution (V5 SDI) Cattiaux et al. (b): SDI based on Z5 Tm correlation patterns. Tobs: E-OBS (ECA&D), Z5: NCEP/NCAR reanalysis. Flow-analogues: Vautard and Yiou (9), NAO index: Jones et al. (998). J. Cattiaux (LSCE) European temperature extremes under CC December, 6 / 4

33 Generalization: dynamics contribution (V5 SDI) r =.8 r =.78 r =.7 r =.88 Cattiaux et al. (b): SDI based on Z5 Tm correlation patterns. Tobs: E-OBS (ECA&D), Z5: NCEP/NCAR reanalysis. Flow-analogues: Vautard and Yiou (9), NAO index: Jones et al. (998). J. Cattiaux (LSCE) European temperature extremes under CC December, 6 / 4

34 Generalization: dynamics contribution (V5 SDI) r = r = r = r = Cattiaux et al. (b): SDI based on Z5 Tm correlation patterns. Tobs: E-OBS (ECA&D), Z5: NCEP/NCAR reanalysis. Flow-analogues: Vautard and Yiou (9), NAO index: Jones et al. (998). J. Cattiaux (LSCE) European temperature extremes under CC December, 6 / 4

35 Generalization: dynamics contribution (V5 SDI) r = r = r = r = Cattiaux et al. (b): SDI based on Z5 Tm correlation patterns. Tobs: E-OBS (ECA&D), Z5: NCEP/NCAR reanalysis. Flow-analogues: Vautard and Yiou (9), NAO index: Jones et al. (998). J. Cattiaux (LSCE) European temperature extremes under CC December, 6 / 4

36 Generalization: SST contribution r =.8 r =.78 r =.7 r =.88 Cattiaux et al. (b). Tobs: E-OBS (ECA&D), Z5: NCEP/NCAR reanalysis. SST: NCEP/NCAR reanalysis. J. Cattiaux (LSCE) European temperature extremes under CC December, 7 / 4

37 Generalization: SST contribution r =.8.83 r =.78.9 r =.7.83 r = Cattiaux et al. (b). Tobs: E-OBS (ECA&D), Z5: NCEP/NCAR reanalysis. SST: NCEP/NCAR reanalysis. J. Cattiaux (LSCE) European temperature extremes under CC December, 7 / 4

38 Generalization: SST contribution MM5 sensitivity experiments over 3 7 SST forcings Tm responses Seasonality highest (lowest) SST trend in autumn (spring). westerlies advection more (less) efficient in autumn winter (spring summer). spring summer: other feedbacks (e.g., soil moisture, aerosols, clouds 3 ). Cattiaux et al. (b). Seneviratne et al. (6), Ruckstuhl et al. (8), 3 Vautard et al. (9). J. Cattiaux (LSCE) European temperature extremes under CC December, 8 / 4

39 Generalization: SST contribution MM5 sensitivity experiments over 3 7 SST forcings Tm responses Seasonality highest (lowest) SST trend in autumn (spring). westerlies advection more (less) efficient in autumn winter (spring summer). spring summer: other feedbacks (e.g., soil moisture, aerosols, clouds 3 ). Cattiaux et al. (b). Seneviratne et al. (6), Ruckstuhl et al. (8), 3 Vautard et al. (9). J. Cattiaux (LSCE) European temperature extremes under CC December, 8 / 4

40 Generalization: SST contribution MM5 sensitivity experiments over 3 7 SST forcings Tm responses Seasonality highest (lowest) SST trend in autumn (spring). westerlies advection more (less) efficient in autumn winter (spring summer). spring summer: other feedbacks (e.g., soil moisture, aerosols, clouds 3 ). Cattiaux et al. (b). Seneviratne et al. (6), Ruckstuhl et al. (8), 3 Vautard et al. (9). J. Cattiaux (LSCE) European temperature extremes under CC December, 8 / 4

41 Recent European temperature extremes and trends: summary All seasonal warm records have been broken since 3, which acts in the recent long-term increase in European temperatures. The NAE atmospheric dynamics is the main driver of the temperatures interannual variability, but does not explain upward trends. The simultaneous warming of the NAE SST contributes to the European warming, through the advection of both water vapor and sensible heat by the westerlies. The SST contribution is higher (lower) in autumn winter (spring summer). Other feedbacks amplify European temperatures in spring summer (e.g., deficit in soil moisture, clouds, aerosols). Main limit: how to identify causality, since both dynamics and SST are coupled. J. Cattiaux (LSCE) European temperature extremes under CC December, 9 / 4

42 Back to outline II.... Drivers of recent extremes & trends Introduction: NAE dynamics as the main driver of European climate Case study of the exceptionally warm autumn of 6 J. Cattiaux et al. (9), Origins of the extremely warm European fall of 6, Geophysical Research Letters, 36 (6), pp. L673. DOI:.9/9GL37339 Generalization to other seasons J. Cattiaux et al. (b), North-Atlantic SST amplified recent wintertime European land temperature extremes and trends, Climate Dynamics, published online. DOI:.7/s Role of the natural variability in future European warming Analysis of future climate projections from IPCC-AR4 (7) J. Cattiaux et al. (a), Dynamics of future seasonal temperature trends and extremes in Europe: a multi-model analysis from IPCC-AR4, to be submitted to Climate Dynamics, in prep.. Application: case study of the cold winter 9/ J. Cattiaux et al. (c), Winter in Europe: A cold extreme in a warming climate, Geophysical Research Letters, 37, pp. L74. DOI:.9/GL4463 J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

43 In the future? Questions: Is the dynamics/temperature inconsistency expected to continue/amplify? How to estimate the contribution of changes in internal variability to the temperature increase? What are the atmospheric circulations associated with projected extremes? Methodology: Multi-model analysis from 3 CMIP3 GCMs: Tm & SLP. Periods 96 (c3m) & 8, in SRES A. Technique: flow-analogues & weather regimes (not shown today). Selection of best models: skillful model-ensembles per season. Cattiaux et al. (a), submitted. J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

44 In the future? Questions: Is the dynamics/temperature inconsistency expected to continue/amplify? How to estimate the contribution of changes in internal variability to the temperature increase? What are the atmospheric circulations associated with projected extremes? Methodology: Multi-model analysis from 3 CMIP3 GCMs: Tm & SLP. Periods 96 (c3m) & 8, in SRES A. Technique: flow-analogues & weather regimes (not shown today). Selection of best models: skillful model-ensembles per season. Cattiaux et al. (a), submitted. J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

45 Models selection per season Criterion: ability to reproduce observed SLP Tm correlation patterns. ERA-4 Cattiaux et al. (a), submitted. J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

46 Models selection per season Criterion: ability to reproduce observed SLP Tm correlation patterns. ERA-4 IPSL-CM4 Cattiaux et al. (a), submitted. J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

47 Models selection per season Criterion: ability to reproduce observed SLP Tm correlation patterns. ERA-4 Best models MAM JJA CCC47 MIR3 GFDL GFDL 3 GISSer ECH5 4 MRI3 CCC47 5 ECHOg ECH4 SON DJF ECHOg CCC47 CNRM3 GFDL 3 GFDL CNRM3 4 CSI3 MRI3 5 IPSL4 ECH4 Cattiaux et al. (a), submitted. J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

48 Models selection per season Criterion: ability to reproduce observed SLP Tm correlation patterns. ERA-4 Model-ensembles Cattiaux et al. (a), submitted. J. Cattiaux (LSCE) European temperature extremes under CC December, / 4

49 Projected changes in European temperatures From seasonal model-ensembles c3m Density Density. (a) MAM Tm in degc. (c) SON Tm in degc Density Density. (b) JJA Tm in degc. (d) DJF Tm in degc Increase in mean. Increase (decrease) in variability in summer (winter). Cattiaux et al. (a), submitted. J. Cattiaux (LSCE) European temperature extremes under CC December, 3 / 4

50 Projected changes in European temperatures From seasonal model-ensembles c3m Density Density. (a) MAM Tm in degc. (c) SON Tm in degc Density Density. (b) JJA Tm in degc. (d) DJF Tm in degc Increase in mean. Increase (decrease) in variability in summer (winter). Cattiaux et al. (a), submitted. J. Cattiaux (LSCE) European temperature extremes under CC December, 3 / 4

51 Projected changes in European temperatures From seasonal model-ensembles c3m Density Density. (a) MAM Tm in degc. (c) SON Tm in degc Density Density. (b) JJA Tm in degc. (d) DJF Tm in degc Increase in mean. Increase (decrease) in variability in summer (winter). Cattiaux et al. (a), submitted. J. Cattiaux (LSCE) European temperature extremes under CC December, 3 / 4

52 Projected changes in NAE dynamics From seasonal model-ensembles Difference 8 & 96 SLP (hpa): Mean Variance Strengthening and northward shift of the jet stream increase in NAO+ conditions (e.g., Terray et al., 4). Generalized decrease in interannual variability. Cattiaux et al. (a), submitted. J. Cattiaux (LSCE) European temperature extremes under CC December, 4 / 4

53 Projected changes in NAE dynamics From seasonal model-ensembles Difference 8 & 96 SLP (hpa): Mean Variance Strengthening and northward shift of the jet stream increase in NAO+ conditions (e.g., Terray et al., 4). Generalized decrease in interannual variability. Cattiaux et al. (a), submitted. J. Cattiaux (LSCE) European temperature extremes under CC December, 4 / 4

54 NAE dynamics & future temperature increase Flow-analogues: method SLP Dec, 6 (CNRM-CM3) Flow-analogues in 96 Reference temperature.?? N analog temperatures. Limit? Is a 4-year sampling period sufficient? Cattiaux et al. (a), submitted. J. Cattiaux (LSCE) European temperature extremes under CC December, 5 / 4

55 NAE dynamics & future temperature increase Flow-analogues: method SLP Dec, 6 (CNRM-CM3) Flow-analogues in 96 r =.75 r =.75 r =.68 r =.63 Reference temperature.?? N analog temperatures. Limit? Is a 4-year sampling period sufficient? Cattiaux et al. (a), submitted. J. Cattiaux (LSCE) European temperature extremes under CC December, 5 / 4

56 NAE dynamics & future temperature increase Flow-analogues: method SLP Dec, 6 (CNRM-CM3) Flow-analogues in 96 r =.75 r =.75 r =.68 r =.63 Reference temperature.?? N analog temperatures. Limit? Is a 4-year sampling period sufficient? Cattiaux et al. (a), submitted. J. Cattiaux (LSCE) European temperature extremes under CC December, 5 / 4

57 NAE dynamics & future temperature increase Flow-analogues: method SLP Dec, 6 (CNRM-CM3) Flow-analogues in 96 r =.75 r =.75 r =.68 r =.63 Reference temperature.?? N analog temperatures. Limit? Is a 4-year sampling period sufficient? Cattiaux et al. (a), submitted. J. Cattiaux (LSCE) European temperature extremes under CC December, 5 / 4

58 NAE dynamics & future temperature increase Flow-analogues: correlations Correlations of the 5 th flow-analogue Correlations (a) MAM Correlations (b) JJA Correlations Years Years (c) SON.8.75 Correlations Years Years Cattiaux et al. (a), submitted (d) DJF J. Cattiaux (LSCE) European temperature extremes under CC December, 6 / 4

59 NAE dynamics & future temperature increase Flow-analogues: correlations Correlations of the 5 th flow-analogue Correlations (a) MAM Correlations (b) JJA Correlations Years Years (c) SON.8.75 Correlations Years Years Cattiaux et al. (a), submitted (d) DJF J. Cattiaux (LSCE) European temperature extremes under CC December, 6 / 4

60 NAE dynamics & future temperature increase Flow-analogues: temperatures Normalized anomalies Normalized anomalies (a) MAM Ensemble reference Reference temperatures Years (c) SON Ensemble reference Normalized anomalies Normalized anomalies Analog temperatures (b) JJA Ensemble reference Years (d) DJF Ensemble reference Years Cattiaux et al. (a), submitted Years J. Cattiaux (LSCE) European temperature extremes under CC December, 7 / 4

61 NAE dynamics & future temperature increase Flow-analogues: temperatures Normalized anomalies Normalized anomalies (a) MAM Ensemble r =.67 r' =.69 reference Reference temperatures r =.7 r' =.7 analog r =.6 r' = Years (c) SON Ensemble r =.6 r' =.65 reference r =.47 r' =.54 analog r =.38 r' =.39 Normalized anomalies Normalized anomalies Analog temperatures (b) JJA Ensemble r =.43 r' =.48 reference r =.43 r' =.47 analog r =.5 r' = Years (d) DJF Ensemble r =.76 r' =.77 reference r =.66 r' =.66 analog r =.69 r' = Years Cattiaux et al. (a), submitted Years J. Cattiaux (LSCE) European temperature extremes under CC December, 7 / 4

62 Focus on extremely warm/cold seasons Exceeding ±σ of detrended temperature distributions. Reference temperatures Analog temperatures Norm. anoms (a) MAM reference Norm. anoms (b) JJA reference Norm. anoms (c) SON Norm. anoms 4 4 (d) DJF reference reference Cattiaux et al. (a), submitted. J. Cattiaux (LSCE) European temperature extremes under CC December, 8 / 4

63 Focus on extremely warm/cold seasons Exceeding ±σ of detrended temperature distributions. Reference temperatures Analog temperatures Norm. anoms (a) MAM reference analog Norm. anoms (b) JJA reference analog Norm. anoms (c) SON Norm. anoms 4 4 (d) DJF reference analog reference analog Cattiaux et al. (a), submitted. J. Cattiaux (LSCE) European temperature extremes under CC December, 8 / 4

64 NAE dynamics & future temperature increase: summary CMIP3 models generally perform well in reproducing NAE dynamics/temperature relationships. Changes in temperatures: increase in the mean. Widening (tightening) of summer (winter) distribution (see also Christensen et al., 7; Fischer and Schär, 9). Changes in dynamics: northward shift of the jet stream & decrease in the variability increase in NAO+ (see also Terray et al., 4). NAE dynamics is and remains the main driver of European temperatures variability, but only a minor part of long-term warming can be attributed to circulation changes (confirmed by weather-regime approach, not shown). Circulations associated with warm/cold seasons do not change. But feedbacks may change (e.g., amplify summer extremes (Fischer and Schär, 9)). Main limit: are flow-analogues able to capture any trend? J. Cattiaux (LSCE) European temperature extremes under CC December, 9 / 4

65 Back to outline III.... Drivers of recent extremes & trends Introduction: NAE dynamics as the main driver of European climate Case study of the exceptionally warm autumn of 6 J. Cattiaux et al. (9), Origins of the extremely warm European fall of 6, Geophysical Research Letters, 36 (6), pp. L673. DOI:.9/9GL37339 Generalization to other seasons J. Cattiaux et al. (b), North-Atlantic SST amplified recent wintertime European land temperature extremes and trends, Climate Dynamics, published online. DOI:.7/s Role of the natural variability in future European warming Analysis of future climate projections from IPCC-AR4 (7) J. Cattiaux et al. (a), Dynamics of future seasonal temperature trends and extremes in Europe: a multi-model analysis from IPCC-AR4, to be submitted to Climate Dynamics, in prep.. Application: case study of the cold winter 9/ J. Cattiaux et al. (c), Winter in Europe: A cold extreme in a warming climate, Geophysical Research Letters, 37, pp. L74. DOI:.9/GL4463 J. Cattiaux (LSCE) European temperature extremes under CC December, 3 / 4

66 Winter 9/: what happened? January 7, (NASA). Are European cold spells of winter 9/ incompatible with global warming? J. Cattiaux (LSCE) European temperature extremes under CC December, 3 / 4

67 Winter 9/: locally cold, globally warm GISS-temp ( 5 th warmest DJF on record since 88 (NOAA, see Cohen et al. ()). J. Cattiaux (LSCE) European temperature extremes under CC December, 3 / 4

68 Winter 9/: an extreme persistence of NAO (a) (m) Cattiaux et al. (c). Z5: NCEP/NCAR reanalysis. NAO index: Jones et al. (998). J. Cattiaux (LSCE) European temperature extremes under CC December, 33 / 4

69 Winter 9/: an extreme persistence of NAO (a) (m) Cattiaux et al. (c). Z5: NCEP/NCAR reanalysis. NAO index: Jones et al. (998). J. Cattiaux (LSCE) European temperature extremes under CC December, 33 / 4

70 Winter 9/ in Europe Cold, but not extreme (a) obs Normalized anomalies Cattiaux et al. (c). Tobs: E-OBS (ECA&D). 3 (d) Tavg (obs) Years 3 3 J. Cattiaux (LSCE) European temperature extremes under CC December, 34 / 4

71 Winter 9/ in Europe Warmer than flow-analogues (a) obs (b) ana (c) obs ana Normalized anomalies Cattiaux et al. (c). 3 (d) Tavg (obs) Tavg (ana) Years Tobs: E-OBS (ECA&D), Flow-analogues from NCEP/NCAR Z J. Cattiaux (LSCE) European temperature extremes under CC December, 35 / 4

72 Winter 9/ in Europe Warmer than flow-analogues (a) obs (b) ana (c) obs ana Normalized anomalies Cattiaux et al. (c). 3 (d) Detrended Tavg (obs) Tavg (ana) Years Tobs: E-OBS (ECA&D), Flow-analogues from NCEP/NCAR Z J. Cattiaux (LSCE) European temperature extremes under CC December, 35 / 4

73 Winter 9/ in Europe Less snowy than flow-analogues (a) obs (b) ana (c) obs ana Normalized anomalies 3 (d) SC (obs) SC (ana) Years Snow cover: E-OBS (ECA&D), Flow-analogues from NCEP/NCAR Z5. J. Cattiaux (LSCE) European temperature extremes under CC December, 36 / 4

74 Cold winter 9/ in Europe: summary Regionally cold and snowy, globally warm (Cohen et al., ). Caused by an exceptional persistence of negative NAO conditions. In Europe, winter was cold (especially Tmax), but not extreme and warmer than expected from flow-analogues estimates. Observed analog difference consistent with the 98 9 long-term observed trend. J. Cattiaux (LSCE) European temperature extremes under CC December, 37 / 4

75 Main conclusions Recent trends & extremes in European seasonal temperatures are not explained by the NAE atmospheric dynamics. Seasonality in the amplifying mechanisms: local feedbacks in spring summer, heat & water vapor advection from Atlantic in autumn winter. Projected changes in NAE circulations (enhanced NAO+) only have a minor contribution to projected temperature increase. Future warm/cold extremes are associated with similar circulations as recent warm/cold extremes. Recent wintertime cold spells in Europe are all but incompatible with global warming. J. Cattiaux (LSCE) European temperature extremes under CC December, 38 / 4

76 Main conclusions Recent trends & extremes in European seasonal temperatures are not explained by the NAE atmospheric dynamics. Seasonality in the amplifying mechanisms: local feedbacks in spring summer, heat & water vapor advection from Atlantic in autumn winter. Projected changes in NAE circulations (enhanced NAO+) only have a minor contribution to projected temperature increase. Future warm/cold extremes are associated with similar circulations as recent warm/cold extremes. Recent wintertime cold spells in Europe are all but incompatible with global warming. J. Cattiaux (LSCE) European temperature extremes under CC December, 38 / 4

77 Limits Difficulty to isolate individual signatures in coupled systems: Linear regression model? No evidence for causality. Sensitivity experiments? Physics must be broken somewhere. Flow-analogues method: could analog temperatures show up any trend, given that flow-analogues days are sampled in a quite stationary climate? Simplistic hypothesis in multi-model approach: Selection of best models: arbitrary. Equi-probability of models in combining future climate projections. Using multi-model range as an estimate for uncertainties. J. Cattiaux (LSCE) European temperature extremes under CC December, 39 / 4

78 Prospects Methodology, statistical & modeling tools could be applied to: CMIP5 models (IPCC-AR5) Post-doc at CNRM. other regions Arctic (with V. Masson-Delmotte), Asia (P. Yiou). other variables (snow, rainfall (Vautard and Yiou, 9)). other periods (last millenium, volcanic eruptions (T. Salameh)). model development IPSL-CM4/5 analysis (with LMD people). Need for going to higher time & spatial scales (downscaling): sub-seasonal scale: persistance/intensity of heat-waves & cold spells (PhD B. Quesada), weather-regime approach (with M. Vrac). more local scale: regional climate projections (projects DRIAS, CORDEX, SALUTAIR). Towards prediction of temperature extreme events? operational: looking for precursors (e.g., spring soil moisture, autumn arctic sea ice). Not shown today: discussion about different forcings of the NAO (case study of winter 9/). statistical: probabilistic methods for weighted combinations of multi-model projections, wrt. individual model skills. J. Cattiaux (LSCE) European temperature extremes under CC December, 4 / 4

79 Thanks for your attention. Questions? J. Cattiaux (LSCE) European temperature extremes under CC December, 4 / 4

80 Energy budget of the mild autumn of 6 Physical mechanisms E = LW + SW + LH + SH increased solar flux caused by the anticyclonic circulation. water vapor from warm SST, enhancing local greenhouse effect. sensible heat advected from warm SST. See also: Shongwe et al. (9) & Philipona et al. (5). J. Cattiaux (LSCE) European temperature extremes under CC December, 4 / 4

81 Warming of North Atlantic ocean Natural or anthropogenic? (a) SST North-Atlantic North-Atlantic SST trend natural multi-decadal oscillation (6 years): the AMO (Kerr, ). global warming, very likely due to human activities (IPCC-AR4). Great debate in recent years! Knight et al. (5) vs. Trenberth and Shea (6) or Ting et al. (9). (b) SST Atlantic (c) SST AMO Trenberth and Shea (6). J. Cattiaux (LSCE) European temperature extremes under CC December, 4 / 4

82 European temperatures & NAE dynamics Winter weather regimes Preferential states of the NAE daily atmospheric circulation. Clustering algorithms 4 winter regimes. Associated temperatures: composites. Sea-Level Pressure (SLP), ERA-4 NDJFM 96. m-temperature (Tm), ERA-4 NDJFM 96. J. Cattiaux (LSCE) European temperature extremes under CC December, 4 / 4

83 European temperatures & NAE dynamics Winter weather regimes Preferential states of the NAE daily atmospheric circulation. Clustering algorithms 4 winter regimes. Associated temperatures: composites. Sea-Level Pressure (SLP), ERA-4 NDJFM 96. Sea-Level Pressure (SLP), IPSL-CM4 NDJFM 96. J. Cattiaux (LSCE) European temperature extremes under CC December, 4 / 4

84 European temperatures & NAE dynamics Winter weather regimes Preferential states of the NAE daily atmospheric circulation. Clustering algorithms 4 winter regimes. Associated temperatures: composites. Summer Winter ECHOg CCC47 MRI3 GFDL 3 IPSL4 CSI35 4 CSI35 ECHOg 5 ECH5 IPSL4 Sea-Level Pressure (SLP), ERA-4 NDJFM 96. J. Cattiaux (LSCE) European temperature extremes under CC December, 4 / 4

85 European temperatures & NAE dynamics Winter weather regimes Preferential states of the NAE daily atmospheric circulation. Clustering algorithms 4 winter regimes. Associated temperatures: composites. Sea-Level Pressure (SLP), ERA-4 NDJFM 96. Sea-Level Pressure (SLP), Ensemble NDJFM 96. J. Cattiaux (LSCE) European temperature extremes under CC December, 4 / 4

86 Temperature increase: changes in WR occurrences? X = N i= ( f i x P i ) } {{ } Inter + }{{} x + Intra com N ( ) N fi P x i + ( f i x i ) i= } {{ } Intra spe i= } {{ } Residu J. Cattiaux (LSCE) European temperature extremes under CC December, 4 / 4

87 Temperature increase: changes in WR occurrences? X = X F X P = N i= ( fi F xi F fi P x P i ) X = = = N i = N i = N i = [( f F i [( f F i [( f F i f i P ) x i F f i P ) ( x i P f i P ) x i P + f i P x i F x P i + f i P ( x i F f i P x i P ] + x F i ) x P i + f i P ( x i F ) ( + f i F x P i )] f i P ) ( x i F x P i )] () X = X = N i= N i= ( f i x P i } {{ } Inter ( f i x P i ) } {{ } Inter ) N + i= + }{{} x + Intra com ) N i x i + ( f i x i ) ( f P } {{ } Intra i= i= } {{ } Intra spe } {{ } Residu N ( ) N fi P x i + ( f i x i ) i= } {{ } Residu J. Cattiaux (LSCE) European temperature extremes under CC December, 4 / 4

88 Temperature increase: changes in WR occurrences? X = N i= ( f i x P i ) } {{ } Inter + }{{} x + Intra com N ( ) N fi P x i + ( f i x i ) i= } {{ } Intra spe i= } {{ } Residu J. Cattiaux (LSCE) European temperature extremes under CC December, 4 / 4

89 Winter 9/: why such an NAO? (a) Normalized time series Forcings ENSO (Cassou, ). Tropical-Atlantic SST (Cassou et al., 4). Eurasian snow over in October (Cohen et al., ). Late-summer Arctic sea-ice extent (Francis et al., 9). AOi (DJF) SOI (DJF) SST Atl Trop (DJF) Eurasian Snow (Prev Oct) (ND) Arctic Sea Ice (Prev Sep) Years /!\ Not systematic! 3 3 (b) AO index AOi AOi' =.3.SOI.7.SST.3.Snow +.6.SeaIce (r =.4) Years J. Cattiaux (LSCE) European temperature extremes under CC December, 4 / 4

90 Winter 9/: a unbreakable record for future years? Density Density Normalized anomalies Normalized anomaly Probability (%) (a) Sigma level of winter Tm Years (b) Probability of below Tm Years Probability of below- Tm ) Pr ( Ty T =. si y c3m. si 46 y 65. si 8 y Hypothesis Gaussian distributions. equi-probable models. models range as uncertainties. J. Cattiaux (LSCE) European temperature extremes under CC December, 4 / 4