Modelling ENSO in GCMs: overview, progress and challenges

Transcription

1 Modelling ENSO in GCMs: overview, progress and challenges Eric Guilyardi IPSL/LOCEAN, Paris, France & NCAS-Climate, Univ. Reading, UK With contributions from: Andrew Wittenberg, Alexey Fedorov, Mat Collins, Pascale Braconnot, Fei-Fei Jin, Seon-Tae Kim, James Lloyd, Benoît Vannière Workshop on Sub-seasonal to Seasonal Prediction Met Office Hadley Centre, Exeter, Dec Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec. 2010

2 El Niño in coupled models Ocean-atmosphere coupled anomaly in the tropical Pacific IPCC (2007) «All IPCC AR4 models show continued ENSO interannual variability in the future no matter what the change in average background conditions, but changes in ENSO interannual variability differ from model to model. Based on various assessments of the current multi-model archive, in which present-day El Niño events are now much better simulated than in the TAR, there is no consistent indication at this time of discernible future changes in ENSO amplitude or frequency.» IPCC AR4 report, WG1, Chap. 10, Executive Summary Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

3 Outline 1. ENSO in coupled GCMs 2. Atmosphere feedbacks during ENSO Dynamical (Bjerknes) feedback Heat flux feedback 3. New strategies to improve ENSO in models Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

4 El Niño in coupled GCMs mean state Zonal wind stress in central Pacific (mean and annual cycle) Trade winds too strong Taux Niño 4 (Pa) Both mean and annual cycle OBS. IPCC-class models Guilyardi et al. (BAMS 2009) Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

5 El Niño in coupled GCMs - amplitude ENSO Amplitude picntrl 2xco2 Standard deviation SSTA (C) ENSO amplitude in IPCC AR4 : much too large diversity! Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

6 El Niño in coupled GCMs - frequency Maximum power of Niño3 SSTA spectra IPCC AR4: improved towards low freq. but still large diversity IPCC TAR: to high frequency AchutaRao & Sperber (2006) Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

7 El Niño in coupled GCMs - structure and timing Westward zonal extension Too small meridional extension With impacts on periodicity (Capotondi et al. 2007) SST standard deviation Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

8 El Niño in coupled GCMs - structure and timing Westward zonal extension Too small meridional extension With impacts on periodicity (Capotondi et al. 2007) SST standard deviation Time sequence of El Niño/La Niña also has errors El Niño termination spatial characteristics e.g.: Model events terminate in West rather than in East Pacific Leloup et al. (2008) Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

9 El Niño in coupled GCMs - teleconnections Tropical teleconnections: Well established Via modulation of Walker circulation and/or equatorial waves ENSO influence over-dominant in models (e.g. IO, Saji et al. 2006, WAM, Joly et al. 2008, SAM, Annamalai et al. 2007, Cai et al. 2008) Extra-tropical teleconnections: Atmospheric bridge via Rossby wave train Still a debate on robustness (Sterl et a. 2007) Timing/amplitude/structure of events all key Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

10 El Niño in coupled GCMs - summary Clear improvement since ~15 years some models get Mean and Annual cycle and ENSO right! but: Amplitude: models diversity much larger than (recent) observed diversity Frequency: progress towards low frequency/wider spectra but still errors Seasonal phase lock: very few models have the spring relaxation and the winter variability maximum Structure and timing: westward extension and narrowing around equator, issues with time sequence (onset, termination) Modes: very few model exhibits the diversity of observed ENSO modes; most are locked into a S-mode (coherent with too strong trade winds) Teleconnections: ENSO influence over-dominant Guilyardi et al. (BAMS 2009) Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

11 ENSO Amplitude chnage ENSO Amplitude El Niño in a warming climate Model biases dominate over scenario picntrl 2xCO2 SRES A2 vs. pdcntrl Evolution of mean state has little impact Some studies detect change in structure and/or balance of processes (e.g. Philip and van Oldenborgh 2009) Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec IPCC (2007), Guilyardi et al. (2009a) 10

12 Reviews in BAMS and Nature Geoscience Coordinated within CLIVAR Pacific Panel Bull. Amer. Met. Soc. (March 2009) Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

13 Attributing ENSO errors: physical mechanisms Atmosphere response to SSTA Bjerknes wind stress feedback (van Oldenborgh al. 2005, Guilyardi 2006) Meridional response of wind stress (An & Wang 2000, Capotondi al. 2006, Merryfield 2006) Radiative and cloud feedbacks (Sun al. 2006, Bony al. 2006, Sun et al. 2009) van Oldenborgh al Other processes: NL dynamical heating ( x T + U in phase, An & Jin 2004) "Multiplicative noise" - MJO (Lengaigne et al. 2004, Perez et al. 2005, Philip & van Oldenborgh 2007) Ocean response to τ and HF anomalies Upwelling, mixing, ("thermocline feedback", "cold tongue dynamics") (Meehl al. 2001, Burgers & van Oldenborgh 2003) Zonal advection (Picaut al. 1997) Wave dynamics Energy Dissipation (Fedorov 2006) Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

14 Attributing ENSO errors: use of simplified frameworks BJ index Theory: linear stability of recharged oscillator s SST equation (Jin et al. 2006) Apply to CMIP3 models (Kim and Jin 2010) MA TD ZA EK TH ENSO BJ = MA + TD + ZA + EK + TH Analysis of contributing terms (ratio to observed ) (see also Belmadani et al. 2009, Roberts and Battisti 2010,...) Model diversity/errors comes from thermal damping ( ) and Thermocline and Ekman feedbacks Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

15 Atmosphere feedbacks during ENSO Multi-model and sensitivity studies show that AGCM has a dominant role (e.g. Schneider 2002, Guilyardi et al. 2004, Kim et al. 2008, Neale et al. 2008, Sun et al. 2008,...) Two types of feedbacks: Dynamical: Bjerknes feedback µ Heat flux feedback Guilyardi et al. (2004) East-west SST gradient Trade winds +ve SST increase in the east Modified heat fluxes (SHF, LHF) -ve Equatorial upwelling in the east Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

16 Evaluating the Bjerknes feedback µ Niño 4 TauX anomaly Niño 3 SST anomaly Seasonal evolution of µ Monthly variability = measure of seasonal phase lock Bjerknes amplification stronger in July-December Obs. values of µ vary from 10 to 15 (10-3 N.m -2 /C) Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

17 Evaluating the heat flux feedback α Niño 3 Heat Flux anomaly Niño 3 SST anomaly Defined as slope of heat flux QA = F(SSTA) α varies from -10 W.m -2 /C to -40 W.m -2 /C Damping stronger in January-May Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

18 Impact of atmosphere convection scheme on ENSO Observations (0.9 C) - HadiSST1.1 Hourdin et al. (2006) Braconnot et al. (2007) IPSL (KE) Kerry Emanuel (1.0 C) - in IPCC IPSL/Tiedke (TI) (0.3 C) old scheme ENSO has disappeared! Guilyardi et al. J.Clim (2009) Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

19 Impact of atmosphere convection scheme on ENSO Observations (0.9 C) - HadiSST1.1 Hourdin et al. (2006) Braconnot et al. (2007) IPSL (KE) Kerry Emanuel (1.0 C) - in IPCC IPSL/Tiedke (TI) (0.3 C) old scheme ENSO has disappeared! What role for α and µ? Guilyardi et al. J.Clim (2009) Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

20 BJ index for KE and TI Guilyardi et al. J.Clim (2009) Linear theory: α dominant factor in TI/KE difference Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

21 Impact of deep convection scheme on atmosphere feedbacks during ENSO Obs KE TI µ El Niño Amplitude ~ N.m -2 /C W.m -2 /C o C Error compensation! Too weak (improves with atmosphere resolution) Due to shortwave feedback difference (convection too strong in TI) Need to get the right ENSO amplitude for the right reasons! Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

22 Seasonal evolution of feedbacks Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

23 Seasonal evolution of feedbacks -25 W.m -2 /C <=> 1 o C/month in SST cooling!!! (MXL 50 m, SSTA of 2 o C) Shortwave HF feedback SW in second half of year explains most of the difference Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

24 Can we test this, i.e. suppress ENSO in KE? Perform KE run with increased SW Interannual Flux Correction: SHFO= SHFSC KE + SW mod (SSTO-SSTSC KE ) SW mod = -15 W.m -2 Mean state (SC) unchanged Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

25 Can we test this, i.e. suppress ENSO in KE? Perform KE run with increased SW Interannual Flux Correction: SHFO= SHFSC KE + SW mod (SSTO-SSTSC KE ) SW mod = -15 W.m -2 Mean state (SC) unchanged KE TI KE? mod TI Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

26 Can we test this, i.e. suppress ENSO in KE? Perform KE run with increased SW Interannual Flux Correction: SHFO= SHFSC KE + SW mod (SSTO-SSTSC KE ) SW mod = -15 W.m -2 Mean state (SC) unchanged µ El Niño Amplitude KE TI KE mod TI Obs KE TI KE mod TI ~ N/m 2 /C W/m 2 /C o C ENSO gone as well! Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

27 SW feedback distribution Point-wise regression of SHF anomaly vs. SSTA during JASOND (correl. less than 0.2 blanked out) Negative feedback (blue) = convective/ascent regime Positive feedback (red/orange) = subsidence regime ERA40 has large errors in East Pacific (Cronin et al. 2006) AMIP KE closer to ISCCP AMIP TI has too strong convection In KE, subsidence/+ve SW invades central Pacific In TI, convection/-ve SW invades east Pacific Coupled vs. forced (Yu & Kirtman 2007) Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

28 ENSO atmosphere feedbacks in CMIP3 models µ Lloyd et al. (2009, 2010) Both positive µ and negative α feedbacks generally underestimated in models (error compensation) αsw is the dominant source of diversity for α (clouds and regimes) Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

29 Summary El Niño in IPCC-class GCMs: significant progress in CMIP3 vs. previous generations still major errors (too much diversity, structure, timing,...) Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

30 Summary El Niño in IPCC-class GCMs: significant progress in CMIP3 vs. previous generations still major errors (too much diversity, structure, timing,...) Atmosphere GCM is a dominant contributor: Dynamical +ve (µ) and heat flux -ve (α) feedbacks both likely to control El Niño properties in CGCMs Both feedbacks are usually too weak in models Convection/low cloud scheme has direct impact on α Atmosphere GCM horizontal resolution can improve Bjerknes feedback µ Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

31 Summary El Niño in IPCC-class GCMs: significant progress in CMIP3 vs. previous generations still major errors (too much diversity, structure, timing,...) Atmosphere GCM is a dominant contributor: Dynamical +ve (µ) and heat flux -ve (α) feedbacks both likely to control El Niño properties in CGCMs Both feedbacks are usually too weak in models Convection/low cloud scheme has direct impact on α Atmosphere GCM horizontal resolution can improve Bjerknes feedback µ Errors in these feedbacks: are already visible in AMIP mode (good for model development) have an impact on the mean as well (improve the mean and you ll improve ENSO) Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

32 Community strategies to improve ENSO in models CLIVAR Workshop, Paris, France, November 2010 New strategies for evaluating ENSO processes in climate models Initiated by CLIVAR Pacific Panel (following BAMS and NGEO review papers) Brought together range of experts: ENSO theory, modelling, observations, forecasting,... The challenges lying ahead: Improve the quality and utility of historical records Maintain present ENSO observing system into the future Continue promoting intercomparison studies (ENSO metrics) Isolate the main sources of model error, guided by theory (BJ & other ENMICS), observations, and rigourous evaluation of the models Testing climate models in initialised mode (seasonal forecast, TAMIP,...) to understand the evolution of model biases Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

33 Using seasonal hindcasts: a new strategy to understand IPCC-class model errors Case studies rather than 100+ year long statistics Analyse initial adjustment phase of hindcast Detailed and process-based comparison to wealth of recent observations (TAO, JASON, A-train,...) Good configuration to look at cloud-sst and wind-sst interactions Links with - WGSIP s Climate Forecast Historical Project (CFHP) - CMIP5 near-term decadal simulations

34 Using seasonal hindcasts to understand ENSO errors EU ENSEMBLES seasonal hindcasts (stream 2) 5 models, 4x50 hindcasts Analysis of model errors as a function of hindcast leadtime Mean annual cycle Vannière et al. (2011) Models simulate correctly either El Niño or La Niña, but not both Models converge towards their own annual cycle when departing from obs.

35 Using seasonal hindcasts to understand ENSO errors Analysis of ENSO atmosphere feedbacks as a function of hindcast leadtime Vannière et al. (2011) µ and SW stabilized after 1-2 months (although mean state still drifting)

36 Outcome of ENSO CLIVAR workshop Science discussions, e.g.: Consensus on basic mechanisms driving ENSO, but not on ENSO triggers (WWBs, mid-latitudes footprinting, Indian ocean, etc.) Process-oriented tools/models to diagnose ENSO properties in GCMs Decadal variability / diversity of ENSO ENSO problem is far from understood or solved Opportunity to narrow the gap between ENSO theoreticians, observationalists, and operational community Catalogue of ENSO evaluation methods for GCMs (incl. coordination of CMIP5 analysis) Review paper on ENSO progress and challenges Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

37 Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

38 Observations uncertainty How long do we need to observe the tropical Pacific for? 20 centuries of nino3 SSTs Wittenberg (GRL, 2009)

39 Observations uncertainty How long do we need to observe the tropical Pacific for? Range of 100yr spectra Modulation of nino3 SST spectra Wittenberg (GRL, 2009)

40 ENSO atmosphere feedbacks: what about µ? (Zonal x Merid.) µ El Niño Amplitude Obs R97E (3.75 x 2.5) R99A (3.75 x 1.8) R149A (2.5 x 1.8) R1414A (2.5 x 1.2) R1914E (1.8 x 1.2) Varying the horizontal atmosphere resolution in IPSL-CM N/m 2 /C W/m 2 /C o C AGCM resolution affects µ (but not ): Atmosphere grid can see ocean equatorial wave guide Added non-linearities in AGCM: better circulation (on/off convection behavior reduced,...) Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

41 Composite analysis of large-scale regimes during El Niño Subsidence Convection in far east Pacific Positive SW has the potential to amplify El Niño during growing phase, i.e. before convective/ascent threshold is reached AMIP KE potential much larger than that of AMIP TI KE and TI have opposite regimes during ENSO in far east Pacific TI ascent threshold 2 o C lower than KE TI physics triggers convection too easily which prevents ENSO from developping Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

42 Heat flux feedback components Split net feedback into four components: shortwave (SW), longwave (LW), latent heat (LH) and sensible heat (SH) flux feedbacks. SW (blue bars) and LH (green bars) components dominate. Lloyd et al. (Clim. Dyn. 2010) The SW feedback shows the greatest model diversity linked to cloud uncertainties in East Pacific region (e.g. Sun et al. 2009)

43 ENSO performance metrics a few examples... Devised by CLIVAR Pacific Panel WG for CMIP5 Amplitude error in Niño 3 and Niño 4 ENSO Frequency RMS Zonal wind RMS error in Equatorial Pacific SST annual cycle amplitude error in Niño 3 Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

44 SST threshold for ascent/convection Bin vertical velocity at 500 hpa in SST bins Subsidence Ascent/Convective threshold when regime switches from subsidence to convection (Bony et al. 2004) Convection AMIP KE threshold larger by 1 o C / AMIP TI (same SST!) KE threshold unchanged TI threshold even lower: 2 o C difference with KE! TI physics triggers convection too easily which prevents ENSO from developping Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

45 Bjerknes feedback µ Bjerknes feedback in AMIP KE and AMIP TI similar and within re-analysis estimates Bjerknes feedback in KE = 1/3 rd of that of AMIP KE Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

46 Heat flux feedback α Heat flux feedback α in AMIP TI is double that of AMIP KE (and closer to re-analysis estimates) Heat flux feedback in KE = half of that of AMIP KE Value unchanged in TI Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

47 Evolution of El Nino composite Wind Stress Wind stress anomaly (shading) SST anomaly (solid contours) 3 year composite at equator Bjerknes feedback in AMIP KE and AMIP TI similar Bjerknes feedback in KE much weaker than in AMIP KE Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

48 Evolution of El Nino composite Heat Flux Heat flux anomaly (shading) SST anomaly (solid contours) 3 year composite at equator Negative Heat Flux anomaly during positive SST anomaly Negative Heat Flux anomaly in AMIP TI much larger than in AMIP KE SHF main contributor to differences between AMIP TI and AMIP KE (LHF also contributes) Negative Heat Flux anomaly in KE too far west and weak Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

49 Mean seasonal cycle at Eq. Wind stress (shading) SST (solid contours) Precipitation (3 and 8 mm/day dashed) AMIP KE performs rather well Convection in AMIP TI too strong and triggered too soon Biases amplified in coupled mode Semi-annual cycle in TI Equinoctial Central American monsoon too strong in TI (Braconnot et al. 2007) Workshop on Sub-seasonal to Seasonal Prediction, Exeter, Dec

50 Feedback Relationship to ENSO amplitude Reanalyses Reanalyses α Feedback: Models with stronger heat flux damping (more nega>ve α) tend to exhibit weaker ENSO, and vice versa. Lloyd et al. (2009)