Exploring Key Physics in Modeling the Madden Julian Oscillation (MJO)

Transcription

1 UCLA Exploring Key Physics in Modeling the Madden Julian Oscillation (MJO) A Joint YOTC / MJO Task Force and GEWEX GASS Global Model Evaluation Project Component I: Climate Simulations Xian an Jiang 1,2 1 Joint Institute for Regional Earth System Sci. & Engineering / UCLA, USA 2 Jet Propulsion Laboratory, California Institute of Technology, USA Duane Waliser 2, Prince Xavier 3, Jon Petch 3, Nick Klingaman 4,and Steve Woolnough 4 3 Met Office, Exeter, UK ; 4 National Center for Atmospheric Sciences Climate, Reading University, UK Participating Modeling Groups Acknowledgments: NSF/Climate & Large scale Dynamics

2 Vertical Structure and Diabatic Processes of the MJO: Global Model Evaluation Project MJO Task Force/YOTC and GASS Model Experiment Science Focus Exp. POC I. 20 Yr Climatological Simulations ( if AGCM) 6-hr, Global Output Vertical Structure, Physical Tendencies Model MJO Fidelity Vertical structure Multi-scale Interactions: (e.g., TCs, Monsoon, ENSO) UCLA/JPL X. Jiang D. Waliser II. 2-Day MJO Hindcasts YOTC MJO Cases E & F (winter 2009)* Time Step, Indo-Pacific Domain Output Very Detailed Physical/Model Processes Heat and moisture budgets Model Physics Evaluation (e.g. Convection/Cloud/BL) Short range Degradation Met Office P. Xavier J. Petch III. 20-Day MJO Hindcasts YOTC MJO Cases E & F (winter 2009)* 3-hr, Global Output Elements of I & II MJO Forecast Skill State Evolution/Degradation Elements of I & II NCAS/Walker in. N. Klingaman S. Woolnough *DYNAMO Case TBD Commitments: About 30 Modeling Groups with AGCM and/or CGCM

3 Primary Goal of the Climate Simulation Component MJO Fidelity score Process-oriented score

4 Participating GCMs for Climate Simulation Component Model Horizontal Resolution Vertical Resolution References 1 01_NASAGMAO_GEOS o lon x 0.5 o lat 72 Molod et al., a_SPCCSM (CAM3 + POP) T42 (~2.8 o ) 30 Super parameterization (Stan et al., 2010) 3 03b_SPCAMP_AMIP T42 30 (Khairoutdinov et al 2008) _GISS_ModelE2 2.5 o lon x 2 o lat 40 (Schmidt et al. 2014) Notes 5 05_EC_GEM ~1.4 o 64 (Cote et al, 1998) 6 07_MIROC T85 (~1.5 o ) 40 (Watanabe et al. 2010) 7 10_MRI GCM T (Yukimoto et al, 2012) 8 11_CWB_GFS T119 (~1 o ) 40 (Liou et al., 1997) AMIP SST _PNU_CFSv1 T62 (~2 o ) 64 (Saha et al. 2006) 10 16_MPI_ECHAM6 (ECHAM6 + MPIOM) T63 ( ~2 o ) 47 (Stevens et al, 2013) 11 17_MetUM_GA o lon x 1.25 o lat 85 (Walters et al. 2011) 12 21_NCAR_CAM o lon x 0.9 o lat 30 (Neal et al, 2012) 13 22_NRL_NAVGEMv.01 T359 (37km) 42 (Hogan et al. 2014) 14 24_UCSD_CAM T42 (~ 2.8 o ) 30 (Zhang & Mu 2005) 15 27_NCEPCPC_CFSv2 T126 (~ 1 o ) 64 (Saha et al. 2013) 16 31a_CNRM_AM 17 31b_CNRM_CM (CNRM_AM+ NEMO) T127 (~1.4 o ) 31 (Voldoireet al. 2013) 18 31c_CNRM_ACM 19 34_CCCma_CanCM4 2.8 o 35 (Merryfield et al. 2013) 20 35_BCCAGCM2.1 T42 (~2.8 deg) 26 (Wu et al 2010) 21 36_FGOALS2.0 s R42 (2.8 o lonx1.6 o lat) 26 (Bao et al. 2013) 22 37_NCHU_ECHAM5 SIT T63 31 (Tseng et al. 2014) 23 39_TAMU_Modi CAM4 (CCSM4) 2.5 o lon x 1.9 o lat 26 (Lappen & Schmumacher 2012) 24 40_ACCESS (modified METUM) o lon x 1.25 o lat 85 (Zhu et al. 2013) 25 43_ISUGCM T42 (~ 2.8 o ) 18 (Wu and Deng 2013) 26 44_LLNL_CAM5ZMMicro 1.25 o lon x 0.9 o lat 30 (Song & Zhang 2011) 27 45_SMHI_ecearth3 T255(80km) 91 IFS cy36r4 Idealized tilted vertical heating

5 MJO Fidelity Lag regression of rainfall to Indian Ocean base point (70 90 o E; 5 o S 5 o N) day filtered dash line 5 m/s

6 MJO Skill Score by Rainfall Hovmoller Diagram (Indian Ocean & western Pacific averaged) Top 25% models Bottom 25% models SPCCSM SPCAM GISS_ModelE2 MRI_AGCM CNRM_CM NCHU_ECHAM5 TAMU_modCAM4 PNU_CFS

7 Wavenumber Frequency Spectra (Nov Apr) Kelvin Rossby MJO Symmetric E/W Ratio: Wave number 1 3; day Kelvin wave & MJO (Guo et al. 2014)

8 MJO skill by HovmÖller diagram Patt. Cor. vs E/W Ratio CORR=0.78 Pattern Correlation Coefficients East/Westward Ratio

9 Process-oriented metrics for the MJO Rainfall PDF Large scale rainfall partition Mean zonal wind over Indo Pacific warm pool Radiative vs convective heating ratio Vertical moisture profiles versus rainfall rate Normalized gross moist stability (NGMS)

10 Process-oriented metrics for the MJO Rainfall PDF Large scale rainfall partition Mean zonal wind over Indo Pacific warm pool Radiative vs convective heating ratio Vertical moisture profiles versus rainfall rate Normalized gross moist stability (NGMS)

11 Metric I. Vertical moisture profiles versus rainfall rate Composite RH Profile vs Rainrate Difference in lowertropospheric between high and low rain events 10% 5% Kim et al. 2014; Maloney et al log 10 (Rain) (Courtesy of D. Kim) Boreal winter, Indo Pacific ( o E; 10 o S 10 o N)

12 MJO fidelity vs hPa RH Diff (top 5% bottom 10%)

13 Metric II. Normalized Gross Moist Stability (NGMS) (Raymond et al. 2008; Benedict et al. 2013; Maloney et al. 2014) Γ v v s - moist entropy r - mixing ratio Efficiency of convection and associated circulation in discharging column moisture Weak positive or negative GMS is necessary to destabilize the MJO in an idealized moisture mode framework (Raymond et al. 2008; Sobel & Maloney 2013). Benedict et al. (2014); Maloney et al. (2014) (Courtesy of E. Maloney & J. Benedict)

14 MJO Fidelity vs Normalized Gross Moist Stability (NGMS) CNRM_ACM CNRM_CM CNRM_AM

15 Vertical Structure and MJO Fidelity

16 Vertical structures of the MJO (ERA Interim; lag 0 regression) u (m/s) T (k) w (hpa/s) Q (k/day) q (g/kg) MJO Score Model Skill in Vertical Structure vs MJO Fidelity u T w Q q Corr=0.77 Corr=0.51 Corr=0.78 Corr=0.70 Corr=0.65 Pattern Correlations of Vertical Profiles Corr=0.79 MJO Score Average Pattern Correlations of Five Variables

17 Vertical MJO Structures in Strong and Weak MJO Models u (m/s) T (k) w (hpa/s) Q (k/day) q (g/kg) ERA Interim Good GCMs Poor GCMs (10 o S 10 o N)

18 Horizontal Structure of the MJO at 850hPa ERA Interim Contour: rainfall Shading: divergence Vector: winds Good GCMs Poor GCMs

19 Summary About ¼ of total 28 GCMs are able to reasonably well capture the observed eastward propagating MJO; Atmosphere ocean coupling can significantly lead to improved simulations of the MJO in some GCMs; Environmental moisture constraint on convection and circulation convection feedback are two critical factors associated with model MJO performances; Significant differences in vertical structures of u wind, T, q, Q, w, associated with the intraseasonal rainfall are noted in strong and weak GCMs.

20 Thank you! Jiang, X., D. E. Waliser, P. K. Xavier, J. Petch, N. P. Klingaman, S. J. Woolnough, Bin Guan, Gilles Bellon, T. Crueger, Charlotte DeMott, C. Hannay, H. Lin, W. Hu, D. Kim, C. L. Lappen, M. M. Lu, H. Y. Ma, T. Miyakawa, J. A. Ridout, S. D. Schubert, J. Scinocca, K. H. Seo, E. Shindo, X. Song, C. Stan, W. L. Tseng, W. Wang, T. Wu, X. Wu, K. Wyser, G. J. Zhang, and H. Zhu, 2014: Vertical structure and physical processes of the Madden Julian Oscillation: Exploring key model physics in climate simulations, Journal of Geophysical Research Atmosphere, under review