• Arakawa, A., and W. H. Schubert, 1974: Interaction of a cumulus cloud ensemble with the large-scale environment. Part I. J. Atmos. Sci., 31 , 674701.

    • Search Google Scholar
    • Export Citation
  • Chen, T-C., and M-C. Yen, 1991: A study of the diabiatic heating associated with the Madden–Julian oscillation. J. Geophys. Res., 96 , (D7). 1316313177.

    • Search Google Scholar
    • Export Citation
  • Chen, T-C., and J-M. Chen, 1997: On the relationship between the streamfunction and velocity potential of the Madden–Julian oscillation. J. Atmos. Sci., 54 , 679685.

    • Search Google Scholar
    • Export Citation
  • Duffy, P. B., B. Govindasamy, J. P. Iorio, J. Milovich, K. R. Sperber, K. E. Taylor, M. F. Wehner, and S. L. Thompson, 2003: High-resolution simulations of global climate, part 1: present climate. Climate Dyn., 21 , 371390.

    • Search Google Scholar
    • Export Citation
  • Hack, J. J., 1994: Parameterization of moist convection in the National Center for Atmospheric Research Community Climate Model (CCM2). J. Geophys. Res., 99 , 55515568.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., and M. L. Salby, 1994: Life cycle of the Madden–Julian oscillation. J. Atmos. Sci., 51 , 22252237.

  • Inness, P. M., and J. M. Slingo, 2003: Simulation of the MJO in a coupled GCM. Part I: Comparison with observations and an atmosphere-only GCM. J. Climate, 16 , 345364.

    • Search Google Scholar
    • Export Citation
  • Inness, P. M., J. M. Slingo, E. Guilyardi, and J. Cole, 2003: Simulation of the Madden–Julian oscillation in a coupled general circulation model. Part II: The role of the basic state. J. Climate, 16 , 365382.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77 , 437471.

  • Kiehl, J. T., J. J. Hack, G. B. Bonan, B. A. Boville, D. L. Williamson, and P. J. Rasch, 1998: The National Center for Atmospheric Research Community Climate Model: CCM3. J. Climate, 11 , 11311149.

    • Search Google Scholar
    • Export Citation
  • Kiladis, G. N., K. H. Straub, and P. T. Haertel, 2005: Zonal and vertical structure of the Madden–Julian oscillation. J. Atmos. Sci., 62 , 27902809.

    • Search Google Scholar
    • Export Citation
  • Knutson, R. R., K. M. Weickmann, and J. E. Kutzbach, 1986: Global-scale intraseasonal oscillations of outgoing longwave radiation and 250-mb zonal wind during Northern Hemisphere summer. Mon. Wea. Rev., 114 , 605623.

    • Search Google Scholar
    • Export Citation
  • Krishnamurti, T., and D. Subrahmanyam, 1982: The 30–50 day mode at 850 mb during MONEX. J. Atmos. Sci., 39 , 20882095.

  • Lanczos, C., 1956: Applied Analysis. Prentice-Hall, 539 pp.

  • Lau, K-M., and D. E. Waliser, Eds.,. 2005: Intraseasonal Variability of the Atmosphere-Ocean Climate System. Springer, 436 pp.

  • Liebmann, B., and C. A. Smith, 1996: Description of a complete (interpolated) OLR dataset. Bull. Amer. Meteor. Soc., 77 , 12751277.

  • Lin, J-L., and Coauthors, 2006: Tropical intraseasonal variability in 14 IPCC AR4 climate models. Part I: Convective signals. J. Climate, 19 , 26652690.

    • Search Google Scholar
    • Export Citation
  • Liu, P., B. Wang, K. R. Sperber, T. Li, and G. A. Meehl, 2005: MJO in the NCAR CAM2 with the Tiedtke convective scheme. J. Climate, 18 , 30073020.

    • Search Google Scholar
    • Export Citation
  • Madden, R. A., and P. R. Julian, 1971: Detection of a 40–50 day oscillation in the zonal wind in the tropical Pacific. J. Atmos. Sci., 28 , 702708.

    • Search Google Scholar
    • Export Citation
  • Madden, R. A., and P. R. Julian, 1972: Description of global-scale circulation cells in the tropics with a 40–50 day period. J. Atmos. Sci., 29 , 11091123.

    • Search Google Scholar
    • Export Citation
  • Madden, R. A., and P. R. Julian, 1994: Observations of the 40-50-day tropical oscillation—A review. Mon. Wea. Rev., 122 , 814837.

  • Maloney, E. D., and D. L. Hartmann, 2001: The sensitivity of intraseasonal variability in the NCAR CCM3 to changes in convective parameterization. J. Climate, 14 , 20152034.

    • Search Google Scholar
    • Export Citation
  • Nakazawa, T., 1988: Tropical super clusters within intraseasonal variations over the western Pacific. J. Meteor. Soc. Japan, 66 , 823839.

    • Search Google Scholar
    • Export Citation
  • Rui, H., and B. Wang, 1990: Development characteristics and dynamic structure of tropical intraseasonal convection anomalies. J. Atmos. Sci., 47 , 357379.

    • Search Google Scholar
    • Export Citation
  • Salby, M. L., and H. H. Hendon, 1994: Intraseasonal behavior of clouds, temperature, and motion in the tropics. J. Atmos. Sci., 51 , 22072224.

    • Search Google Scholar
    • Export Citation
  • Slingo, J. M., and Coauthors, 1996: The intraseasonal oscillation in 15 atmospheric general circulation models: Results from an AMIP diagnostic subproject. Climate Dyn., 12 , 325357.

    • Search Google Scholar
    • Export Citation
  • Song, X., X. Wu, G. J. Zhang, and R. W. Arritt, 2008a: Dynamical effects of convective momentum transports on global climate simulations. J. Climate, 21 , 180194.

    • Search Google Scholar
    • Export Citation
  • Song, X., X. Wu, G. J. Zhang, and R. W. Arritt, 2008b: Understanding the effects of convective momentum transport on climate simulations: The role of convective heating. J. Climate, 21 , 50345047.

    • Search Google Scholar
    • Export Citation
  • Sperber, K. R., 2003: Propagation and the vertical structure of the Madden–Julian oscillation. Mon. Wea. Rev., 131 , 30183037.

  • Sperber, K. R., 2004: Madden–Julian variability in NCAR CAM2.0 and CCSM2.0. Climate Dyn., 23 , (3–4). 259278.

  • Sperber, K. R., S. Gualdi, S. Legutke, and V. Gayler, 2005: The Madden–Julian oscillation in ECHAM4 coupled and uncoupled GCMs. Climate Dyn., 25 , 117140.

    • Search Google Scholar
    • Export Citation
  • Sud, Y. C., and G. K. Walker, 1999: Microphysics of Clouds with the Relaxed Arakawa–Schubert Scheme (McRAS). Part I: Design and evaluation with GATE phase III data. J. Atmos. Sci., 56 , 31963220.

    • Search Google Scholar
    • Export Citation
  • Tiedtke, M., 1989: A comprehensive mass flux scheme for cumulus parameterization in large-scale models. Mon. Wea. Rev., 117 , 17791800.

    • Search Google Scholar
    • Export Citation
  • Tokioka, T., K. Yamazaki, A. Kitoh, and T. Ose, 1988: The equatorial 30-60 day oscillation and the Arakawa-Schubert penetrative cumulus parameterization. J. Meteor. Soc. Japan, 66 , 883901.

    • Search Google Scholar
    • Export Citation
  • Tung, W. W., and M. Yanai, 2002: Convective momentum transport observed during the TOGA COARE IOP. Part II: Case studies. J. Atmos. Sci., 59 , 25352549.

    • Search Google Scholar
    • Export Citation
  • Waliser, D. E., K. M. Lau, and J. H. Kim, 1999: The influence of coupled sea surface temperatures on the Madden–Julian oscillation: A model perturbation experiment. J. Atmos. Sci., 56 , 333358.

    • Search Google Scholar
    • Export Citation
  • Wang, W., and M. E. Schlesinger, 1999: The dependence on convection parameterization of the tropical intraseasonal oscillation simulated by the UIUC 11-layer atmospheric GCM. J. Climate, 12 , 14231457.

    • Search Google Scholar
    • Export Citation
  • Weickmann, K. M., 1983: Intraseasonal circulation and outgoing longwave radiation modes during Northern Hemisphere winter. Mon. Wea. Rev., 111 , 18381858.

    • Search Google Scholar
    • Export Citation
  • Weickmann, K. M., G. R. Lussky, and J. E. Kutzbach, 1985: Intraseasonal (30–60 day) fluctuations of outgoing longwave radiation and 250 mb streamfunction during northern winter. Mon. Wea. Rev., 113 , 941961.

    • Search Google Scholar
    • Export Citation
  • Wu, X., and M. Yanai, 1994: Effects of vertical wind shear on the cumulus transport of momentum: Observations and parameterization. J. Atmos. Sci., 51 , 16401660.

    • Search Google Scholar
    • Export Citation
  • Wu, X., and X-Z. Liang, 2005a: Radiative effects of cloud horizontal inhomogeneity and vertical overlap identified from a month-long cloud-resolving simulation. J. Atmos. Sci., 62 , 41054112.

    • Search Google Scholar
    • Export Citation
  • Wu, X., and X-Z. Liang, 2005b: Effect of subgrid cloud-radiation interaction on climate simulations. Geophys. Res. Lett., 32 , L24806. doi:10.1029/2005GL024432.

    • Search Google Scholar
    • Export Citation
  • Wu, X., X-Z. Liang, and G. J. Zhang, 2003: Seasonal migration of ITCZ precipitation across the equator: Why can’t GCMs simulate it? Geophys. Res. Lett., 30 , 1824. doi:10.1029/2003GL017198.

    • Search Google Scholar
    • Export Citation
  • Wu, X., L. Deng, X. Song, G. Vettoretti, W. R. Peltier, and G. J. Zhang, 2007a: Impact of a modified convective scheme on the Madden-Julian Oscillation and El Niño–Southern Oscillation in a coupled climate model. Geophys. Res. Lett., 34 , L16823. doi:10.1029/2007GL030637.

    • Search Google Scholar
    • Export Citation
  • Wu, X., L. Deng, X. Song, and G. J. Zhang, 2007b: Coupling of convective momentum transport with convective heating in global climate simulations. J. Atmos. Sci., 64 , 13341349.

    • Search Google Scholar
    • Export Citation
  • Xie, P., and P. A. Arkin, 1996: Analyses of global monthly precipitation using gauge observations, satellite estimates, and numerical model predictions. J. Climate, 9 , 840858.

    • Search Google Scholar
    • Export Citation
  • Xie, P., and P. A. Arkin, 1997: Global precipitation 17-year monthly analysis based on gauge observations, satellite estimates, and numerical model outputs. Bull. Amer. Meteor. Soc., 78 , 25392558.

    • Search Google Scholar
    • Export Citation
  • Yanai, M., B. Chen, and W-W. Tung, 2000: The Madden–Julian oscillation observed during the TOGA COARE IOP: Global view. J. Atmos. Sci., 57 , 23742396.

    • Search Google Scholar
    • Export Citation
  • Zangvil, A., 1975: Temporal and spatial behavior of large-scale disturbances in tropical cloudiness deduced from satellite brightness data. Mon. Wea. Rev., 103 , 904920.

    • Search Google Scholar
    • Export Citation
  • Zangvil, A., and M. Yanai, 1981: Upper tropospheric waves in the tropics. Part II: Association with clouds in the wavenumber-frequency domain. J. Atmos. Sci., 38 , 939953.

    • Search Google Scholar
    • Export Citation
  • Zhang, C., 2005: Madden-Julian oscillation. Rev. Geophys., 43 , RG2003. doi:10.1029/2004RG000158.

  • Zhang, C., and M. Dong, 2004: Seasonality in the Madden–Julian oscillation. J. Climate, 17 , 31693180.

  • Zhang, G. J., 2002: Convective quasi-equilibrium in midlatitude continental environment and its effect on convective parameterization. J. Geophys. Res., 107 , 4220. doi:10.1029/2001JD001005.

    • Search Google Scholar
    • Export Citation
  • Zhang, G. J., and H. R. Cho, 1991: Parameterization of the vertical transport of momentum by cumulus clouds. Part I: Theory. J. Atmos. Sci., 48 , 14831492.

    • Search Google Scholar
    • Export Citation
  • Zhang, G. J., and N. A. McFarlane, 1995: Sensitivity of climate simulations to the parameterization of cumulus convection in the Canadian Climate Centre general circulation model. Atmos.–Ocean, 33 , 407446.

    • Search Google Scholar
    • Export Citation
  • Zhang, G. J., and X. Wu, 2003: Convective momentum transport and perturbation pressure field from a cloud-resolving model simulation. J. Atmos. Sci., 60 , 11201139.

    • Search Google Scholar
    • Export Citation
  • Zhang, G. J., and M. Mu, 2005: Simulation of the Madden–Julian oscillation in the NCAR CCM3 using a revised Zhang–McFarlane convection parameterization scheme. J. Climate, 18 , 40464064.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    October–April climatology of precipitation rate (mm day−1) for (a) CMAP, (b) ISUCCM3, and (c) CTL.

  • View in gallery

    1987 Hovmöller diagrams of 850-hPa zonal wind anomaly (m s−1) averaged between 5°N and 5°S for (a) NCEP–NCAR reanalysis, (b) ISUCCM3, and (c) CTL. Areas of negative (easterly anomaly) are shaded. The contour interval is 5 m s−1.

  • View in gallery

    Spatial distribution of variance of 20–70-day bandpassed OLR for the four seasons from 1979 to 1988: (a) AVHRR, (b) ISUCCM3, (c) NOCMT, (d) NOTRI, and (e) CTL. (top left) December–February (DJF), (top right) March–May (MAM), (bottom left) June–August (JJA), and (bottom right) September–November (SON). Unit: W2 m−4. Regions of deep convection are highlighted (shaded > 240 W2 m−4). The contour interval is 80 W2 m−4.

  • View in gallery

    Ten-years (October–April 1979–88) lag correlations of 20–100-day bandpassed OLR with 200-hPa velocity potential for (a) NCEP–NCAR reanalysis, (b) ISUCCM3, (c) NOCMT, (d) NOTRI, and (e) CTL. The white solid line represents a phase speed of ∼5 m s−1.

  • View in gallery

    Wavenumber–frequency spectra of 200-hPa zonal wind averaged between 10°N and 10°S for (a) NCEP–NCAR reanalysis, (b) ISUCCM3, (c) NOCMT, (d) NOTRI, and (e) CTL (years 1979–88). The contour starts from 0.05 m2 s−2, with an interval of 0.05 m2 s−2.

  • View in gallery

    Analysis of 20–100-day bandpassed AVHRR OLR for 10 yr (October–April 1979–88): (a) EOF1 and (b) EOF2.

  • View in gallery

    Zero-lag linear regressions of PC1 for 20–100-day filtered 10-yr (October–April 1979–88) (a) NCEP–NCAR reanalysis 200-hPa wind (m s−1) and OLR (W m−2); (b) NCEP–NCAR reanalysis 850-hPa wind (m s−1) and CMAP (mm day−1), (c) NCEP–NCAR reanalysis 200-hPa wind (m s−1) and streamfunction (m2 s−1), (d) NCEP–NCAR reanalysis 850-hPa wind (m s−1) and streamfunction (m2 s−1), (e) NCEP–NCAR reanalysis latent heat flux (W m−2), and (f) NCEP–NCAR reanalysis 500-hPa vertical velocity (Pa s−1; negative sign means upward motion). (g)–(l) Same as (a)–(f), but for regressions using PC2. All regressions have been scaled by a one standard deviation of PC1 to give units. The variables are plotted where the regression is 95% confidence or better.

  • View in gallery

    Same as Fig. 7, but for ISUCCM3.

  • View in gallery

    Same as Fig. 7, but for CTL.

  • View in gallery

    Same as Fig. 7, but for NOCMT.

  • View in gallery

    Same as Fig. 7, but for NOTRI.

  • View in gallery

    NCEP–NCAR reanalysis longitude–height cross sections of zero-lag linear regressions of PC1 with 5°N–5°S-averaged 20–100-day bandpass-filtered (a) divergence (s−1); (b) vertical velocity (Pa s−1); (c) zonal wind (m s−1) and vertical velocity (Pa s−1) vectors (vertical velocity times −100) and contours of zonal wind with the interval of 0.5 m s−1; and (d) specific humidity (g kg−1). All regressions have been scaled by a one standard deviation of PC1 to give units.

  • View in gallery

    Same as Fig. 12, but for ISUCCM3.

  • View in gallery

    Same as Fig. 12, but for CTL.

  • View in gallery

    Same as Fig. 12, but for NOCMT.

  • View in gallery

    Same as Fig. 12, but for NOTRI.

  • View in gallery

    Time lag vs height plots of linear regressions of PC1 with 90°E (5°N–5°S averaged) 20–100-day bandpass-filtered NCEP–NCAR reanalysis (a) divergence (s−1); (b) vertical velocity (Pa s−1; negative sign means upward motion); (c) zonal wind (m s−1) and vertical velocity (Pa s−1) vectors (vertical velocity times −100) and contours of zonal wind with the interval of 0.5 m s−1; and (d) specific humidity (g kg−1). All regressions have been scaled by a one standard deviation of PC1 to give units. Time lags run from −25 to 25 days.

  • View in gallery

    Same as Fig. 17, but for ISUCCM3.

  • View in gallery

    Same as Fig. 17, but for CTL.

  • View in gallery

    Same as Fig. 17, but for NOCMT.

  • View in gallery

    Same as Fig. 17, but for NOTRI.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 51 47 5
PDF Downloads 20 19 5

Effects of Convective Processes on GCM Simulations of the Madden–Julian Oscillation

View More View Less
  • 1 Department of Geological and Atmospheric Sciences, Iowa State University, Ames, Iowa
© Get Permissions
Full access

Abstract

Weak temporal variability in tropical climates such as the Madden–Julian oscillation (MJO) is one of major deficiencies in general circulation models (GCMs). The uncertainties in the representation of convection and cloud processes are responsible for these deficiencies. With the improvement made to the convection scheme, the Iowa State University (ISU) GCM, which is based on a version of the NCAR Community Climate Model, is able to simulate many features of MJO as revealed by observations. In this study, four 10-yr (1979–88) ISU GCM simulations with observed sea surface temperatures are analyzed and compared to examine the effects of the revised convection closure, convection trigger condition, and convective momentum transport (CMT) on the MJO simulations. The modifications made in the convection scheme improve the simulations of amplitude, spatial distribution, eastward propagation, and horizontal and vertical structures, especially for the coherent feature of eastward-propagating convection and the precursor sign of convective center. The revised convection closure plays a key role in the improvement of the eastward propagation of MJO. The convection trigger helps produce less frequent but more vigorous moist convection and enhance the amplitude of the MJO signal. The inclusion of CMT results in a more coherent structure for the MJO deep convective center and its corresponding atmospheric variances.

Corresponding author address: Liping Deng, 3010 Agronomy Hall, Iowa State University, Ames, IA 50011. Email: liping@iastate.edu

Abstract

Weak temporal variability in tropical climates such as the Madden–Julian oscillation (MJO) is one of major deficiencies in general circulation models (GCMs). The uncertainties in the representation of convection and cloud processes are responsible for these deficiencies. With the improvement made to the convection scheme, the Iowa State University (ISU) GCM, which is based on a version of the NCAR Community Climate Model, is able to simulate many features of MJO as revealed by observations. In this study, four 10-yr (1979–88) ISU GCM simulations with observed sea surface temperatures are analyzed and compared to examine the effects of the revised convection closure, convection trigger condition, and convective momentum transport (CMT) on the MJO simulations. The modifications made in the convection scheme improve the simulations of amplitude, spatial distribution, eastward propagation, and horizontal and vertical structures, especially for the coherent feature of eastward-propagating convection and the precursor sign of convective center. The revised convection closure plays a key role in the improvement of the eastward propagation of MJO. The convection trigger helps produce less frequent but more vigorous moist convection and enhance the amplitude of the MJO signal. The inclusion of CMT results in a more coherent structure for the MJO deep convective center and its corresponding atmospheric variances.

Corresponding author address: Liping Deng, 3010 Agronomy Hall, Iowa State University, Ames, IA 50011. Email: liping@iastate.edu

1. Introduction

After the discovery of the Madden–Julian oscillation (MJO) by Madden and Julian (1971, 1972), its characteristics and structure have been extensively analyzed by many observational studies (e.g., Madden and Julian 1994; Lau and Waliser 2005; Zhang 2005). The MJO is a dominant intraseasonal variability in the tropical atmosphere. The MJO-related convection develops and propagates eastward along the equator in the Indian Ocean, tends to propagate into the South Pacific convergence zone (SPCZ) in the western Pacific, and decays in the central Pacific (Madden and Julian 1972). Convectively coupled MJO signals have been detected in various properties associated with deep convection and atmospheric circulations, such as the outgoing longwave radiation (OLR), surface precipitation, 200-hPa velocity potential and zonal wind, 850-hPa zonal wind, and surface latent heat flux (e.g., Zangvil 1975; Zangvil and Yanai 1981; Krishnamurti and Subrahmanyam 1982; Weickmann et al. 1985; Chen and Yen 1991; Slingo et al. 1996; Chen and Chen 1997; Maloney and Hartmann 2001; Tung and Yanai 2002; Sperber 2004; Zhang and Dong 2004). The major power of the MJO is concentrated at wavenumbers 1–3 and eastward periods of 30–90 days (e.g., Salby and Hendon 1994; Zhang 2005). The speed of eastward-propagating MJO is about 5 m s−1 (Weickmann et al. 1985; Knutson et al. 1986). Within the MJO, a hierarchical structure of cloud systems is identified by Nakazawa (1988), which includes several eastward-moving supercloud clusters (SCCs) near the equator over the western Pacific and several westward-moving cloud clusters within each SCC. The horizontal structure of the MJO shows the coupling of deep convection with the large-scale motion, with a pair of upper-level (200 hPa) anticyclonic circulations (with easterlies in between) and a pair of low-level (850 hPa) cyclonic circulations (with westerlies in between) on both sides of the equator in the Indian Ocean and western Pacific with the minimum OLR (e.g., Weickmann 1983; Rui and Wang 1990; Hendon and Salby 1994; Yanai et al. 2000; Kiladis et al. 2005). The MJO temperature, specific humidity, wind, divergence, and diabatic heating fields display asymmetry and westward tilt in the vertical (Sperber 2003; Kiladis et al. 2005). Low-level convergence, upward motion, and a positive moisture anomaly that is favored for the development of new convection and the eastward propagation are present on the east side of the MJO convective center, while low-level divergence, downward motion, and negative moisture anomaly exist on the west side.

Despite the progress in understanding the MJO by the observational study, the MJO simulation remains a major challenge for GCMs and NWP models. The unrealistic features in the MJO simulations include the weak amplitude, more power at higher frequencies than that in the observations, temporal and spatial distributions of MJO variances differing from those observed, eastward propagation speed being too fast, and a lack of coherent structure for the eastward propagation from the Indian Ocean to the Pacific (e.g., Slingo et al. 1996; Inness and Slingo 2003; Sperber 2004; Sperber et al. 2005; Zhang 2005). While some improvement in simulating MJO variance and coherent eastward propagation has been attributed to model resolutions (e.g., Inness et al. 2001; Sperber et al. 2005), model mean background state (e.g., Slingo et al. 1996; Inness et al. 2003; Sperber et al. 2005), and air–sea interaction (e.g., Waliser et al. 1999; Inness and Slingo 2003; Sperber 2004; Sperber et al. 2005), studies have shown that the model physics, and especially the representation of convective processes, may be the key to producing the realistic MJO simulations in GCMs. Convection affects large-scale circulation and wave disturbances through precipitation, latent heat release, and the redistribution of heat, moisture, and momentum. Because the organization and evolution of tropical convection is a major component of the MJO, the coupling of the convection scheme with large-scale dynamics is crucial for modeling the MJO.

Tokioka et al. (1988) showed that the addition of a minimum value of the cumulus entrainment rate of the environmental air in the Arakawa and Schubert (1974) convection scheme is a key factor for simulating the MJO by a GCM. Slingo et al. (1996) evaluated 15 atmospheric GCMs and found that the convection schemes with the convective available potential energy (CAPE)-type closure tend to produce better MJO signals than the moisture convergence–type closure. Wang and Schlesinger (1999) showed that the use of a large threshold of relative humidity allows for the accumulation of moist static energy to a certain amount to trigger the convection in three different convection schemes and improves the simulation of MJO signals. Maloney and Hartmann (2001) improved the MJO simulations by using the microphysics of cloud together with the relaxed Arakawa–Schubert convection scheme (Sud and Walker 1999), and they suggested that the simulations are sensitive to the parameterization of convective precipitation evaporation in an unsaturated airenvironment and unsaturated downdrafts. Liu et al. (2005) showed that a GCM with the Tiedtke (1989) convection scheme simulates an improved mean state, intraseasonal variability, space–time power spectra, and coherent eastward propagation of MJO-related precipitation. Lin et al. (2006) evaluated 14 coupled GCM simulations of MJO and found that the intraseasonal variance is too weak in most of the models; however, two models that have convective closure/triggers tied to the moisture convergence produce better MJO simulations than the others.

The above studies indicate that the factors in the convection schemes identified to be crucial for improving MJO simulations are not universal for GCMs. Therefore, understanding the physical mechanism responsible for the MJO simulations remains a major challenge. Physically sounded and observationally validated convection schemes are needed for undertaking this task. Recently, Zhang (2002) improved the closure assumption of the Zhang and McFarlane (1995) convection scheme by relating convection to the destabilization of the tropospheric layer above the planetary boundary layer by the large-scale processes based on the observations from the Atmospheric Radiation Measurement Program (ARM) and Tropical Ocean and Global Atmosphere Coupled Ocean–Atmosphere Response Experiment (TOGA COARE). Zhang and Mu (2005) showed that the revised closure assumption helps the MJO simulations by the National Center for Atmospheric Research (NCAR) GCM. Uncoupled and coupled simulations using the Iowa State University (ISU) GCM with various modifications to the convection, cloud, and radiation schemes demonstrate improvements on many aspects of global climate simulations such as the ENSO, the MJO, and the precipitation and energy budget (e.g., Wu et al. 2003; Wu and Liang 2005a,b; Wu et al. 2007a,b).

The ultimate goal of this project is to understand the mechanisms and physical processes through which convection affects the MJO. In this paper, the objectives are to evaluate the MJO simulated by the ISU GCM against observations, and to examine the impacts of revised convection closure, convection trigger, and convective momentum transport (CMT) on the simulations. The data and analysis techniques are described in section 2. The characteristics and structure of the simulated MJO are presented in section 3. The summary is given in section 4.

2. ISU GCM simulations, observational datasets, and analysis techniques

The ISU GCM is based on the NCAR Community Climate Model version 3 (CCM3), which is a spectral global climate model (Kiehl et al. 1998). It has been used worldwide for climate modeling and climate change studies. The resolution is user specifiable, with the most common implementation having 18 hybrid vertical levels extending from the surface to 4 hPa, and a horizontal resolution of T42 (a roughly 2.8° × 2.8° Gaussian grid). It has a highly sophisticated physical parameterization package for subgrid-scale processes such as boundary layer turbulence, radiation, clouds, and convection. Deep precipitating convection and shallow convection are treated by two different schemes, that is, the Zhang and McFarlane (1995) and Hack (1994) schemes, respectively. The Zhang and McFarlane deep convection scheme makes use of the ensemble plume concept to represent convective clouds (Arakawa and Schubert 1974) and simplifies it to a bulk mass flux form. It also includes representation of saturated convective-scale downdrafts as an inverted plume. The Hack shallow convection scheme is a modified moist convective adjustment scheme in the mass flux form. The scheme checks the local instability of the temperature stratification, as opposed to the deep instability used in the deep convection scheme, and adjusts the three adjacent model layers, where local instability is present, to a neutral state.

Three modifications, that is, a revised convection closure assumption, convection trigger condition, and CMT, are made to the deep convection scheme in the ISU GCM. The revised closure assumption is based on ARM and TOGA COARE observations. It relates convection to the destabilization of the tropospheric layer above the planetary boundary layer by the large-scale processes (Zhang 2002). The trigger condition for deep convection is based on the cloud-resolving simulations, that is, the convection is activated when the CAPE increase resulting from the large-scale forcing exceeds certain threshold (70 J kg−1 h−1; Wu et al. 2007a). The CMT parameterization scheme is validated by and simplified based on the cloud-resolving simulations (Zhang and Cho 1991; Wu et al. 2003; Zhang and Wu 2003). The CMT scheme considers the vertical redistribution of the horizontal momentum by convection, and accounts for the role of perturbation pressure field generated by the interaction of convection with large-scale circulation in vertical momentum transport (Wu and Yanai 1994). The CMT-induced convective heating plays an important role in shaping up the Hadley circulation (Song et al. 2008a,b).

Four 10-yr (1979–88) simulations with observed sea surface temperatures are analyzed (Table 1). The control simulation (CTL) is conducted using the ISU GCM with the original deep convection scheme as in standard CCM3, while the simulation ISUCCM3 is performed with the inclusion of all three modifications in the convection scheme. Two sensitivity simulations, that is, NOCMT and NOTRI, are performed to investigate the impacts of each of three modifications on the MJO. The simulation NOCMT only includes the revised closure and trigger condition in the convection scheme, and the simulation NOTRI only applies the revised closure in the scheme.

The pentad (5-day mean) precipitation product of the Climate Prediction Center (CPC) Merged Analysis of Precipitation (CMAP) is used to evaluate the climate simulations. Observations from rain gauges and precipitation estimates from several satellite-based algorithms (infrared and microwave) are merged by the technique described in Xie and Arkin (1996, 1997). The Advanced Very High Resolution Radiometer (AVHRR) OLR data are interpolated in time and space from the National Oceanic and Atmospheric Administration (NOAA) twice-daily OLR values and are averaged to once daily (Liebmann and Smith 1996). National Centers for Environmental Prediction (NCEP)–NCAR reanalysis datasets (Kalnay et al. 1996), including the wind, OLR, precipitation, vertical velocity, latent heat flux, and specific humidity, are used for the comparison. The datasets are regridded to T42 resolution for comparison with the model output.

In this paper, two analysis methods are applied to examine the MJO characteristics and structure. The empirical orthogonal function (EOF) analysis is used to obtain physical and dynamic independent patterns from the datasets. Each of the spatial EOF pattern is associated with a temporal principal component (PC). The first several EOFs can explain the majority of the data variability. Many analyses presented in the following sections are obtained by projecting simulated fields onto the observed EOF patterns and using the PC time series for the regression. The wavenumber–frequency spectra analysis (Maloney and Hartmann 2001) is also used to examine the MJO power in the study. This method resolves the transient wave along the latitudinal belt into eastward- and westward-moving components. Positive (negative) frequency and period represent the eastward (westward) propagation.

3. Characteristics and structure of the simulated MJO

a. Basic features

Before presenting the MJO analysis, the 10-yr (1979–88) October–April mean precipitation is first checked to identify the impacts of modified convection scheme on the mean state, which provides important background for the MJO activities. The distribution of precipitation in CTL shows a split ITCZ, and the amplitude over the equatorial western Pacific and Indian Ocean is much weaker than CMAP (Figs. 1a,c). The split-ITCZ problem is improved with the inclusion of new closure in the convection scheme (not shown). With the convection trigger condition added in the scheme, the amplitude of precipitation along the equator is enhanced (not shown). With all three modifications in the convection scheme, the ISUCCM3 precipitation shows a better agreement with CMAP in both amplitude and distribution (Fig. 1b). The simulation of the SPCZ precipitation belt is also improved with the modified convection scheme. The ISUCCM3 produces a northwest–southeast precipitation belt like CMAP, but a west–east belt is simulated by CTL.

The impacts of a modified convection scheme on the MJO simulations can be visually identified from the Hovmöller diagram of 850-hPa zonal wind anomalies averaged across the equatorial belt (5°N–5°S) during 1987 (Fig. 2). The anomalies are obtained by subtracting the zonal and time means. Several eastward-propagating bands of anomalous westerlies can be found in the observations (Fig. 2a). These MJO-related disturbances start from the Indian Ocean to the western Pacific, and in some cases it extends to the eastern Pacific. ISUCCM3 shows an eastward-propagating MJO signal, but with more high-frequency variability than the observations (Fig. 2b). The life cycle of the MJO signal in ISUCCM3 is also shorter than that in the observations. However, the dominant westward-propagating anomalies are produced by CTL (Fig. 2c).

To identify the MJO variability, the variance of daily observed AVHRR OLR for the period of 10 yr (1979–88) is presented in Fig. 3a. A 20–70-day Lanczos (1956) filter has been used to highlight the MJO-related variability. The observed largest variance of OLR is coincident with the regions of the heaviest mean precipitation and the MJO-related strong convective center over the Indian Ocean and western Pacific (not shown). The minimum over the Maritime Continent may be due to the strong land heating (cooling) during the day (night), which tends to favor the variability of convections occurs on shorter time scales (e.g., diurnal cycle) than the 20–70-day period. The ISUCCM3 catches those features with a 10-yr (1979–88)-averaged 20–70-day OLR variance (Fig. 3b), which is stronger than the observations, but CTL does not show a well-defined MJO variance over the Indian Ocean and western Pacific, especially near the equator (Fig. 3e). The largest OLR variance, which corresponds to the center of deep convection, shifts from the southern tropics to the Asian monsoon region in both the ISU GCM and the observations from winter [December–February (DJF)] to summer [June–August (JJA)]. The simulations of spatial distribution and amplitude of the MJO OLR variance for all four seasons are improved in the ISUCCM3 as compared to the CTL. In the NOCMT run, which excludes the CMT from the convection scheme, the spatial structure of OLR variances (Fig. 3c) is similar to that in ISUCCM3 (Fig. 3b), but the amplitude is weaker than that in the ISUCCM3. For example, over the western Pacific, the NOCMT OLR variance around 160°E is smaller than the ISUCCM3 variance in DJF. In the NOTRI run, which only includes the revised convection closure, the spatial distribution and amplitude of the OLR variance (Fig. 3d) are quite different from those in NOCMT (Fig. 3c), but are similar to those in CTL (Fig. 3e), which suggests that the trigger condition plays an important role in simulating the intraseasonal variability.

Figure 4 shows lag correlations of daily values of OLR at each longitude with a base time series of daily 200-hPa velocity potential at 90°E, both averaged between 10°N and 10°S and filtered to retain the variability at periods of 20–100 days. The observed eastward propagation of the MJO convective activity shows clearly, with positive correlations extending from the western Indian Ocean at a lag of −15 days to the date line at a lag of around +20 days (a phase speed of ∼5 m s−1; see Fig. 4a). The ISUCCM3, like the observations, shows the MJO eastward propagation from the Indian Ocean to the western Pacific with a phase speed of ∼5 m s−1 (Fig. 4b), but with smaller amplitudes and shorter periods (the MJO signal sustains an about 20-day span compared to the 30-day span in the NCEP–NCAR reanalysis). The main signal of CTL is a westward propagation centered at 90°E, and a weak signal appears to propagate eastward from 120°E to 180° (Fig. 4e). When the CMT is removed, the eastward propagation of the MJO convective activity in NOCMT (Fig. 4c) is similar to that in ISUCCM3 and observations, but the MJO phase speed of ∼10 m s−1 is greater than that in ISUCCM3. When both the CMT and convection trigger condition are removed and only the revised convection closure is kept in the scheme, NOTRI still simulates an eastward propagation, but the phase speed of ∼15 m s−1 is even faster than the speed simulated by NOCMT and the eastward propagation stops around 150°E (Fig. 4d). The impacts of both CMT and the convection trigger condition decrease the speed of eastward propagation. The inclusion of the revised convection closure in NOTRI results in the eastward-propagating signal, while the old convection closure produces the westward propagation in CTL.

Averaged wavenumber–frequency spectra for observed equatorial (10°N–10°S) 200-hPa zonal wind are plotted in Fig. 5a. Ten-year data are used and a high-pass filter is applied to remove the periods of the annual cycle and lower before the spectra computation. The observed zonal wind spectrum is dominated by the power at wavenumber 1 and eastward periods of 30–90 days. The maximum variances are near 60 and 35 days and wavenumber 1. Both ISUCCM3 and CTL show a preference for the eastward power, especially in wavenumber 1 (Figs. 5b,e), which is similar to the observations. However, the variance at intraseasonal time scales for ISUCCM3 is larger than that for CTL, and is closer to the observations. With the removal of CMT from the scheme, the maximum power in the NOCMT simulation is near a period of 60 days and wavenumber 1 (Fig. 5c), which is similar to the ISUCCM3. However, the peak of power at wavenumber 1 and eastward periods of 25–33 days exits in NOCMT, but disappears in ISUCCM3. This suggests that, with the impact of CMT, ISUCCM3 tends to have less power at slightly higher frequency (around 30 days). With both CMT and the convective trigger excluded from the scheme, NOTRI produces much less variance at intraseasonal time scales than NOCMT for eastward-propagating disturbances, especially near 60 days and wavenumber 1 (Fig. 5d). This indicates that the impact of the convection trigger tends to enhance the MJO convection activity. The comparison between NOTRI and CTL shows that the revised convection closure plays an important role in simulating the eastward-propagating signal at higher frequencies (Figs. 5d,e).

b. Horizontal structure

In this section, the horizontal structure of simulated MJO in the upper and lower troposphere is compared with observations using the EOF analysis. Following the approach used by Duffy et al. (2003) and Sperber (2004), the simulated fields are projected onto the observed leading patterns to ensure that all four simulations are treated identically. The observed leading patterns defined here are based on 10-yr (1979–88) AVHRR OLR. The EOF analysis on bandpassed 10-yr AVHRR OLR is performed to show the MJO convection signature (Fig. 6). The total explained variance of first two modes is 23.7%. EOF1 suggests an MJO convective center over the Indian Ocean around 90°E (Fig. 6a). EOF2 shows an MJO convective center over the western Pacific around 125°E (Fig. 6b). The maximum positive correlation coefficient between AVHRR OLR PC1 and PC2 time series is about 0.63 around −12 days (not shown). With this lagged correlation between PC1 and PC2, the convective anomaly center over the Indian Ocean leads that over the western Pacific. The model and NCEP–NCAR reanalysis OLR will be projected to the bandpassed 10-yr AVHRR OLR patterns in the following analysis.

After projection, the close correspondence between the reduced OLR (negative values) and enhanced precipitation (positive values) resulting from the MJO deep convection over the Indian Ocean and vicinity of the western Pacific is clearly seen in the lag-0 regressions of the PC time series onto the bandpassed NCEP–NCAR reanalysis wind, OLR, and CMAP (Figs. 7a,b and 7g,h). The MJO deep convective center with the corresponding enhanced precipitation and reduced OLR (hereafter MDC) is over the region of 5°N–15°S, 70°–100°E in the PC1 regressions (Figs. 7a,b), and moves eastward to 5°N–15°S, 100°–160°E in the PC2 regressions (Figs. 7g,h). In the lower troposphere (850 hPa), the westerly anomalies dominate west of the MDC and reach the east edge of the MDC, and the easterly inflows dominate the east side of the MDC for NCEP–NCAR reanalysis (Figs. 7b,h). In the upper troposphere (200 hPa), the dominant outflow from the east edge of the MDC to its whole west side is composed of easterly anomalies, and the westerly anomaly is the dominant wind in the east side of the MDC (Figs. 7a,g).

The ISUCCM3 has an improved MDC simulation compared to the CTL (Figs. 8 and 9). The MDC is located over 5°N–15°S, 70°E–180° in both the NCEP–NCAR reanalysis and ISUCCM3, but it shifts slightly north in CTL. The maxima of the MDC shift eastward in ISUCCM3 compared to CTL. The amplitudes of OLR and precipitation anomalies in ISUCCM3 are larger than those in CTL and are close to the observations (Figs. 8a,b and 9a,b). The coherent relationship between the low-level equatorial convergent westerly inflow and the MDC is better represented in ISUCCM3 than in CTL (Figs. 8b,h and 9b,h). Both the location and strength of near-equatorial upper-tropospheric divergent flow in ISUCCM3 is closer to the observations than those in CTL (Figs. 8a,g and 9a,g).

The surface heat flux is an important factor for the evolution of MJO through its interaction with deep convection. Figures 7e,k present the observed latent heat flux regressions using PC1 and PC2. The latent heat flux has a positive center over the Indian Ocean around 0°N, 60°–90°E corresponding to the MDC (Fig. 7e). The center moves eastward to the Maritime Continent around 120°E in Fig. 7k. These features are well represented in ISUCCM3 compared to CTL, especially over the Indian Ocean (Figs. 8e and 9e). The latent heat flux centers of both observations and ISUCCM3 are around 60°–90°E over the Indian Ocean for regressions using PC1. However, in CTL, the centers shift westward, around 40°–60°E, close to Africa.

Figures 7f,l show that the observed 500-hPa vertical velocity regressions are consistent with the reduced OLR and enhanced precipitation over the Indian Ocean and western Pacific. The anomalies center of upward motion is around 80°E over the Indian Ocean in regressions obtained using PC1, and it moves eastward to the western Pacific around 125°E with the use of PC2. The maximum of the ISUCCM3 500-hPa upward motion anomalies is larger than the observations, but it is located in the same area as the NCEP–NCAR reanalysis (Figs. 8f,l). The CTL vertical velocity anomalies are comparable to the observations, but the location of the upward motion anomaly center is not as well presented as in ISUCCM3 (Figs. 9f,l). For example, in regressions using PC1, corresponding to the latent heat flux center, the upward anomaly center is around 60°–90°E in both the observations and ISUCCM3, but is mainly around 60°E in CTL, corresponding to its latent heat flux center at 40°–60°E.

Figures 7c,d and 7i,j show 200- and 850-hPa streamfunction regressions for the NCEP–NCAR reanalysis. The low-level leading anticyclone and trailing cyclone responding to the convection are well developed around the equator with the larger amplitude in the Southern Hemisphere, which is consistent with the maxima of the MDC being displaced south of the equator. The pattern of anticyclone and cyclone in the upper troposphere is just opposite to that in the lower troposphere, which indicates a baroclinic structure in the wind fields. The anticyclone and cyclone move eastward from the Indian Ocean (Figs. 7c,d) to the western Pacific (Figs. 7i,j). Comparing with the observations, the forced Rossby wave response and the baroclinic wind response to the MDC are represented in the ISUCCM3 (Figs. 8c,d and 8i,j). Regressions with the 850- and 200-hPa winds in the ISUCCM3 and NCEP–NCAR reanalysis indicate the near-equatorial convergent flow centered on the convective maxima. While the structure of upper- and lower-tropospheric streamfunction is simulated by ISUCCM3, the amplitudes of the anticyclone and cyclone in ISUCCM3 are smaller than those in the NCEP–NCAR reanalysis, but are larger than those in CTL (Figs. 9c,d and 9i,j).

The impacts of three modifications to the convection scheme on the horizontal structure of simulated MJO are illustrated in the above analysis. To examine the contribution of each modification, three pairs of simulations are compared, that is, NOCMT versus ISUCCM3, NOTRI versus NOCMT, and CTL versus NOTRI. The analysis of NOCMT in comparison with ISUCCM3 will illustrate which feature of the MJO is influenced by the CMT. The amplitudes of OLR and precipitation anomalies in NOCMT are comparable to those in ISUCCM3, and the location of MDC in NOCMT is also similar to that in ISUCCM3 (Figs. 10a,b and 10g,h). However, NOCMT produces smaller amplitudes of the low-level leading anticyclone and trailing cyclone responding to the convection compared to ISUCCM3, especially over the Indian Ocean for the regressions using PC1 (Fig. 10d), and the corresponding equatorial convergent inflow is also weaker in NOCMT than that in ISUCCM3. The coherent relationship between the low-level equatorial convergent flow and the MDC is not represented well in NOCMT as compared to ISUCCM3 (Figs. 10b,h). For example, in the lower troposphere for NOCMT, the equatorial convergent westerly inflow stops west of the MDC and does not reach the east edge of the MDC for the regressions using PC1, and the equatorial convergent easterly inflow coupled with the MDC is produced from 140° to 80°E along the equator (Fig. 10b). The comparison between ISUCCM3 and NOCMT shows that the inclusion of CMT in ISUCCM3 enhances the amplitudes of lower-tropospheric anticyclone and cyclone, and results in the more coherent lower-tropospheric equatorial convergent inflow and MDC.

Figures 11a,b and 11g,h show the simulation of MDC in NOTRI. The amplitude of MDC in NOTRI is smaller than that in NOCMT. Corresponding to the weak MDC in NOTRI, the anomalies of 500-hPa vertical velocity, latent heat flux, upper- and lower-tropospheric anticyclone and cyclone, and equatorial divergent and convergent flow are all weaker than those in NOCMT (Figs. 11c,f and 11i,l). The 500-hPa vertical velocity anomalies in NOTRI are generally weaker than those in NOCMT (Figs. 11f,l). The latent heat flux anomalies in NOTRI are also smaller than those in NOCMT around 90° (Figs. 11e and 10e) and 125°E (Figs. 11k and 10k). The weak responses of the anticyclone, cyclone, and corresponding equatorial divergent and convergent flow with respect to the MDC are seen in Figs. 11c,d and 11i,j, especially for the Maritime Continent around 120°E in regressions using PC2. The comparison between NOTR and NOCMT indicates that the convection trigger enhances the MJO signal.

To examine the contribution of revised convection closure, NOTRI is compared to CTL, and the MDC eastward shift is seen in this comparison. In regressions using PC1, the MDC appears between 50° and 75°E over the Indian Ocean for CTL (Figs. 9a,b), but shifts eastward of 20° in NOTRI (Figs. 11a,b). In regressions using PC2, over the western Pacific, the MDC also shifts eastward of 10° in NOTRI compared to CTL (Figs. 11g,h and 9g,h). Corresponding to this eastward relocation of the MDC, the 500-hPa upward anomaly center around 65°E over the Indian Ocean in PC1 regressions of CTL moves 20° eastward in NOTRI (Figs. 9f and 11f). The positive latent heat flux center also moves eastward about 20° from the west of the Indian Ocean in CTL to the central of the Indian Ocean in NOTRI (Figs. 9e and 11e). These eastward relocations of MDC and its related atmospheric fields between NOTRI and CTL are mainly due to the use of revised convection closure.

c. Vertical structure

Applying the similar projection from the analysis of horizontal structure, the longitude–height cross sections of lag-0 regressions of PC1 with space (from 5°N to 5°S)-averaged bandpassed 10-yr NCEP–NCAR reanalysis data are used to describe the observed vertical structures of MJO. The upper-tropospheric divergence and lower-tropospheric convergence anomalies are the dominant features at 90°E in Fig. 12a, which is consistent with convective anomaly center. The westward tilt of divergence anomalies with height around 90°E, coupled with the eastward propagation of the convective anomaly center, suggests the deep convective signal starting from near surface (e.g., Sperber 2003; Sperber et al. 2005). The observed upward velocity anomalies, corresponding to the anomalies of convergence, are around 700–200 hPa near 90°E (Fig. 12b). The low-level easterly anomalies with the positive moisture anomalies on the east side of the convective center (Figs. 12c,d) help build up a moisture convergence anomaly. Also, the moisture anomalies show a westward tilt from 60° to 145°E, with the largest enhancement around 600–700 hPa (Fig. 12d). The low-level convergence, upward velocity, and wet anomalies favor the development of new convection, and the eastward propagation is present on the east side of the MJO convective center, while the low-level divergence, downward velocity, and dry anomalies exist on the west side (Sperber 2003).

In ISUCCM3, dominant upper-tropospheric divergence and lower-tropospheric convergence are present at 90°E (Fig. 13a), which is in agreement with observations, but is not represented well in CTL (Fig. 14a). The westward tilt of divergence with height is also produced by ISUCCM3, while not show in CTL. The upward velocity anomalies around 700–200 hPa and the low-level wetter-than-normal easterly anomalies on the east side of the convective center are better simulated in ISUCCM3 than those in CTL (Figs. 13b,d and 14b,d). ISU CMM3 shows the westward tilt of the specific humidity anomalies around 90°E (Fig. 13d), while CTL does not represent this westward tilt (Fig. 14d). With the inclusion of the revised convection closure, convection trigger, and CMT, ISUCCM3 represents the observed divergence, upward motion, and moisture anomalies coupled with the MJO convection better than CTL, especially for the westward tilt of moisture anomalies. The westward tilt of divergence and moisture coupled with the low-level easterly anomalies shows the preconditions of eastward propagation with the low-level moisture convergence on the east side of the convective center.

To explore the role of each modification in the simulation of MJO, two sensitivity experiments NOCMT and NOTRI are analyzed and compared with ISUCCM3 and CTL. The simulation of divergence, upward motion, zonal wind, and moisture anomalies coupled with the convective center in NOCMT (Fig. 15) is similar to that in ISUCCM3 (Fig. 13), which indicates that the CMT has little influence on the vertical structure of MJO in the Indian Ocean around 90°E. Therefore, the improvement of the vertical structure shown in ISUCCM3 is largely due to the use of the revised convection closure and convection trigger. Figure 16 shows the analysis of MJO vertical structure for NOTRI, which does not includes the new trigger and CMT. In the vicinity of the convective center, the upper-level divergence and low-level convergence anomalies coupled with the upward velocity anomalies in NOTRI are much smaller than those in NOCMT (Figs. 16a,b and 15a,b). Also, on the east side of the convective anomaly center, the low-level wetter-than-normal easterly anomalies in NOTRI are weaker than those in NOCMT (Figs. 16c,d and 15c,d). These results demonstrate that the convection trigger enhances the MJO signal in the vertical. To examine the contribution of revised convection closure to the improvement, Fig. 16 is compared with Fig. 14. A convective anomaly center with the coherent atmospheric variances is present better in NOTRI than those in CTL. The strongest anomalies of upward velocity between 200 and 600 hPa coupled with the divergence and moisture anomalies in NOTRI shift eastward about 15° compared to those in CTL (Figs. 16 and 14), and this eastward shift of the convective anomaly center is consistent with the eastward relocation of MDC in section 3b. The low-level moisture convergence on the east side of the convective center features the low-level wetter-than-normal easterly anomalies in NOTRI in contrast with the drier-than-normal easterly anomalies in CTL (Figs. 16c,d and 14c,d). The comparison between NOTRI and CTL demonstrates that the revised convection closure is largely responsible for the improved simulations of the MJO convective center and the preconditions of eastward propagation in divergence and moisture fields on the east side of the convective center.

To further depict the evolution of physical properties during the life cycle of the MJO with respect to the convective maxima, the lag regressions of the bandpassed 10-yr divergence, vertical velocity, moisture, and wind fields (averaged between 5°N and 5°S) as a function of pressure at 90°E for PC1 regressions after the projection are analyzed for observations and simulations (e.g., Sperber 2003). In the observations, the near-surface convergence anomalies and moistening of the boundary layer appear around −10 day, and the upward motion appears earlier in the low level in advance of the deep convection (Fig. 17). These anomalies develop quickly and reach the upper level around lag-0 days with the deep convection building up. During the life cycle of the MJO with respect to the deep convection center, the wind anomaly field changes from easterly through upward to westerly in the low level (Fig. 17c). With the modified convection scheme, ISUCCM3 simulates the evolution of atmospheric conditions close to the observations compared to CTL (Figs. 18 and 19). The convergence and upward motion anomalies initially occur around −15 days in the low level and penetrate to the upper level, with the deep convection building up in ISUCCM3 (Figs. 18a,b); while in CTL, those anomalies are stationary (Figs. 19a,b). The wind anomaly field of ISUCCM3 shows the evolution from easterly through upward to westerly in the low level within the life cycle of the MJO, which is not present in CTL (Figs. 18c and 19c). Also, ISUCCM3 shows the evolution of moisture during the life cycle of the MJO closer to the observations compared to CTL (Figs. 18d and 19d), although the location of the initial signal for the wetter-than-normal moisture in ISUCCM3 is a little higher than that in the observations. Overall, ISUCCM3 represents the development of atmospheric conditions during the life cycle of the MJO better than CTL, especially for the near-surface precursor signals of divergence and upward motion anomalies in advance of the deep convection.

Figure 20 shows the evolution of the convective anomaly center coupled with the convergence, upward motion, low-level zonal wind anomalies, and moisture anomalies during the MJO life cycle in NOCMT, which does not include the CMT in the scheme. Most features in Fig. 20 are similar to those in Fig. 18, except for the near-surface moisture precursor signal. In advance of the convective maxima, the positive moisture anomalies begin near surface around −20 days in NOCMT (Fig. 20d), but the positive anomalies start around −14 days in ISUCCM3 (Fig. 18d). With both CMT and the convection trigger excluded from the scheme, the lag regressions of NOTRI shown in Fig. 21 are quite different from the NOCMT (Fig. 20). First, the precursor signals of anomalies in advance of the convective center initially appear in the upper level during the MJO life cycle in NOTRI (Fig. 21), while in NOCMT, the precursor signals begin first at the surface (Fig. 20). Second, the amplitude of the MJO-related anomalies in NOTRI is smaller than that in NOCMT. For example, the anomalies of convergence, vertical velocity, and moisture are weaker around lag-0 days with the deep convection building up in NOTRI compared to NOCMT (Figs. 21 and 20). The difference between NOTRI and NOCMT indicates that the impact of the convection trigger not only improves the evolution of atmospheric conditions during the life cycle of the MJO with respect to the convective center, but also enhances the MJO signal. Finally, the comparison between NOTRI (Fig. 21) and CTL (Fig. 19) illustrates the impact of revised convection closure on the MJO simulations. With the inclusion of revised convection closure, NOTRI produces the evolution of the convective anomaly center coupled with the convergence, upward motion, zonal wind, and moisture anomalies during the MJO life cycle (Fig. 21). However, the atmospheric anomalies associated with the development of MJO convective anomaly center are stationary in CTL (Fig. 19).

4. Summary

In this study, the MJO simulated by ISUGCM is evaluated against observations. The results demonstrate that the improved MJO simulations are obtained by the ISU GCM with the inclusion of the revised convection closure assumption, convection trigger condition, and CMT in the convection scheme. The basic features in the simulations include 1) the improved spatial distribution and amplitude of the MJO OLR variance, which represents the MJO deep convective center, 2) a more realistic MJO eastward propagation from the Indian Ocean to the western Pacific with a phase speed of about 5 m s−1 instead of a westward propagation in the control run, and 3) the large variance at eastward intraseasonal time scales, which suggests an enhanced MJO signal. Despite the improvement, the ISU GCM-simulated MJO has more high-frequency variability than the observations. To identify the horizontal and vertical structure of MJO, the ISU GCM simulations are projected to the robust lead–lag relationship of eastward intraseasonal propagating OLR to ensure all simulations are treated identically. The results indicate that the eastward-propagating convection and its related atmospheric variances during the MJO life cycle in the ISU GCM simulation with the modified convection scheme are in general agreement with the observations, especially for the preconditions of eastward propagation on the east side of the convective center and the amplitude of the MJO signal.

The MJO is analyzed in three simulations (NOTRI, NOCMT, and ISUCCM3), which have the revised closure assumption, convection trigger condition, and CMT added one at a time. This enables us to examine the impacts of each modification on the MJO simulation. The use of the revised convection closure assumption results in the eastward-propagating signal while the original convection closure produces the strong westward signal, and this improvement is related to the eastward shift of MJO deep convective center and its associated variances. A more realistic eastward location of the deep convective center favors the eastward-propagating signal from the Indian Ocean to the western Pacific, while the unrealistic location of convective center results in strong westward propagation. With the convection trigger condition added in the scheme, the MJO convection activity is enhanced. The explanation for this strong MJO signal is that moist deep convection occurs less frequently but is more vigorous, and the convection is activated only when the increase of CAPE reaches a certain threshold (70 J kg−1 h−1; Wu et al. 2007a). The inclusion of the CMT in the convection scheme leads to less eastward power at a slightly higher frequency around 30 days, with the decreased eastward phase speed and more coherent structure for the MJO deep convection and dynamic fields. The decreased eastward power in the slightly higher frequency (25–33 days) within the intraseasonal scale is consistent with the reduced kinetic energy corresponding to the decreased phase speed. With the space–time structure analysis, the impact of the CMT is more notable on the horizontal structure than the vertical structure over the Indian Ocean. The possible reason for this difference is that the variance of CMT-induced convective heating is stronger on the horizontal direction than that on the vertical direction over the Indian Ocean. Further analysis is needed to understand the physical processes responsible for the improvements of the MJO simulations resulting from the revised convection closure assumption, convection trigger condition, and CMT in the convection scheme.

Acknowledgments

The first author (Deng) would like to thank her Ph.D. committee members Profs. Mike T. C. Chen, William Gutowski, William Gallus, Raymond Arritt, and Guangjun Zhang. Computing support by Daryl Herzmann is greatly appreciated. Comments and suggestions by two reviewers and the editor help improve the presentation of the paper. This research was partly supported by the Biological and Environmental Research Program (BER), U.S. Department of Energy under Grant DE-FG02-08ER64559, and by the National Science Foundation under Grant ATM-0935263.

REFERENCES

  • Arakawa, A., and W. H. Schubert, 1974: Interaction of a cumulus cloud ensemble with the large-scale environment. Part I. J. Atmos. Sci., 31 , 674701.

    • Search Google Scholar
    • Export Citation
  • Chen, T-C., and M-C. Yen, 1991: A study of the diabiatic heating associated with the Madden–Julian oscillation. J. Geophys. Res., 96 , (D7). 1316313177.

    • Search Google Scholar
    • Export Citation
  • Chen, T-C., and J-M. Chen, 1997: On the relationship between the streamfunction and velocity potential of the Madden–Julian oscillation. J. Atmos. Sci., 54 , 679685.

    • Search Google Scholar
    • Export Citation
  • Duffy, P. B., B. Govindasamy, J. P. Iorio, J. Milovich, K. R. Sperber, K. E. Taylor, M. F. Wehner, and S. L. Thompson, 2003: High-resolution simulations of global climate, part 1: present climate. Climate Dyn., 21 , 371390.

    • Search Google Scholar
    • Export Citation
  • Hack, J. J., 1994: Parameterization of moist convection in the National Center for Atmospheric Research Community Climate Model (CCM2). J. Geophys. Res., 99 , 55515568.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., and M. L. Salby, 1994: Life cycle of the Madden–Julian oscillation. J. Atmos. Sci., 51 , 22252237.

  • Inness, P. M., and J. M. Slingo, 2003: Simulation of the MJO in a coupled GCM. Part I: Comparison with observations and an atmosphere-only GCM. J. Climate, 16 , 345364.

    • Search Google Scholar
    • Export Citation
  • Inness, P. M., J. M. Slingo, E. Guilyardi, and J. Cole, 2003: Simulation of the Madden–Julian oscillation in a coupled general circulation model. Part II: The role of the basic state. J. Climate, 16 , 365382.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77 , 437471.

  • Kiehl, J. T., J. J. Hack, G. B. Bonan, B. A. Boville, D. L. Williamson, and P. J. Rasch, 1998: The National Center for Atmospheric Research Community Climate Model: CCM3. J. Climate, 11 , 11311149.

    • Search Google Scholar
    • Export Citation
  • Kiladis, G. N., K. H. Straub, and P. T. Haertel, 2005: Zonal and vertical structure of the Madden–Julian oscillation. J. Atmos. Sci., 62 , 27902809.

    • Search Google Scholar
    • Export Citation
  • Knutson, R. R., K. M. Weickmann, and J. E. Kutzbach, 1986: Global-scale intraseasonal oscillations of outgoing longwave radiation and 250-mb zonal wind during Northern Hemisphere summer. Mon. Wea. Rev., 114 , 605623.

    • Search Google Scholar
    • Export Citation
  • Krishnamurti, T., and D. Subrahmanyam, 1982: The 30–50 day mode at 850 mb during MONEX. J. Atmos. Sci., 39 , 20882095.

  • Lanczos, C., 1956: Applied Analysis. Prentice-Hall, 539 pp.

  • Lau, K-M., and D. E. Waliser, Eds.,. 2005: Intraseasonal Variability of the Atmosphere-Ocean Climate System. Springer, 436 pp.

  • Liebmann, B., and C. A. Smith, 1996: Description of a complete (interpolated) OLR dataset. Bull. Amer. Meteor. Soc., 77 , 12751277.

  • Lin, J-L., and Coauthors, 2006: Tropical intraseasonal variability in 14 IPCC AR4 climate models. Part I: Convective signals. J. Climate, 19 , 26652690.

    • Search Google Scholar
    • Export Citation
  • Liu, P., B. Wang, K. R. Sperber, T. Li, and G. A. Meehl, 2005: MJO in the NCAR CAM2 with the Tiedtke convective scheme. J. Climate, 18 , 30073020.

    • Search Google Scholar
    • Export Citation
  • Madden, R. A., and P. R. Julian, 1971: Detection of a 40–50 day oscillation in the zonal wind in the tropical Pacific. J. Atmos. Sci., 28 , 702708.

    • Search Google Scholar
    • Export Citation
  • Madden, R. A., and P. R. Julian, 1972: Description of global-scale circulation cells in the tropics with a 40–50 day period. J. Atmos. Sci., 29 , 11091123.

    • Search Google Scholar
    • Export Citation
  • Madden, R. A., and P. R. Julian, 1994: Observations of the 40-50-day tropical oscillation—A review. Mon. Wea. Rev., 122 , 814837.

  • Maloney, E. D., and D. L. Hartmann, 2001: The sensitivity of intraseasonal variability in the NCAR CCM3 to changes in convective parameterization. J. Climate, 14 , 20152034.

    • Search Google Scholar
    • Export Citation
  • Nakazawa, T., 1988: Tropical super clusters within intraseasonal variations over the western Pacific. J. Meteor. Soc. Japan, 66 , 823839.

    • Search Google Scholar
    • Export Citation
  • Rui, H., and B. Wang, 1990: Development characteristics and dynamic structure of tropical intraseasonal convection anomalies. J. Atmos. Sci., 47 , 357379.

    • Search Google Scholar
    • Export Citation
  • Salby, M. L., and H. H. Hendon, 1994: Intraseasonal behavior of clouds, temperature, and motion in the tropics. J. Atmos. Sci., 51 , 22072224.

    • Search Google Scholar
    • Export Citation
  • Slingo, J. M., and Coauthors, 1996: The intraseasonal oscillation in 15 atmospheric general circulation models: Results from an AMIP diagnostic subproject. Climate Dyn., 12 , 325357.

    • Search Google Scholar
    • Export Citation
  • Song, X., X. Wu, G. J. Zhang, and R. W. Arritt, 2008a: Dynamical effects of convective momentum transports on global climate simulations. J. Climate, 21 , 180194.

    • Search Google Scholar
    • Export Citation
  • Song, X., X. Wu, G. J. Zhang, and R. W. Arritt, 2008b: Understanding the effects of convective momentum transport on climate simulations: The role of convective heating. J. Climate, 21 , 50345047.

    • Search Google Scholar
    • Export Citation
  • Sperber, K. R., 2003: Propagation and the vertical structure of the Madden–Julian oscillation. Mon. Wea. Rev., 131 , 30183037.

  • Sperber, K. R., 2004: Madden–Julian variability in NCAR CAM2.0 and CCSM2.0. Climate Dyn., 23 , (3–4). 259278.

  • Sperber, K. R., S. Gualdi, S. Legutke, and V. Gayler, 2005: The Madden–Julian oscillation in ECHAM4 coupled and uncoupled GCMs. Climate Dyn., 25 , 117140.

    • Search Google Scholar
    • Export Citation
  • Sud, Y. C., and G. K. Walker, 1999: Microphysics of Clouds with the Relaxed Arakawa–Schubert Scheme (McRAS). Part I: Design and evaluation with GATE phase III data. J. Atmos. Sci., 56 , 31963220.

    • Search Google Scholar
    • Export Citation
  • Tiedtke, M., 1989: A comprehensive mass flux scheme for cumulus parameterization in large-scale models. Mon. Wea. Rev., 117 , 17791800.

    • Search Google Scholar
    • Export Citation
  • Tokioka, T., K. Yamazaki, A. Kitoh, and T. Ose, 1988: The equatorial 30-60 day oscillation and the Arakawa-Schubert penetrative cumulus parameterization. J. Meteor. Soc. Japan, 66 , 883901.

    • Search Google Scholar
    • Export Citation
  • Tung, W. W., and M. Yanai, 2002: Convective momentum transport observed during the TOGA COARE IOP. Part II: Case studies. J. Atmos. Sci., 59 , 25352549.

    • Search Google Scholar
    • Export Citation
  • Waliser, D. E., K. M. Lau, and J. H. Kim, 1999: The influence of coupled sea surface temperatures on the Madden–Julian oscillation: A model perturbation experiment. J. Atmos. Sci., 56 , 333358.

    • Search Google Scholar
    • Export Citation
  • Wang, W., and M. E. Schlesinger, 1999: The dependence on convection parameterization of the tropical intraseasonal oscillation simulated by the UIUC 11-layer atmospheric GCM. J. Climate, 12 , 14231457.

    • Search Google Scholar
    • Export Citation
  • Weickmann, K. M., 1983: Intraseasonal circulation and outgoing longwave radiation modes during Northern Hemisphere winter. Mon. Wea. Rev., 111 , 18381858.

    • Search Google Scholar
    • Export Citation
  • Weickmann, K. M., G. R. Lussky, and J. E. Kutzbach, 1985: Intraseasonal (30–60 day) fluctuations of outgoing longwave radiation and 250 mb streamfunction during northern winter. Mon. Wea. Rev., 113 , 941961.

    • Search Google Scholar
    • Export Citation
  • Wu, X., and M. Yanai, 1994: Effects of vertical wind shear on the cumulus transport of momentum: Observations and parameterization. J. Atmos. Sci., 51 , 16401660.

    • Search Google Scholar
    • Export Citation
  • Wu, X., and X-Z. Liang, 2005a: Radiative effects of cloud horizontal inhomogeneity and vertical overlap identified from a month-long cloud-resolving simulation. J. Atmos. Sci., 62 , 41054112.

    • Search Google Scholar
    • Export Citation
  • Wu, X., and X-Z. Liang, 2005b: Effect of subgrid cloud-radiation interaction on climate simulations. Geophys. Res. Lett., 32 , L24806. doi:10.1029/2005GL024432.

    • Search Google Scholar
    • Export Citation
  • Wu, X., X-Z. Liang, and G. J. Zhang, 2003: Seasonal migration of ITCZ precipitation across the equator: Why can’t GCMs simulate it? Geophys. Res. Lett., 30 , 1824. doi:10.1029/2003GL017198.

    • Search Google Scholar
    • Export Citation
  • Wu, X., L. Deng, X. Song, G. Vettoretti, W. R. Peltier, and G. J. Zhang, 2007a: Impact of a modified convective scheme on the Madden-Julian Oscillation and El Niño–Southern Oscillation in a coupled climate model. Geophys. Res. Lett., 34 , L16823. doi:10.1029/2007GL030637.

    • Search Google Scholar
    • Export Citation
  • Wu, X., L. Deng, X. Song, and G. J. Zhang, 2007b: Coupling of convective momentum transport with convective heating in global climate simulations. J. Atmos. Sci., 64 , 13341349.

    • Search Google Scholar
    • Export Citation
  • Xie, P., and P. A. Arkin, 1996: Analyses of global monthly precipitation using gauge observations, satellite estimates, and numerical model predictions. J. Climate, 9 , 840858.

    • Search Google Scholar
    • Export Citation
  • Xie, P., and P. A. Arkin, 1997: Global precipitation 17-year monthly analysis based on gauge observations, satellite estimates, and numerical model outputs. Bull. Amer. Meteor. Soc., 78 , 25392558.

    • Search Google Scholar
    • Export Citation
  • Yanai, M., B. Chen, and W-W. Tung, 2000: The Madden–Julian oscillation observed during the TOGA COARE IOP: Global view. J. Atmos. Sci., 57 , 23742396.

    • Search Google Scholar
    • Export Citation
  • Zangvil, A., 1975: Temporal and spatial behavior of large-scale disturbances in tropical cloudiness deduced from satellite brightness data. Mon. Wea. Rev., 103 , 904920.

    • Search Google Scholar
    • Export Citation
  • Zangvil, A., and M. Yanai, 1981: Upper tropospheric waves in the tropics. Part II: Association with clouds in the wavenumber-frequency domain. J. Atmos. Sci., 38 , 939953.

    • Search Google Scholar
    • Export Citation
  • Zhang, C., 2005: Madden-Julian oscillation. Rev. Geophys., 43 , RG2003. doi:10.1029/2004RG000158.

  • Zhang, C., and M. Dong, 2004: Seasonality in the Madden–Julian oscillation. J. Climate, 17 , 31693180.

  • Zhang, G. J., 2002: Convective quasi-equilibrium in midlatitude continental environment and its effect on convective parameterization. J. Geophys. Res., 107 , 4220. doi:10.1029/2001JD001005.

    • Search Google Scholar
    • Export Citation
  • Zhang, G. J., and H. R. Cho, 1991: Parameterization of the vertical transport of momentum by cumulus clouds. Part I: Theory. J. Atmos. Sci., 48 , 14831492.

    • Search Google Scholar
    • Export Citation
  • Zhang, G. J., and N. A. McFarlane, 1995: Sensitivity of climate simulations to the parameterization of cumulus convection in the Canadian Climate Centre general circulation model. Atmos.–Ocean, 33 , 407446.

    • Search Google Scholar
    • Export Citation
  • Zhang, G. J., and X. Wu, 2003: Convective momentum transport and perturbation pressure field from a cloud-resolving model simulation. J. Atmos. Sci., 60 , 11201139.

    • Search Google Scholar
    • Export Citation
  • Zhang, G. J., and M. Mu, 2005: Simulation of the Madden–Julian oscillation in the NCAR CCM3 using a revised Zhang–McFarlane convection parameterization scheme. J. Climate, 18 , 40464064.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

October–April climatology of precipitation rate (mm day−1) for (a) CMAP, (b) ISUCCM3, and (c) CTL.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 2.
Fig. 2.

1987 Hovmöller diagrams of 850-hPa zonal wind anomaly (m s−1) averaged between 5°N and 5°S for (a) NCEP–NCAR reanalysis, (b) ISUCCM3, and (c) CTL. Areas of negative (easterly anomaly) are shaded. The contour interval is 5 m s−1.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 3.
Fig. 3.

Spatial distribution of variance of 20–70-day bandpassed OLR for the four seasons from 1979 to 1988: (a) AVHRR, (b) ISUCCM3, (c) NOCMT, (d) NOTRI, and (e) CTL. (top left) December–February (DJF), (top right) March–May (MAM), (bottom left) June–August (JJA), and (bottom right) September–November (SON). Unit: W2 m−4. Regions of deep convection are highlighted (shaded > 240 W2 m−4). The contour interval is 80 W2 m−4.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 4.
Fig. 4.

Ten-years (October–April 1979–88) lag correlations of 20–100-day bandpassed OLR with 200-hPa velocity potential for (a) NCEP–NCAR reanalysis, (b) ISUCCM3, (c) NOCMT, (d) NOTRI, and (e) CTL. The white solid line represents a phase speed of ∼5 m s−1.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 5.
Fig. 5.

Wavenumber–frequency spectra of 200-hPa zonal wind averaged between 10°N and 10°S for (a) NCEP–NCAR reanalysis, (b) ISUCCM3, (c) NOCMT, (d) NOTRI, and (e) CTL (years 1979–88). The contour starts from 0.05 m2 s−2, with an interval of 0.05 m2 s−2.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 6.
Fig. 6.

Analysis of 20–100-day bandpassed AVHRR OLR for 10 yr (October–April 1979–88): (a) EOF1 and (b) EOF2.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 7.
Fig. 7.

Zero-lag linear regressions of PC1 for 20–100-day filtered 10-yr (October–April 1979–88) (a) NCEP–NCAR reanalysis 200-hPa wind (m s−1) and OLR (W m−2); (b) NCEP–NCAR reanalysis 850-hPa wind (m s−1) and CMAP (mm day−1), (c) NCEP–NCAR reanalysis 200-hPa wind (m s−1) and streamfunction (m2 s−1), (d) NCEP–NCAR reanalysis 850-hPa wind (m s−1) and streamfunction (m2 s−1), (e) NCEP–NCAR reanalysis latent heat flux (W m−2), and (f) NCEP–NCAR reanalysis 500-hPa vertical velocity (Pa s−1; negative sign means upward motion). (g)–(l) Same as (a)–(f), but for regressions using PC2. All regressions have been scaled by a one standard deviation of PC1 to give units. The variables are plotted where the regression is 95% confidence or better.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 8.
Fig. 8.

Same as Fig. 7, but for ISUCCM3.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 9.
Fig. 9.

Same as Fig. 7, but for CTL.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 10.
Fig. 10.

Same as Fig. 7, but for NOCMT.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 11.
Fig. 11.

Same as Fig. 7, but for NOTRI.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 12.
Fig. 12.

NCEP–NCAR reanalysis longitude–height cross sections of zero-lag linear regressions of PC1 with 5°N–5°S-averaged 20–100-day bandpass-filtered (a) divergence (s−1); (b) vertical velocity (Pa s−1); (c) zonal wind (m s−1) and vertical velocity (Pa s−1) vectors (vertical velocity times −100) and contours of zonal wind with the interval of 0.5 m s−1; and (d) specific humidity (g kg−1). All regressions have been scaled by a one standard deviation of PC1 to give units.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 13.
Fig. 13.

Same as Fig. 12, but for ISUCCM3.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 14.
Fig. 14.

Same as Fig. 12, but for CTL.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 15.
Fig. 15.

Same as Fig. 12, but for NOCMT.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 16.
Fig. 16.

Same as Fig. 12, but for NOTRI.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 17.
Fig. 17.

Time lag vs height plots of linear regressions of PC1 with 90°E (5°N–5°S averaged) 20–100-day bandpass-filtered NCEP–NCAR reanalysis (a) divergence (s−1); (b) vertical velocity (Pa s−1; negative sign means upward motion); (c) zonal wind (m s−1) and vertical velocity (Pa s−1) vectors (vertical velocity times −100) and contours of zonal wind with the interval of 0.5 m s−1; and (d) specific humidity (g kg−1). All regressions have been scaled by a one standard deviation of PC1 to give units. Time lags run from −25 to 25 days.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 18.
Fig. 18.

Same as Fig. 17, but for ISUCCM3.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 19.
Fig. 19.

Same as Fig. 17, but for CTL.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 20.
Fig. 20.

Same as Fig. 17, but for NOCMT.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Fig. 21.
Fig. 21.

Same as Fig. 17, but for NOTRI.

Citation: Journal of Climate 23, 2; 10.1175/2009JCLI3114.1

Table 1.

List of four model simulations.

Table 1.
Save