• Bellenger, H., Y. N. Takayabu, T. Ushiyama, and K. Yoneyama, 2010: Role of diurnal warm layers in the diurnal cycle of convection over the tropical Indian Ocean during MISMO. Mon. Wea. Rev., 138, 24262433, doi:10.1175/2010MWR3249.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Benedict, J., and D. A. Randall, 2007: Observed characteristics of the MJO relative to maximum rainfall. J. Atmos. Sci., 64, 23322354, doi:10.1175/JAS3968.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bernie, D. J., S. J. Woolnough, and J. M. Slingo, 2005: Modeling diurnal and intraseasonal variability of the ocean mixed layer. J. Climate, 18, 11901202, doi:10.1175/JCLI3319.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bernie, D. J., E. Guilyardi, G. Madec, J. M. Slingo, and S. J. Woolnough, 2007: Impact of resolving the diurnal cycle in an ocean–atmosphere GCM. Part 1: A diurnally forced OGCM. Climate Dyn., 29, 575590, doi:10.1007/s00382-007-0249-6.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bernie, D. J., E. Guilyardi, G. Madec, J. M. Slingo, S. J. Woolnough, and J. Cole, 2008: Impact of resolving the diurnal cycle in an ocean–atmosphere GCM. Part 2: A diurnally coupled CGCM. Climate Dyn., 31, 909925, doi:10.1007/s00382-008-0429-z.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bladé, I., and D. L. Hartmann, 1993: Tropical intraseasonal oscillation in a simple nonlinear model. J. Atmos. Sci., 50, 29222939, doi:10.1175/1520-0469(1993)050<2922:TIOIAS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Duchon C. E., 1979: Lanczos filtering in one and two dimensions. J. Appl. Meteor., 18, 10161022, doi:10.1175/1520-0450(1979)018<1016:LFIOAT>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Edson, J. B., and Coauthors, 2013: On the exchange of momentum over the open ocean. J. Phys. Oceanogr., 43, 15891610, doi:10.1175/JPO-D-12-0173.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fairall, C. W., E. F. Bradley, J. E. Hare, A. A. Grachev, and J. B. Edson, 2003: Bulk parameterization of air–sea fluxes: Updates and verification for the COARE algorithm. J. Climate, 16, 571590, doi:10.1175/1520-0442(2003)016<0571:BPOASF>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Flatau, M., P. J. Flatau, P. Phoebus, and P. P. Niiler, 1997: The feedback between equatorial convection and local radiative and evaporative processes: The implication for intraseasonal oscillations. J. Atmos. Sci., 54, 23732386, doi:10.1175/1520-0469(1997)054<2373:TFBECA>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fu, X., B. Wang, D. E. Waliser, and L. Tao, 2007: Impact of atmosphere–ocean coupling on the predictability of monsoon intraseasonal oscillations. J. Atmos. Sci., 64, 157174, doi:10.1175/JAS3830.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fu, X., W. Wang, J.-Y. Lee, B. Wang, K. Kikuchi, J. Xu, J. Li, and S. Weaver, 2015: Distinctive roles of air–sea coupling on different MJO events: A new perspective revealed from the DYNAMO/CINDY field campaign. Mon. Wea. Rev., 143, 794812, doi:10.1175/MWR-D-14-00221.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gent, P. R., and J. C. McWilliams, 1990: Isopycnal mixing in ocean circulation models. J. Phys. Oceanogr., 20, 150155, doi:10.1175/1520-0485(1990)020<0150:IMIOCM>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Griffies, S. M., 2012: Elements of the Modular Ocean Model (MOM): 2012 release. NOAA/GFDL Ocean Group Tech. Rep. 7, 631 pp. [Available online at http://www.mom-ocean.org/web/docs/project/MOM5_elements.pdf.]

  • Griffies, S. M., and R. W. Hallberg, 2000: Biharmonic friction with a Smagorinsky viscosity for use in large-scale eddy-permitting ocean models. Mon. Wea. Rev., 128, 29352946, doi:10.1175/1520-0493(2000)128<2935:BFWASL>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Haertel, P. T., G. N. Kiladis, A. Denno, and T. M. Rickenbach, 2008: Vertical-mode decompositions of 2-day waves and the Madden–Julian oscillation. J. Atmos. Sci., 65, 813833, doi:10.1175/2007JAS2314.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Huang, B., M. L’Heureux, J. Lawrimore, C. Liu, H.-M. Zhang, V. Banzon, Z.-Z. Hu, and A. Kumar, 2013: Why did large differences arise in the sea surface temperature datasets across the tropical Pacific during 2012? J. Atmos. Oceanic Technol., 30, 29442953, doi:10.1175/JTECH-D-13-00034.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Inness, P. M., and J. M. Slingo, 2003: Simulation of the Madden–Julian Oscillation in a coupled general circulation model. Part I: Comparison with observations and an atmosphere-only GCM. J. Climate, 16, 345364, doi:10.1175/1520-0442(2003)016<0345:SOTMJO>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kemball-Cook, S., B. Wang, and X. Fu, 2002: Simulation of the intraseasonal oscillation in the ECHAM-4 model: Impact of coupling with an ocean model. J. Atmos. Sci., 59, 14331453, doi:10.1175/1520-0469(2002)059<1433:SOTIOI>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kikuchi, K., and Y. N. Takayabu, 2004: The development of organized convection associated with the MJO during TOGA COARE IOP: Trimodal characteristics. Geophys. Res. Lett., 31, L10101, doi:10.1029/2004GL019601.

    • Crossref
    • 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, doi:10.1175/JAS3520.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Large, W. G., J. C. McWilliams, and S. C. Doney, 1994: Oceanic vertical mixing: Review and a model with a nonlocal boundary layer parameterization. Rev. Geophys., 32, 363403, doi:10.1029/94RG01872.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, Y., W. Han, T. Shinoda, C. Wang, R.-C. Lien, J. N. Moum, and J. W. Wang, 2013: Effects of solar radiation diurnal cycle on the tropical Indian Ocean mixed layer variability during wintertime Madden-Julian oscillation events. J. Geophys. Res. Oceans, 118, 49454964, doi:10.1002/jgrc.20395.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lin, J.-L., and Coauthors, 2006: Tropical intraseasonal variability in 14 IPCC AR4 climate models. Part I: Convective signals. J. Climate, 19, 26652690, doi:10.1175/JCLI3735.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Loeb, N. G., K. J. Priestley, D. P. Kratz, E. B. Geier, R. N. Green, B. A. Wielicki, P. O’Rawe Hinton, and S. K. Nolan, 2001: Determination of unfiltered radiances from the Clouds and the Earth’s Radiant Energy System instrument. J. Appl. Meteor., 40, 822835, doi:10.1175/1520-0450(2001)040<0822:DOURFT>2.0.CO;2.

    • Crossref
    • 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, doi:10.1175/1520-0469(1972)029<1109:DOGSCC>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • McPhaden, M. J., and Coauthors, 2009: RAMA: The Research Moored Array for African–Asian–Australian Monsoon Analysis and Prediction. Bull. Amer. Meteor. Soc., 90, 459480, doi:10.1175/2008BAMS2608.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Morel, A., and D. Antoine, 1994: Heating rate within the upper ocean in relation to its bio-optical state. J. Phys. Oceanogr., 24, 16521665, doi:10.1175/1520-0485(1994)024<1652:HRWTUO>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Moum, J. N., and Coauthors, 2014: Air–sea interactions from westerly wind bursts during the November 2011 MJO in the Indian Ocean. Bull. Amer. Meteor. Soc., 95, 11851199, doi:10.1175/BAMS-D-12-00225.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Reynolds, R. W., T. M. Smith, C. Liu, D. B. Chelton, K. S. Casey, and M. G. Schlax, 2007: Daily high-resolution-blended analyses for sea surface temperature. J. Climate, 20, 54735496, doi:10.1175/2007JCLI1824.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rienecker, M. M., and Coauthors, 2011: MERRA: NASA’s Modern-Era Retrospective Analysis for Research and Applications. J. Climate, 24, 36243648, doi:10.1175/JCLI-D-11-00015.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Saha, S., and Coauthors, 2010: The NCEP Climate Forecast System Reanalysis. Bull. Amer. Meteor. Soc., 91, 10151067, doi:10.1175/2010BAMS3001.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Seo, H., A. C. Subramanian, A. J. Miller, and N. R. Cavanaugh, 2014: Coupled impacts of the diurnal cycle of sea surface temperature on the Madden–Julian oscillation. J. Climate, 27, 84228443, doi:10.1175/JCLI-D-14-00141.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shinoda, T., 2005: Impact of the diurnal cycle of solar radiation on intraseasonal SST variability in the western equatorial Pacific. J. Climate, 18, 26282636, doi:10.1175/JCLI3432.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shinoda, T., and H. H. Hendon, 1998: Mixed layer modeling of intraseasonal variability in the tropical western Pacific and Indian Oceans. J. Climate, 11, 26682685, doi:10.1175/1520-0442(1998)011<2668:MLMOIV>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shinoda, T., and H. H. Hendon, 2001: Upper-ocean heat budget in response to the Madden–Julian oscillation in the western equatorial Pacific. J. Climate, 14, 41474165, doi:10.1175/1520-0442(2001)014<4147:UOHBIR>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shinoda, T., H. H. Hendon, and J. Glick, 1998: Intraseasonal variability of surface fluxes and sea surface temperature in the tropical western Pacific and Indian Oceans. J. Climate, 11, 16851702, doi:10.1175/1520-0442(1998)011<1685:IVOSFA>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Waliser, D. E., K. E. 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, doi:10.1175/1520-0469(1999)056<0333:TIOCSS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, W., M. Chen, and A. Kumar, 2009: Impacts of ocean surface on the northward propagation of the boreal summer intraseasonal oscillation in the NCEP Climate Forecast System. J. Climate, 22, 65616576, doi:10.1175/2009JCLI3007.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, W., A. Kumar, and X. Fu, 2015: How important is the sea surface temperature uncertainty in the prediction of tropical convection associated with the MJO? Mon. Wea. Rev., 143, 31563175, doi:10.1175/MWR-D-14-00385.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Webster, P. J., C. A. Clayson, and J. A. Curry, 1996: Clouds, radiation, and the diurnal cycle of sea surface temperature in the tropical western Pacific Ocean. J. Climate, 9, 17121730, doi:10.1175/1520-0442(1996)009<1712:CRATDC>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Weller, R. A., and S. P. Anderson, 1996: Surface meteorology and air–sea fluxes in the western equatorial Pacific warm pool during the TOGA Coupled Ocean–Atmosphere Response Experiment. J. Climate, 9, 19591990, doi:10.1175/1520-0442(1996)009<1959:SMAASF>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wentz, F. J., C. L. Gentemann, D. K. Smith, and D. B. Chelton, 2000: Satellite measurements of sea surface temperature through clouds. Science, 288, 847850, doi:10.1126/science.288.5467.847.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Woolnough, S. J., F. Vitart, and M. A. Balmaseda, 2007: The role of the ocean in the Madden-Julian oscillation: Implications for MJO prediction. Quart. J. Roy. Meteor. Soc., 133, 117128, doi:10.1002/qj.4.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Yoneyama, K., C. Zhang, and C. N. Long, 2013: Tracking pulses of the Madden–Julian oscillation. Bull. Amer. Meteor. Soc., 94, 18711891, doi:10.1175/BAMS-D-12-00157.1.

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

  • Zhang, C., 2013: Madden–Julian oscillation: Bridging weather and climate. Bull. Amer. Meteor. Soc., 94, 18491870, doi:10.1175/BAMS-D-12-00026.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, C., and S. P. Anderson, 2003: Sensitivity of intraseasonal perturbations in SST to the structure of the MJO. J. Atmos. Sci., 60, 21962206, doi:10.1175/1520-0469(2003)060<2196:SOIPIS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • View in gallery

    Diurnal cycle of ocean temperature (°C) averaged during the intensive observing period in the DYNAMO field campaign (1 Oct 2011–15 Jan 2012). Observations from DYNAMO are at 0°, 79°E and 1.5°S, 79°E, and those from RAMA are at other locations. Contours are plotted at 0.1°C interval. The x axis is the local time. (right) The thick solid and dashed curves for DYNAMO locations are mixed layer depth and isothermal depth, respectively.

  • View in gallery

    Comparisons between hourly in situ temperature (°C) at the depth of 1 m (red) and 5 m (black), and their differences (blue; right y axis) during the DYNAMO period.

  • View in gallery

    Composited vertical daily mean temperature profiles for two groups at (a) 0°, 79°E and (b) 1.5°S, 79°E. The STRONG group consists of cases with the amplitude of DT1m5m greater than one STD of DT1m5m. The composite for the ALL group uses all observations.

  • View in gallery

    Differences in downward solar radiation flux (W m−2) between reanalyses and CERES estimate averaged from September 2011 to January 2012: (a) CFSR minus CERES, (b) MERRA minus CERES, and (c) hybrid minus CERES.

  • View in gallery

    Comparisons at the DYNAMO observation location 1.5°S, 79°E. (a) Monthly mean downward shortwave radiation flux (SW). (b) Mean SW diurnal cycle average from October 2011 to January 2012. (c) Intraseasonal SW anomalies and SST. (d) As in (c), but for W10m. The hybrid reanalysis in each panel is taken as the average of CFSR and MERRA reanalyses. For SW and W10m, the DYNAMO observations and CFSR, MERRA, and hybrid reanalyses are plotted with dotted, long dashed, short dashed, and thin solid curves, respectively. The CERES SW in (a) and DYNAMO SST in (c) and (d) are plotted with thick solid curves.

  • View in gallery

    Evolution of the daily mean SST observation and 1M simulation with MOM5 at 1.5°S, 79°E.

  • View in gallery

    The October 2011–January 2012 mean SST (°C): (a) 1M simulation, (b) 10M simulation, (c) TMI, (d) differences between 1M simulation and TMI, (e) differences between 10M simulation and TMI, and (f) differences between the two simulations.

  • View in gallery

    Evolution of hourly SSTs at 1.5°S, 79°E in simulations and observations during the DYNAMO period: (a) 24 Sep–12 Dec 2011 and (b) 4–17 Nov 2011. Black, blue, and red curves are for the observation, 1M simulation, and 10M simulation, respectively.

  • View in gallery

    Diurnal variation of vertical temperature profile in the upper ocean at 1.5°S, 79°E in the (a)–(c) 1M simulation and (d)–(f) 10M simulation. (left) The average for September 2011–January 2012 and for (center) warming and (right) cooling phases of MJO events during this period. The thick solid and dashed curves are mixed layer depth and isothermal depth. Horizontal dotted lines indicate the bottom of model layers.

  • View in gallery

    Horizontal distributions of mean temperature diurnal cycle range (°C). (a) SST in the 1M simulation, (b) SST in the 10M simulation, and (c) SST differences between the two simulations and temperature at 15-m depth in the (d) 1M simulation, (e) 10M simulation, and (f) the differences between the two simulations.

  • View in gallery

    Horizontal distributions of mean intraseasonal STD (°C) in the (a) 1M simulation, (b) 10M simulation, and (c) differences between the 1M and 10M simulations. (d) As in (c), but the differences are between two additional runs with daily mean forcing instead of hourly forcing as in (a) and (b).

  • View in gallery

    Time–longitude diagram of the simulated 20–80-day bandpass-filtered SST averaged between 5°S and 5°N: (a) 1M simulation, (b) 10M simulation, and (c) their differences.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 36 36 23
PDF Downloads 30 30 19

Importance of the Vertical Resolution in Simulating SST Diurnal and Intraseasonal Variability in an Oceanic General Circulation Model

View More View Less
  • 1 Earth System Science Interdisciplinary Center, University of Maryland, and Climate Prediction Center, NOAA/NWS/NCEP, College Park, Maryland, and Nanjing University of Information Science and Technology, Nanjing, China
  • 2 Climate Prediction Center, NOAA/NWS/NCEP, College Park, Maryland
  • 3 Earth System Science Interdisciplinary Center, University of Maryland, and Climate Prediction Center, NOAA/NWS/NCEP, College Park, Maryland
© Get Permissions
Full access

Abstract

In this paper, the influence of high vertical resolution near the surface in an oceanic general circulation model in simulating the observed sea surface temperature (SST) variability is investigated. In situ observations of vertical temperature profiles are first used to quantify temperature variability with depth near the ocean surface. The analysis shows that there is a sharp vertical temperature gradient within the top 10 m of the ocean. Both diurnal and intraseasonal variabilities of the ocean temperatures are largest near the surface and decrease with the ocean depth. Numerical experiments with an oceanic general circulation model are next carried out with 1- and 10-m vertical resolutions for the upper ocean to study the dependence of the simulated SST and vertical temperature structure on the vertical resolution. It is found that the simulated diurnal and intraseasonal variabilities, as well as the associated vertical temperature gradient near the surface, are strongly influenced by the oceanic vertical resolution, with the 1-m vertical resolution producing a stronger vertical temperature gradient and temporal variability than the 10-m vertical resolution. These results suggest that a realistic representation of SST variability with a high vertical resolution in the upper ocean is required for a coupled atmosphere–ocean model to correctly simulate the observed tropical intraseasonal oscillations, including the Madden–Julian oscillation and the boreal summer monsoon intraseasonal oscillation, which are strongly linked with the underlying SST.

© 2017 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author e-mail: Wanqiu Wang, wanqiu.wang@noaa.gov

Abstract

In this paper, the influence of high vertical resolution near the surface in an oceanic general circulation model in simulating the observed sea surface temperature (SST) variability is investigated. In situ observations of vertical temperature profiles are first used to quantify temperature variability with depth near the ocean surface. The analysis shows that there is a sharp vertical temperature gradient within the top 10 m of the ocean. Both diurnal and intraseasonal variabilities of the ocean temperatures are largest near the surface and decrease with the ocean depth. Numerical experiments with an oceanic general circulation model are next carried out with 1- and 10-m vertical resolutions for the upper ocean to study the dependence of the simulated SST and vertical temperature structure on the vertical resolution. It is found that the simulated diurnal and intraseasonal variabilities, as well as the associated vertical temperature gradient near the surface, are strongly influenced by the oceanic vertical resolution, with the 1-m vertical resolution producing a stronger vertical temperature gradient and temporal variability than the 10-m vertical resolution. These results suggest that a realistic representation of SST variability with a high vertical resolution in the upper ocean is required for a coupled atmosphere–ocean model to correctly simulate the observed tropical intraseasonal oscillations, including the Madden–Julian oscillation and the boreal summer monsoon intraseasonal oscillation, which are strongly linked with the underlying SST.

© 2017 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author e-mail: Wanqiu Wang, wanqiu.wang@noaa.gov

1. Introduction

The Madden–Julian oscillation (Madden and Julian 1972) influences the tropical and extratropical weather and climate variability and is one of the important sources of predictability for extended range forecasts. Accurate simulations and predictions of the MJO, therefore, are of great importance for operational weather forecast and climate modeling. Upper-ocean processes and the air–sea coupling have been shown to have a strong influence on the MJO simulation (e.g., Flatau et al. 1997; Shinoda et al. 1998; Waliser et al. 1999; Kemball-Cook et al. 2002; Zhang 2005; Fu et al. 2007; Woolnough et al. 2007; Seo et al. 2014). In particular, predictions of the MJO strongly depend on the amplitude of the underlying sea surface temperatures (SSTs) (Woolnough et al. 2007; Wang et al. 2015), suggesting that reproducing the observed MJO requires not only the use of coupled dynamical atmosphere–ocean models but also the realistic simulation of SST. The SST has also been shown to have strong impacts on the simulation of the boreal summer intraseasonal oscillation (BSISO) in the Indian monsoon region (Wang et al. 2009).

In the tropics, several factors contribute to SST variability including the surface momentum and heat fluxes through their influence on dynamic and thermodynamic processes modulating the mixed layer heat content, stratification, and vertical mixing (Shinoda and Hendon 2001; Zhang and Anderson 2003; Shinoda and Hendon 1998; Shinoda 2005; Bernie et al. 2007; Li et al. 2013). While it has been shown that simulation and prediction of the tropical intraseasonal oscillation strongly depend on the accuracy of the underlying SSTs, the amplitude of associated SST variability has been a difficult aspect to quantify. For example, large differences in SST variability are found among observational analyses and the coupled atmosphere–ocean model simulations (Wang et al. 2015). The differences among observational SST products may be related to analysis procedure (Huang et al. 2013) and also to how the SST is defined (e.g., ocean skin temperature versus bulk temperature; Wentz et al. 2000; Reynolds et al. 2007). For model simulations, a major factor that influences the SST simulations is the vertical oceanic resolution near the surface, which is typically 10 m in the current generation of coupled atmosphere–ocean general circulation models (CGCMs). Inness and Slingo (2003) and Bernie et al. (2005) emphasized the influence of vertical resolution on the simulation of SST and pointed out that a multilevel ocean mixed layer model was necessary to resolve the diurnal cycle of SST.

Diurnal fluctuations in the SST and stratification of the upper ocean also influence SST variability at longer time scales. Such an influence of diurnal SST fluctuations is through the increase (decrease) of daily mean SST by an enhanced (weakened) diurnal cycle during the period of enhanced (suppressed) convection. Bernie et al. (2007) showed that the inclusion of a diurnal cycle increased the intraseasonal SST response to the MJO by approximately 20% in the simulation with 1-m vertical ocean resolution. Diurnally varying SST has also been argued to play important roles in MJO convection (Webster et al. 1996; Bernie et al. 2008; Bellenger et al. 2010; Woolnough et al. 2007). For instance, the recharge–discharge paradigm that links diurnal SST to deep convection is through the moistening of the troposphere over warmer SST caused by the diurnal cycle (Bladé and Hartmann 1993; Benedict and Randall 2007). This preconditioning of the local atmospheric condition is followed by a rapid moistening of the mid-to-upper troposphere by deep convection (Kikuchi and Takayabu 2004; Kiladis et al. 2005; Benedict and Randall 2007; Haertel et al. 2008). In this context, the realistic simulation of the diurnal cycle of SST is essential to capture the feedback between convection and SST.

The simulation of the MJO has been a long-standing challenge in CGCMs, which generally produce a weaker and more stationary intraseasonal convective oscillation than observed (Lin et al. 2006; Huang et al. 2013). Although deficiencies in the MJO simulation can be caused by errors in atmospheric model physics and model configuration such as the coupling frequency, the near-surface resolution of the ocean model component can also play an important role. Given the commonly used low vertical resolution of 10 m for the upper ocean in the CGCMs, which was found to be too coarse to capture the observed SST diurnal cycle (Bernie et al. 2005), it is interesting to study how the simulated intraseasonal SST variability in an oceanic general circulation model (OGCM) depends on the vertical resolution and, further, how the simulated SST variability is related to the temperature gradient near the surface in the model.

Previous studies on the impact of the vertical resolution in oceanic models used primarily one-dimensional ocean models, regional OGCMs, and regional coupled models, and the results could be model dependent. In this study, we investigate the influence of high vertical resolution in a global OGCM in simulating the observed near-surface vertical temperature structure and the associated SST variability. In particular, surface fields from two recent global climate reanalyses that provide global hourly output are used to force the OGCM. Use of well-validated surface forcing is important for an investigation of the OGCM representation of the SST diurnal cycle and its contribution to the intraseasonal variability. We focus the analysis on the following aspects: 1) observed time mean and diurnal and intraseasonal variations of the vertical temperature structure near the surface, and 2) dependence in the simulation of the observed variability on the model’s vertical resolution. Diurnal variations of temperature vertical profile near the ocean surface are analyzed using in situ observations, which are also used to validate model simulations. Given its demonstrated importance in the simulation of the tropical intraseasonal variability, a realistic representation of SST diurnal variability with enhanced ocean vertical resolution is probably necessary for improved simulation of the tropical MJO as well as the BSISO.

This paper is organized as follows. Section 2 describes the observed data, the ocean model, and the configuration of the experiments. Section 3 analyzes the observed diurnal cycle and intraseasonal variability of the upper-ocean temperature. Section 4 compares and validates the model simulations against observations. A summary is given in section 5.

2. Observations and numerical model

a. Observations

In this study, we first analyze the mean state and diurnal and intraseasonal variability of the upper-ocean temperatures based on in situ measurements at eight locations from the Dynamics of the MJO (DYNAMO) field campaign (Yoneyama et al. 2013; Zhang 2013) and the Research Moored Array for African–Asian–Australian Monsoon Analysis and Prediction (RAMA; McPhaden et al. 2009). The main purpose is to diagnose the depth dependence of the time means and the variability of the temperature and to investigate how realistic the temperature at a depth (e.g., of 5 m) is in representing the SST. Given that most CGCMs use a 10-m top layer for the ocean component, which approximately represent the temperature and other variables at the 5-m depth, we investigate to what extent the ocean surface temperature diurnal and intraseasonal variability can be represented at 5-m depth. Such an analysis requires that the observed data have high resolution in time to delineate the diurnal cycle and resolve vertical variations in the upper 10 m or so.

The DYNAMO field campaign was carried out in the central equatorial Indian Ocean during September 2011–March 2012 (Yoneyama et al. 2013; Zhang 2013). The major objectives of DYNAMO were to improve the understanding of key processes related to the MJO and therefore help improve the MJO simulation and prediction. In situ datasets collected from this project have high vertical and temporal resolutions for both the atmosphere and ocean, allowing detailed analyses of the MJO-related processes, including the vertical stratification of the upper-ocean temperature. Furthermore, the dataset provided an observational validation for model simulations of the processes associated with the MJO variability. Two in situ moored observations at 0°, 79°E and 1.5°S, 79°E from the DYNAMO project are used. Observations at these two locations provide hourly data and are available at a vertical resolution of 1–5 m for the upper 10 m. In addition, observed data from six RAMA buoys are also used. Like the two moored DYNAMO locations, these six RAMA buoys also provide data at high temporal and vertical resolutions.

b. Ocean model and experiments

Previous numerical studies utilized one-dimensional (1D) vertical mixing model to examine the role of ocean vertical resolution in the SST variability (e.g., Bernie et al. 2005). It was argued that the dominant variability is in association with vertical processes rather than horizontal advection, except during westerly wind bursts. In addition, coupled models using an oceanic general circulation model with the low vertical resolution near the ocean surface or using a 1D ocean mixed layer model with the high vertical resolution near the ocean surface, and regional models have been used to explore the role of the ocean on the MJO (Woolnough et al. 2007; Seo et al. 2014). In this study, we extend these previous studies to investigate the influence of the ocean vertical resolution using the global Geophysical Fluid Dynamics Laboratory (GFDL) Modular Ocean Model, version 5 (MOM5; Griffies 2012). Vertical mixing in the MOM5 follows the nonlocal K-profile parameterization (KPP) of Large et al. (1994). Detailed description of the implementation of the KPP is provided in Griffies (2012). The horizontal mixing of tracers uses the isoneutral method developed by Gent and McWilliams (1990). The horizontal mixing of momentum uses the nonlinear scheme of Smagorinsky (Griffies and Halberg 2000). The penetrative solar radiation is approximated by three exponential functions for three wavelength ranges including infrared wavelengths, long visible wavelengths, and short visible and ultraviolet wavelengths. The satellite-derived climatological ocean color data are used to estimate the attenuation length of the exponential profiles (Morel and Antoine 1994).

The model uses a zonal resolution of 0.5° and a meridional resolution of 0.25° between 10°S and 10°N, gradually increasing through the tropics until becoming fixed at 0.5° poleward of 30°S and 30°N. In the vertical, a z* coordinate is used, which is defined as z* = H(zη)/(H + η), where z is the depth of the ocean model level, η the deviation of the ocean surface from a state of rest at z = 0, and H the depth of the ocean bottom of the model grid. For a horizontal ocean grid box, the maximum (minimum) H is set to 4.5 km (50 m). Two sets of vertical resolutions are used. The first set uses 40 layers in the vertical with a 10-m z* thickness for the upper 220 m. The depth of the level for temperature equation zt is in the middle of each layer. The value of ztη (distance between zt and ocean surface) depends on η and H and is about 5 m for the top layer. The second set of the resolution uses 50 layers in the vertical with the z* thickness of 1 m for the upper 9 m, gradually increasing to 10 m and becoming the same as that in the first set. The value of ztη for the top layer in the second resolution is about 0.5 m. For simplicity, we refer to the first and second sets as the 10M and 1M models, emphasizing the difference in the model layer thickness near the surface and refer to the simulations as the 10M and 1M runs, respectively. As commonly used in CGCMs and OGCMs, the temperature of the top layer at about 0.5 m in the 1M model and about 5 m in the 10M model is taken as the simulated SST.

Two parallel sensitivity experiments were performed to investigate the impact of oceanic vertical resolution. The first experiment uses 10-m resolution. The resolution for this experiment represents the commonly used setting in contemporary CGCMs. For each experiment, MOM5 is integrated from 1 September 2011 to 29 February 2012, covering the entire DYNAMO intensive observation period (IOP; 1 October 2011–15 January 2012). Initial fields are taken from the oceanic state of the CFSR. The output is saved at a 1-h interval. The first 20 days of the model simulations are not included in the analysis.

The ocean model is driven by the forcing from atmospheric reanalyses. The quality of the atmospheric forcing influences the mean state and variability of SST, and a high temporal resolution of forcing is required to capture SST diurnal cycle (Seo et al. 2014). In this study, hourly fields from the Climate Forecast System Reanalysis (CFSR; Saha et al. 2010) and the Modern-Era Retrospective Analysis for Research and Applications (MERRA; Rienecker et al. 2011) are used. Both CFSR and MERRA provide the output at a 1-h interval, allowing for a good representation of the diurnal cycle. The atmospheric variables used to force the ocean model include downward longwave radiation flux, net downward shortwave radiative flux, 10-m surface winds, surface pressure, wind stress, and air temperature and specific humidity at 2 m. The momentum flux, downward longwave radiation, and net downward solar radiation are taken directly from the input atmospheric reanalyses, and upward longwave radiation and the turbulent sensible and latent heat fluxes are calculated within the model with wind speed, air temperature, specific humidity, and SST, using a new version of Coupled Ocean–Atmosphere Response Experiment (COARE3.5) algorithm (Fairall et al. 2003; Edson et al. 2013). Surface upward longwave radiation is calculated from a graybody formula with an emissivity of 0.97.

3. Upper-ocean temperatures in the observation

In this section, we analyze mean state and diurnal and intraseasonal variability of the upper-ocean temperatures based on observations at eight mooring locations from the DYNAMO field campaign and RAMA. Figure 1 displays time–depth sections of the mean ocean temperature diurnal cycle at the eight mooring locations. The mean diurnal cycle is calculated as the average over the DYNAMO IOP. The uppermost level in Fig. 1 is at 1-m depth. The temperature at this level is treated as the observed SST in our analysis. There exists a distinct diurnal cycle signal with daily maximum SST in the afternoon and at the uppermost level. Of particular interest is the development of a sharp vertical temperature gradient in the upper ocean within the top 10 m during the daytime. Such a strong vertical gradient is seen at all in situ observations used in the analysis, although the amplitude varies with the buoy locations. The peak of the mean vertical temperature contrast between 1 and 10 m can reach up to 0.5°C. This vertical temperature contrast indicates a strong shallow stratification in the upper ocean. The vertical stratification associated with the diurnal temperature change is shown in the two DYNAMO panels in Fig. 1, right, with the mixed layer depth (MLD; thick solid curves) and isothermal layer depth (ILD; thick dashed curves). The MLD is defined as the depth at which the potential density increases from the surface value is equivalent to that caused by a temperature decrease from the surface temperature by 0.3°C with the salinity held constant at the surface value. The ILD is defined as the depth where the temperature decreases by 0.3°C from surface temperature. Both MLD and ILD show a clear diurnal variation with the shallowest values corresponding to the peak in the near-surface ocean temperature.

Fig. 1.
Fig. 1.

Diurnal cycle of ocean temperature (°C) averaged during the intensive observing period in the DYNAMO field campaign (1 Oct 2011–15 Jan 2012). Observations from DYNAMO are at 0°, 79°E and 1.5°S, 79°E, and those from RAMA are at other locations. Contours are plotted at 0.1°C interval. The x axis is the local time. (right) The thick solid and dashed curves for DYNAMO locations are mixed layer depth and isothermal depth, respectively.

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0689.1

The vertical temperature profile also exhibits a large variability at longer time scales. To reveal its temporal characteristics, hourly temperatures at the depth of 1 m (T1m or SST) and 5 m (T5m) are compared in Fig. 2 for the two DYNAMO locations. The amplitude of vertical contrast at individual times can be more pronounced than the period average in Fig. 1. T1m is generally warmer than T5m, and their differences frequently exceed 0.5°C (blue curves in Fig. 2). A closer examination shows that both the amplitude and variability of thermal contrast are largely determined by T1m, which has a much stronger diurnal variability.

Fig. 2.
Fig. 2.

Comparisons between hourly in situ temperature (°C) at the depth of 1 m (red) and 5 m (black), and their differences (blue; right y axis) during the DYNAMO period.

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0689.1

Near-surface temperatures can differ substantially from the SST, particularly under the conditions of low winds and strong solar insolation. Generally, the warming (cooling) period is during the dry (wet) phase of the MJO event. Physically, the diurnal variation of SST is a consequence of the influence of contrasting processes during daytime and nighttime. The SST diurnal warming is produced by the absorption of solar insolation at the ocean surface and is weakened by mechanical wind-driven mixing, surface latent and sensible heat fluxes, and precipitation cooling (Webster et al. 1996). The radiative forcing of the shortwave flux is absorbed over the upper few tens of meters of the ocean. Accordingly, to reasonably simulate the vertical profile of heating rate in a numerical model, the uppermost layers must be thin enough to capture the nature of the vertical distribution of shortwave radiation.

As pointed out by Weller and Anderson (1996), the intraseasonal SST variation in the equatorial Pacific exhibits markedly different subdaily fluctuations depending on the phases of the MJO convection. The diurnal SST variation was nearly absent during the convectively active and windy periods but was enhanced and accompanied by strong and shallow stratification during calm and clear-sky periods. Moum et al. (2014) observed a large diurnal warming in the upper few meters during the convectively suppressed conditions in the time of the DYNAMO campaign. Our analysis confirms that the temperature gradient near the surface exhibits a marked intraseasonal variability as well, and a sharp vertical temperature gradient develops during a warming period, whereas the vertical temperature gradient is weak when in a cooling period.

One can also argue that the diurnal cycle and vertical temperature gradient will be strongly modulated by the surface flux anomalies associated with the MJO. For example, to the east of the enhanced convection in the Indian Ocean and western Pacific, corresponding to a warming episode with the ocean receiving above-average incoming solar radiation as a result of the less cloudy sky and losing less evaporative heat flux as a result of weakened surface winds, the mixed layer depth would shoal. As a consequence, the heat capacity of the mixed layer is reduced and the daytime temperature of the shoaling boundary layer will become more sensitive to the heat fluxes. This would produce a warmer mixed layer temperature and SST, allowing an increase in buoyancy stratification to build up beneath the mixed layer due to the vertical gradient in the absorption of shortwave radiation. In addition, the relatively small wind-driven turbulence will also be suppressed by the strong vertical gradient of buoyancy forcing.

How strongly does the vertical temperature gradient near the surface vary in time? To highlight changes of the vertical temperature gradient in upper-ocean temperature under different conditions, the daily mean temperature composite of large gradient cases is calculated and compared with the average profile calculated based on all data. Here, one standard deviation (STD) above the mean amplitude of the temperature contrast is used to quantify the strong gradient cases. The STD of upper-ocean temperature difference between 1- and 5-m depth (DT1m5m) is first computed. Cases with DT1m5m amplitude greater than one STD of DT1m5m are classified into the group STRONG and are compared with the group ALL that includes all observations.

Figure 3 compares the composited vertical temperature profiles between STRONG and ALL for two DYNAMO campaign locations. The temperature difference between 1 and 5 m in the STRONG group is more pronounced than that in the ALL group with the DT1m5m in the STRONG group being about 0.45°C at 0°, 79°E and approximately 0.35°C at 1.5°S, 79°E, compared to the corresponding values of 0.15°C at 0°, 79°E and 0.1°C at 1.5°S, 79°E in the ALL group. More importantly, the temperature changes at 1 m in ALL and STRONG are 0.4° and 0.3°C, respectively, at 1.5°S, 79°E, while the changes at 5 m are 0.1°C or less. If the temperature at 5 m is used to represent SST, it would significantly underestimate the surface temperature variation. Composites are also calculated for the RAMA locations, which show a similar contrast between STRONG and ALL groups (not shown). Consistent with earlier discussion, further analyses show that the occurrence of most of the STRONG cases took place east of the MJO-associated convection under the clear-sky condition. This would also indicate that the warm temperatures in the STRONG group may contribute to preconditioning the lower troposphere with a warmer and moister state and thus help organize the eastward propagation of the MJO.

Fig. 3.
Fig. 3.

Composited vertical daily mean temperature profiles for two groups at (a) 0°, 79°E and (b) 1.5°S, 79°E. The STRONG group consists of cases with the amplitude of DT1m5m greater than one STD of DT1m5m. The composite for the ALL group uses all observations.

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0689.1

These results suggest that the use of the bulk temperature representing temperature at a depth of a few meters in an observational SST analysis such as the analysis of Reynolds et al. (2007) or the use of the upper 10 m as the top layer in an ocean model may underestimate the warm ocean surface temperature east of the MJO convection. Various studies have shown the importance of the warm SST anomaly east of the MJO convection in the preconditioning of a warm moist lower atmosphere for eastward propagation of the MJO (Waliser et al. 1999; Zhang 2005). One possible reason for the slow propagation of the CGCM-simulated MJO is the underrepresentation of SST anomalies, resulting in a slower preconditioning of the lower troposphere to the east of the convection than in reality.

4. Impacts of ocean model vertical resolution

The analysis in section 2 indicates the existence of large vertical temperature gradient in the upper 10 m, especially during the dry and clear-sky period of the MJO event. Previous studies (Bernie et al. 2005) pointed out that the diurnal SST variability in model simulations was dependent crucially on the resolution of the uppermost model level, and a realistic representation of the diurnal cycle also led to stronger intraseasonal SST variability. Simulation of the observed diurnal and intraseasonal SST variability, therefore, will require a vertical resolution of about 1 m in the upper ocean. In this section, we investigate the impact of the vertical resolution on the simulation of the sharp and shallow temperature stratification in the upper ocean based on numerical experiments with MOM5. In particular, we analyze to what extent the simulated diurnal and intraseasonal SST variabilities are changed under different vertical resolutions.

Given the dependence on the atmospheric surface forcing, especially on the downward shortwave radiation flux (SW), of the simulated diurnal and intraseasonal SST variability, the uncertainty in the CFSR and MERRA reanalysis data and its impact on the simulated SST were first assessed. Figure 4 evaluates SW mean bias taken as the difference between the reanalyses and the Clouds and the Earth’s Radiant Energy System (CERES) product (Loeb et al. 2001). Overall, SW from CFSR (MERRA) is overestimated (underestimated) by 25 W m−2 or so in the tropical Indian Ocean and western Pacific (Figs. 4a,b), while their average is close to the CERES observational values (Fig. 4c).

Fig. 4.
Fig. 4.

Differences in downward solar radiation flux (W m−2) between reanalyses and CERES estimate averaged from September 2011 to January 2012: (a) CFSR minus CERES, (b) MERRA minus CERES, and (c) hybrid minus CERES.

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0689.1

The CERES, CFSR, and MERRA products are further validated against DYNAMO observations at 1.5°S, 79°E (Fig. 5). Monthly mean SW at this DYNAMO location shows a good agreement between CERES and the DYNAMO observational estimates, while the CFSR (MERRA) produces a positive (negative) bias in SW compared to the DYNAMO and CERES observations (Fig. 5a). Because of the opposite signs of the biases in SW, a hybrid dataset is formed by taking the average of CFSR and MERRA for individual fields. It is shown that the monthly SW from the hybrid dataset is close to the CERES and DYNAMO values, except for October 2011 in which the hybrid is about 25 W m−2 lower than the DYNAMO and CERES observations. The SW diurnal cycle in the hybrid dataset is much more reasonable than in the individual reanalyses (Fig. 5b).

Fig. 5.
Fig. 5.

Comparisons at the DYNAMO observation location 1.5°S, 79°E. (a) Monthly mean downward shortwave radiation flux (SW). (b) Mean SW diurnal cycle average from October 2011 to January 2012. (c) Intraseasonal SW anomalies and SST. (d) As in (c), but for W10m. The hybrid reanalysis in each panel is taken as the average of CFSR and MERRA reanalyses. For SW and W10m, the DYNAMO observations and CFSR, MERRA, and hybrid reanalyses are plotted with dotted, long dashed, short dashed, and thin solid curves, respectively. The CERES SW in (a) and DYNAMO SST in (c) and (d) are plotted with thick solid curves.

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0689.1

Figure 5c compares SW intraseasonal anomalies between DYNAMO and the hybrid dataset. The intraseasonal anomalies are obtained by (i) computing differences between the total field and 31-day running average and (ii) taking the 11-day running average of the resulting differences. The DYNAMO and hybrid SW anomalies are in phase and lead the SST anomalies by 10 days or so, although the intraseasonal amplitude of hybrid SW appears to be weaker than that of the DYNAMO SW. In addition to SW, the surface latent heat flux variability, which is dominated by the surface wind speed anomalies, is important in maintaining a realistic SST variability (Flatau et al. 1997; Shinoda et al. 1998). Figure 5d shows that maximum (minimum) SST anomalies are generally led by the weakest (strongest) 10-m wind speed (W10m) anomalies by 3–5 days, and the hybrid dataset provides realistic W10m intraseasonal anomalies compared to the DYNAMO anomalies.

Consistent with the SW bias, simulated SSTs driven by the forcing fields from CFSR (MERRA) are systematically too warm (cold) in these regions. As shown in Fig. 6, the simulated SST at 1.5°S, 79°E forced by CFSR (MERRA) fields was 0.5°–1°C warmer (cooler) than the observation for most of the period. When used as a forcing for the ocean model, the hybrid dataset was found to reduce the systematic bias in the simulated SST compared to that forced by individual CFSR or MERRA fields (Fig. 6). The model results to be shown next are all from the simulations forced by the hybrid dataset.

Fig. 6.
Fig. 6.

Evolution of the daily mean SST observation and 1M simulation with MOM5 at 1.5°S, 79°E.

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0689.1

To validate the model performance in simulating mean state, the SST averaged from October 2011 to January 2012 in the 1M simulation and 10M simulation is compared with the Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI) retrieval (Wentz et al. 2000) for the tropical Indian Ocean (Fig. 7). Spatially, there is a good agreement between the model and observation (Figs. 7a–c) with the amplitude of the simulated SST bias being generally less than 0.3 K in tropical Indian Ocean (Figs. 7d,e). Except for the coastal areas, the 1M simulation is warmer than the 10M by 0.1–0.2 K, especially for the tropical southern Indian Ocean.

Fig. 7.
Fig. 7.

The October 2011–January 2012 mean SST (°C): (a) 1M simulation, (b) 10M simulation, (c) TMI, (d) differences between 1M simulation and TMI, (e) differences between 10M simulation and TMI, and (f) differences between the two simulations.

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0689.1

Comparisons in the SST at 1.5°S, 79°E between model simulations and in situ observations are shown in Fig. 8a for 24 September–12 December 2011. The model-simulated SST is taken as the temperature at the top model level for respective simulations and at the top available level (1-m depth) for the observations. In general, the model captures the observed SST temporal variations quite well. The simulated SST qualitatively resembles the observed changes in intraseasonal time scales, with warmer and cooler episodes well reproduced. While both the simulation and the observation exhibit diurnal cycle variations, the diurnal amplitude in the 1M simulation is comparable to the observed and the diurnal amplitude in the 10M simulation is much weaker than that in the 1M simulation and observation. To highlight such discrepancies better, the evolution of hourly SST at 1.5°S, 79°E during a shorter period (4–17 November 2011) is shown in Fig. 8b. The SST from the 1M simulation produces larger diurnal variability, with a magnitude similar to the diurnal cycle, than that observed. The strong diurnal signal corresponds to the regimes of high insolation and reduced wind speeds (not shown). By contrast, the amplitude of the diurnal cycle of SST in the 10M simulation is much weaker. Comparisons between the simulations and observation for the TAO/TRITON buoys give a similar result that the temporal SST evolution is more reasonably simulated with 1-m vertical resolution (not shown).

Fig. 8.
Fig. 8.

Evolution of hourly SSTs at 1.5°S, 79°E in simulations and observations during the DYNAMO period: (a) 24 Sep–12 Dec 2011 and (b) 4–17 Nov 2011. Black, blue, and red curves are for the observation, 1M simulation, and 10M simulation, respectively.

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0689.1

a. Diurnal cycle variation

Impacts of the ocean vertical resolution on the simulated near-surface temperature are shown in Fig. 9, which presents the mean diurnal variation of the vertical temperature profile in the upper 40 m at 1.5°S, 79°E for the average from 1 October 2011 to 15 January 2012 and for the warming and cooling phases of MJO events during this period. For the average in the 1M simulation (Fig. 9a), a sharp vertical temperature gradient emanates within the top 10 m, which agrees well with the observation (Fig. 1, bottom right). The corresponding vertical profile of ocean temperature in the 10M simulation (Fig. 9d) is much weaker. The result highlights the importance of fine oceanic vertical resolution in simulating a realistic temperature profile in the upper oceans. Compared to the DYNAMO observation (Fig. 1, bottom right), the MLD and ILD in the 1M simulation show realistic diurnal variations (Fig. 9a) while MLD and ILD diurnal variations in the 10M simulation are much too weak.

Fig. 9.
Fig. 9.

Diurnal variation of vertical temperature profile in the upper ocean at 1.5°S, 79°E in the (a)–(c) 1M simulation and (d)–(f) 10M simulation. (left) The average for September 2011–January 2012 and for (center) warming and (right) cooling phases of MJO events during this period. The thick solid and dashed curves are mixed layer depth and isothermal depth. Horizontal dotted lines indicate the bottom of model layers.

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0689.1

An overall assessment of the influence of vertical resolution on diurnal variability is presented in Fig. 10, which displays the horizontal distribution of mean diurnal cycle range (DCR) during the integration period. The DCR is defined as the temperature difference between the daytime maximum and nighttime minimum. Not surprisingly, for the temperature at the model top level, which is taken as SST, the 1M simulation generates a larger DCR than the 10M simulation does (Figs. 10a–c). In the 10M simulation, the SST DCR is about 0.2°C with small spatial variations. The SST DCR in the 1M simulation is much stronger (~0.4°–0.5°C). Spatially, the SST DCR in the 1M simulation is larger (0.5°–0.6°C) along the equator and relatively smaller (<0.4°C) 10° latitude or so off the equator. The pattern in the 1M simulation bears similarities to the modeling results in Li et al. (2013), although a different model and a different atmospheric forcing dataset were used in their study. The 1M simulation shows that the large diurnal cycle is mostly confined within the upper 5 m, and at lower levels beneath 10-m depth the diurnal cycle is weak in both the 1M and 10M simulations (Figs. 9a,d). A comparison in spatial distribution of temperature DCR at 15-m depth is presented in Fig. 10 to show the weakening of DCR at lower levels. It is seen that the temperature DCR at 15-m depth is much weaker than that at the top level for both simulations and that 15-m temperature DCR differences between the two simulations are near zero in the entire tropical Indian Ocean (Figs. 10d–f).

Fig. 10.
Fig. 10.

Horizontal distributions of mean temperature diurnal cycle range (°C). (a) SST in the 1M simulation, (b) SST in the 10M simulation, and (c) SST differences between the two simulations and temperature at 15-m depth in the (d) 1M simulation, (e) 10M simulation, and (f) the differences between the two simulations.

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0689.1

The average SST DCR at six buoy locations near the equator from the observations and simulations is compared in Table 1. The diurnal range in the 10M simulation is consistently weaker than the observations at all six locations, while the diurnal range from the 1M simulation is close to the observed. The amplitude of DCR in the 10M simulation is generally smaller than 0.2°C, while the amplitude in the 1M simulation is 0.38°–0.53°C, which is in good agreement with the observed range of 0.33°–0.50°C. On average, the diurnal range in the 1M (10M) simulation is 0.43°C (0.19°C), compared to the observed 0.42°C, indicating that the mean diurnal cycle is too weak in the 10M simulation and is dramatically improved in the 1M simulation.

Table 1.

Comparisons of simulated SST diurnal range and intraseasonal STD (K) with observations at six buoy locations.

Table 1.

b. Intraseasonal variability

To analyze differences in the near-surface temperature at the intraseasonal time scale associated with the variations in the atmospheric conditions, composites of temperature profile are made for the warming (Figs. 9b,e) and cooling (Figs. 9c,f) phases. The warming (cooling) phase is taken as the period when the 20–80-day filtered SST increases (decreases) from minimum (maximum) to maximum (minimum). The 1M (10M) simulation produces a strong (weak) diurnal cycle and large (small) vertical temperature gradient during the warming (cooling) phase of the MJO convection (Figs. 9b,c,e,f). In particular, the 1M (10M) simulation produces much stronger (weaker) contrast in diurnal cycle and daily mean of the top-level temperature (i.e., the SST) between the warming and cooling phases.

It has been recognized that the diurnal cycle enhances the intraseasonal SST variability. To validate the impacts of diurnal cycle on intraseasonal variability, we compare the intraseasonal variability of SST between two simulations. The Lanczos digital bandpass filter (Duchon 1979) is first applied to extract the 20–80-day intraseasonal SST anomaly component. The STD of the intraseasonal anomalies is then computed and shown in Fig. 11.

Fig. 11.
Fig. 11.

Horizontal distributions of mean intraseasonal STD (°C) in the (a) 1M simulation, (b) 10M simulation, and (c) differences between the 1M and 10M simulations. (d) As in (c), but the differences are between two additional runs with daily mean forcing instead of hourly forcing as in (a) and (b).

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0689.1

The amplitude of the STD is generally less than 0.15°C in the 10M simulation, while the STD amplitude in the 1M simulation is larger than 0.2°C in most of the Indian Ocean, especially for the equatorial area where the STD is greater than 0.3°C in the western and eastern Indian Ocean. Differences between the two simulations (Fig. 11c) show that the intraseasonal SST amplitude in the 1M simulation is consistently larger than that in the 10M simulation by 0.05°–0.1°C. These results underscore that intraseasonal SST variability is enhanced by increasing the vertical resolution in the upper ocean. It is also noticeable that the areas with significant differences in STD between the two simulations (Fig. 11c) are also the areas of significant differences in DCR (Fig. 10c), reflecting the important role of the diurnal cycle in the enhancement of intraseasonal variability as also revealed in previous studies (Bernie et al. 2005; Shinoda 2005). To confirm the role of diurnal cycle in our simulations with MOM5, we carried out two additional simulations with daily mean, instead of hourly, forcing. Differences in the standard SST STD between these two additional simulations are presented in Fig. 11d, which shows that the 1-m vertical resolution with daily mean forcing produces stronger intraseasonal amplitude than 10-m vertical resolution. However, the amplitude of the STD differences with daily mean forcing (Fig. 11d) is much smaller than that with hourly forcing (Fig. 11c), indicating the important role of diurnal cycle in rectifying intraseasonal variability.

To further compare the intraseasonal SST evolution represented in the two experiments, Fig. 12 displays the time–longitude diagram of the simulated 20–80-day bandpass-filtered SST averaged between 5°S and 5°N during the DYNAMO IOP. There were three MJO events during this period initiated in the Indian Ocean in October, November, and December (Fu et al. 2015). Alternating positive and negative SST anomalies propagate eastward, leading to the enhanced and weakened convection associated with the MJO events (Fu et al. 2015). With the identical surface atmospheric forcing, similar time evolution patterns, dominated by eastward propagations, are simulated in both experiments. However, the amplitude of the intraseasonal SST anomalies is consistently stronger in the 1M simulation than that in the 10M simulation. Further, the SST anomaly differences (1M minus 10M) are in phase with the SST anomalies, leading to enhanced amplitude of both positive and negative anomalies in the 1M simulation compared to the 10M simulation.

Fig. 12.
Fig. 12.

Time–longitude diagram of the simulated 20–80-day bandpass-filtered SST averaged between 5°S and 5°N: (a) 1M simulation, (b) 10M simulation, and (c) their differences.

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0689.1

A quantitative comparison of the intraseasonal SST STD between the experiments and observations is presented in Table 1 for six locations. The intraseasonal SST STD in the 10M simulation is too weak compared with the observed at all locations except for 4°S, 80.5°E, where the STD in the 10M simulation is the same as that observed. The STD difference between the 1M simulation and the observation is generally less than 0.04°C except for 1.5°N, 90°E, where the 1M simulation is 0.08°C too strong. The average SST STD averaged across all six locations in the 1M (10M) simulation is 0.25°C (0.18°C), a slight (significant) overestimate (underestimate) of the observed value of 0.23°C.

5. Summary and discussion

In this study, we examined diurnal and intraseasonal variabilities of the sea surface temperature (SST) based on observations from moored buoys and simulations from the Geophysical Fluid Dynamics Laboratory (GFDL) Modular Ocean Model, version 5 (MOM5), focusing on the impact of the vertical resolution. Compared to previous studies that used primarily one-dimensional ocean models, regional OGCMs, and regional coupled models, this study investigated the influence of high vertical resolution in a global OGCM with surface forcing fields from two recent global climate reanalyses that provide global hourly output.

It was found that there were strong temperature variations near the ocean surface, and a sharp vertical temperature gradient existed within the top 10 m of the ocean. Both diurnal and intraseasonal variabilities of the ocean temperatures were largest near the surface and decreased with the ocean depth. The vertical temperature stratification in the upper ocean favored the diurnal warming and was modulated at intraseasonal time scales depending on the condition of the atmosphere with the largest (weakest) diurnal warming and vertical gradient occurring during the dry clear-sky (wet cloudy sky) period, contributing to intraseasonal SST variability. During the dry period, a sharp temperature contrast, corresponding to a shoaled mixed layer, favored to reduce the heat capacity. The daytime temperature of shoaled mixed layer became increasingly sensitive to the heat fluxes. This produced a warmer upper ocean with increased buoyancy stratification built up beneath the surface because of the vertical gradient in the absorption of shortwave radiation. As a consequence, the relatively small wind-driven turbulence was suppressed by the strong vertical gradient of buoyancy forcing. This indicated that the vertical stratification of the upper ocean was important in modulating the amplitude of diurnal as well as intraseasonal variabilities of the SST.

The existence of the large vertical temperature gradient near the ocean surface indicated that for an OGCM to realistically simulate the SST variability a sufficient vertical resolution must be used. Yet a 10-m upper layer has been commonly used for the ocean component in global coupled atmosphere–ocean models. Bernie et al. (2005) showed that an upper-layer thickness on the order of 1 m is required based on their experiments with a one-dimensional ocean mixed layer model. In this study, the impact of vertical resolution in the upper ocean in simulating SST variability was further investigated with the GFDL MOM5. Two simulations were carried out, one with 1-m and the other with 10-m vertical resolution for the upper layer referred to as the 1M and 10M simulations, respectively. The simulations were assessed against observational data from moored buoys. The representation of the diurnal and intraseasonal variability of SST was found to be highly sensitive to the vertical resolution of the ocean model. The vertical temperature gradient near the surface, as well as the diurnal and intraseasonal variability, was strongly influenced by the oceanic vertical resolution, with the use of the 1-m vertical resolution producing a stronger vertical temperature gradient and enhanced diurnal and intraseasonal variability. The results indicated that a proper vertical resolution is critical for the SST simulation and prediction with realistic representation of ocean processes in the upper ocean.

The simulations in this study are based on an uncoupled ocean model driven by specified atmospheric forcing. In nature, the atmosphere and ocean mutually interact with each other. While the SST varies in response to changes in the atmospheric circulation, the resulting SST variations also affect the atmosphere by contributing the preconditioning of the atmosphere through modifications of heat fluxes prior to deep convection. Wang et al. (2015) showed that the simulation of the observed atmospheric intraseasonal variability strongly depends on the accuracy of the underlying SST. Two observational SST analyses were tested in Wang et al. (2015), the TRMM Microwave Imager (TMI) SST retrieval of Wentz et al. (2000), which was taken as the ocean skin temperature, and the National Climatic Data Center (NCDC) daily SST analysis of Reynolds et al. (2007), which was defined as the bulk temperature of the upper few meters of the ocean. The daily mean SST standard deviation (STD) of the TMI retrieval was about 0.1°C greater than that of the NCDC analysis, and the simulation forced with the TMI retrieval produced much more realistic representation of the eastward propagation of the MJO than that forced with the NCDC SST. Given that the amplitude of the intraseasonal SST STD in the 1M simulation is generally stronger than that in the 10M simulation, it is expected that the use of a 1-m vertical resolution for the ocean component will lead to an improved simulation and prediction of the observed tropical intraseasonal oscillations, including the tropical MJO and boreal summer intraseasonal oscillation, and may also contribute to improved prediction of active and break phases of the Indian summer monsoon.

Acknowledgments

The authors thank Ren-Chieh Lien at the University of Washington for providing us the COARE3.5 code as well as the DYNAMO observational data. The RAMA moored data are provided by the Pacific Marine Environmental Laboratory. We greatly appreciate the helpful reviews by Daniel Harnos, Jieshun Zhu, and three anonymous reviewers. Xuyang Ge gratefully acknowledges the financial support given by the Earth System Science Organization, Ministry of Earth Sciences, government of India, to conduct this research under the Monsoon Mission. Ying Zhang is also supported by the NOAA Climate Program Office CVP program under the project “Improvement of MJO Simulation in NCEP Coupled Forecast System: Upper Ocean and Air-Sea Coupled Processes” (Grant NA15OAR4310173).

REFERENCES

  • Bellenger, H., Y. N. Takayabu, T. Ushiyama, and K. Yoneyama, 2010: Role of diurnal warm layers in the diurnal cycle of convection over the tropical Indian Ocean during MISMO. Mon. Wea. Rev., 138, 24262433, doi:10.1175/2010MWR3249.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Benedict, J., and D. A. Randall, 2007: Observed characteristics of the MJO relative to maximum rainfall. J. Atmos. Sci., 64, 23322354, doi:10.1175/JAS3968.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bernie, D. J., S. J. Woolnough, and J. M. Slingo, 2005: Modeling diurnal and intraseasonal variability of the ocean mixed layer. J. Climate, 18, 11901202, doi:10.1175/JCLI3319.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bernie, D. J., E. Guilyardi, G. Madec, J. M. Slingo, and S. J. Woolnough, 2007: Impact of resolving the diurnal cycle in an ocean–atmosphere GCM. Part 1: A diurnally forced OGCM. Climate Dyn., 29, 575590, doi:10.1007/s00382-007-0249-6.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bernie, D. J., E. Guilyardi, G. Madec, J. M. Slingo, S. J. Woolnough, and J. Cole, 2008: Impact of resolving the diurnal cycle in an ocean–atmosphere GCM. Part 2: A diurnally coupled CGCM. Climate Dyn., 31, 909925, doi:10.1007/s00382-008-0429-z.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bladé, I., and D. L. Hartmann, 1993: Tropical intraseasonal oscillation in a simple nonlinear model. J. Atmos. Sci., 50, 29222939, doi:10.1175/1520-0469(1993)050<2922:TIOIAS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Duchon C. E., 1979: Lanczos filtering in one and two dimensions. J. Appl. Meteor., 18, 10161022, doi:10.1175/1520-0450(1979)018<1016:LFIOAT>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Edson, J. B., and Coauthors, 2013: On the exchange of momentum over the open ocean. J. Phys. Oceanogr., 43, 15891610, doi:10.1175/JPO-D-12-0173.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fairall, C. W., E. F. Bradley, J. E. Hare, A. A. Grachev, and J. B. Edson, 2003: Bulk parameterization of air–sea fluxes: Updates and verification for the COARE algorithm. J. Climate, 16, 571590, doi:10.1175/1520-0442(2003)016<0571:BPOASF>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Flatau, M., P. J. Flatau, P. Phoebus, and P. P. Niiler, 1997: The feedback between equatorial convection and local radiative and evaporative processes: The implication for intraseasonal oscillations. J. Atmos. Sci., 54, 23732386, doi:10.1175/1520-0469(1997)054<2373:TFBECA>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fu, X., B. Wang, D. E. Waliser, and L. Tao, 2007: Impact of atmosphere–ocean coupling on the predictability of monsoon intraseasonal oscillations. J. Atmos. Sci., 64, 157174, doi:10.1175/JAS3830.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fu, X., W. Wang, J.-Y. Lee, B. Wang, K. Kikuchi, J. Xu, J. Li, and S. Weaver, 2015: Distinctive roles of air–sea coupling on different MJO events: A new perspective revealed from the DYNAMO/CINDY field campaign. Mon. Wea. Rev., 143, 794812, doi:10.1175/MWR-D-14-00221.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gent, P. R., and J. C. McWilliams, 1990: Isopycnal mixing in ocean circulation models. J. Phys. Oceanogr., 20, 150155, doi:10.1175/1520-0485(1990)020<0150:IMIOCM>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Griffies, S. M., 2012: Elements of the Modular Ocean Model (MOM): 2012 release. NOAA/GFDL Ocean Group Tech. Rep. 7, 631 pp. [Available online at http://www.mom-ocean.org/web/docs/project/MOM5_elements.pdf.]

  • Griffies, S. M., and R. W. Hallberg, 2000: Biharmonic friction with a Smagorinsky viscosity for use in large-scale eddy-permitting ocean models. Mon. Wea. Rev., 128, 29352946, doi:10.1175/1520-0493(2000)128<2935:BFWASL>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Haertel, P. T., G. N. Kiladis, A. Denno, and T. M. Rickenbach, 2008: Vertical-mode decompositions of 2-day waves and the Madden–Julian oscillation. J. Atmos. Sci., 65, 813833, doi:10.1175/2007JAS2314.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Huang, B., M. L’Heureux, J. Lawrimore, C. Liu, H.-M. Zhang, V. Banzon, Z.-Z. Hu, and A. Kumar, 2013: Why did large differences arise in the sea surface temperature datasets across the tropical Pacific during 2012? J. Atmos. Oceanic Technol., 30, 29442953, doi:10.1175/JTECH-D-13-00034.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Inness, P. M., and J. M. Slingo, 2003: Simulation of the Madden–Julian Oscillation in a coupled general circulation model. Part I: Comparison with observations and an atmosphere-only GCM. J. Climate, 16, 345364, doi:10.1175/1520-0442(2003)016<0345:SOTMJO>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kemball-Cook, S., B. Wang, and X. Fu, 2002: Simulation of the intraseasonal oscillation in the ECHAM-4 model: Impact of coupling with an ocean model. J. Atmos. Sci., 59, 14331453, doi:10.1175/1520-0469(2002)059<1433:SOTIOI>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kikuchi, K., and Y. N. Takayabu, 2004: The development of organized convection associated with the MJO during TOGA COARE IOP: Trimodal characteristics. Geophys. Res. Lett., 31, L10101, doi:10.1029/2004GL019601.

    • Crossref
    • 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, doi:10.1175/JAS3520.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Large, W. G., J. C. McWilliams, and S. C. Doney, 1994: Oceanic vertical mixing: Review and a model with a nonlocal boundary layer parameterization. Rev. Geophys., 32, 363403, doi:10.1029/94RG01872.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, Y., W. Han, T. Shinoda, C. Wang, R.-C. Lien, J. N. Moum, and J. W. Wang, 2013: Effects of solar radiation diurnal cycle on the tropical Indian Ocean mixed layer variability during wintertime Madden-Julian oscillation events. J. Geophys. Res. Oceans, 118, 49454964, doi:10.1002/jgrc.20395.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lin, J.-L., and Coauthors, 2006: Tropical intraseasonal variability in 14 IPCC AR4 climate models. Part I: Convective signals. J. Climate, 19, 26652690, doi:10.1175/JCLI3735.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Loeb, N. G., K. J. Priestley, D. P. Kratz, E. B. Geier, R. N. Green, B. A. Wielicki, P. O’Rawe Hinton, and S. K. Nolan, 2001: Determination of unfiltered radiances from the Clouds and the Earth’s Radiant Energy System instrument. J. Appl. Meteor., 40, 822835, doi:10.1175/1520-0450(2001)040<0822:DOURFT>2.0.CO;2.

    • Crossref
    • 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, doi:10.1175/1520-0469(1972)029<1109:DOGSCC>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • McPhaden, M. J., and Coauthors, 2009: RAMA: The Research Moored Array for African–Asian–Australian Monsoon Analysis and Prediction. Bull. Amer. Meteor. Soc., 90, 459480, doi:10.1175/2008BAMS2608.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Morel, A., and D. Antoine, 1994: Heating rate within the upper ocean in relation to its bio-optical state. J. Phys. Oceanogr., 24, 16521665, doi:10.1175/1520-0485(1994)024<1652:HRWTUO>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Moum, J. N., and Coauthors, 2014: Air–sea interactions from westerly wind bursts during the November 2011 MJO in the Indian Ocean. Bull. Amer. Meteor. Soc., 95, 11851199, doi:10.1175/BAMS-D-12-00225.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Reynolds, R. W., T. M. Smith, C. Liu, D. B. Chelton, K. S. Casey, and M. G. Schlax, 2007: Daily high-resolution-blended analyses for sea surface temperature. J. Climate, 20, 54735496, doi:10.1175/2007JCLI1824.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rienecker, M. M., and Coauthors, 2011: MERRA: NASA’s Modern-Era Retrospective Analysis for Research and Applications. J. Climate, 24, 36243648, doi:10.1175/JCLI-D-11-00015.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Saha, S., and Coauthors, 2010: The NCEP Climate Forecast System Reanalysis. Bull. Amer. Meteor. Soc., 91, 10151067, doi:10.1175/2010BAMS3001.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Seo, H., A. C. Subramanian, A. J. Miller, and N. R. Cavanaugh, 2014: Coupled impacts of the diurnal cycle of sea surface temperature on the Madden–Julian oscillation. J. Climate, 27, 84228443, doi:10.1175/JCLI-D-14-00141.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shinoda, T., 2005: Impact of the diurnal cycle of solar radiation on intraseasonal SST variability in the western equatorial Pacific. J. Climate, 18, 26282636, doi:10.1175/JCLI3432.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shinoda, T., and H. H. Hendon, 1998: Mixed layer modeling of intraseasonal variability in the tropical western Pacific and Indian Oceans. J. Climate, 11, 26682685, doi:10.1175/1520-0442(1998)011<2668:MLMOIV>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shinoda, T., and H. H. Hendon, 2001: Upper-ocean heat budget in response to the Madden–Julian oscillation in the western equatorial Pacific. J. Climate, 14, 41474165, doi:10.1175/1520-0442(2001)014<4147:UOHBIR>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shinoda, T., H. H. Hendon, and J. Glick, 1998: Intraseasonal variability of surface fluxes and sea surface temperature in the tropical western Pacific and Indian Oceans. J. Climate, 11, 16851702, doi:10.1175/1520-0442(1998)011<1685:IVOSFA>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Waliser, D. E., K. E. 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, doi:10.1175/1520-0469(1999)056<0333:TIOCSS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, W., M. Chen, and A. Kumar, 2009: Impacts of ocean surface on the northward propagation of the boreal summer intraseasonal oscillation in the NCEP Climate Forecast System. J. Climate, 22, 65616576, doi:10.1175/2009JCLI3007.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, W., A. Kumar, and X. Fu, 2015: How important is the sea surface temperature uncertainty in the prediction of tropical convection associated with the MJO? Mon. Wea. Rev., 143, 31563175, doi:10.1175/MWR-D-14-00385.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Webster, P. J., C. A. Clayson, and J. A. Curry, 1996: Clouds, radiation, and the diurnal cycle of sea surface temperature in the tropical western Pacific Ocean. J. Climate, 9, 17121730, doi:10.1175/1520-0442(1996)009<1712:CRATDC>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Weller, R. A., and S. P. Anderson, 1996: Surface meteorology and air–sea fluxes in the western equatorial Pacific warm pool during the TOGA Coupled Ocean–Atmosphere Response Experiment. J. Climate, 9, 19591990, doi:10.1175/1520-0442(1996)009<1959:SMAASF>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wentz, F. J., C. L. Gentemann, D. K. Smith, and D. B. Chelton, 2000: Satellite measurements of sea surface temperature through clouds. Science, 288, 847850, doi:10.1126/science.288.5467.847.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Woolnough, S. J., F. Vitart, and M. A. Balmaseda, 2007: The role of the ocean in the Madden-Julian oscillation: Implications for MJO prediction. Quart. J. Roy. Meteor. Soc., 133, 117128, doi:10.1002/qj.4.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Yoneyama, K., C. Zhang, and C. N. Long, 2013: Tracking pulses of the Madden–Julian oscillation. Bull. Amer. Meteor. Soc., 94, 18711891, doi:10.1175/BAMS-D-12-00157.1.

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

  • Zhang, C., 2013: Madden–Julian oscillation: Bridging weather and climate. Bull. Amer. Meteor. Soc., 94, 18491870, doi:10.1175/BAMS-D-12-00026.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, C., and S. P. Anderson, 2003: Sensitivity of intraseasonal perturbations in SST to the structure of the MJO. J. Atmos. Sci., 60, 21962206, doi:10.1175/1520-0469(2003)060<2196:SOIPIS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
Save