• Aldrian, E., , and R. Dwi Susanto, 2003: Identification of three dominant rainfall regions within Indonesia and their relationship to sea surface temperature. Int. J. Climatol., 23, 14351452.

    • Search Google Scholar
    • Export Citation
  • Aldrian, E., , L. Dümenil Gates, , and F. H. Widodo, 2007: Seasonal variability of Indonesian rainfall in ECHAM4 simulations and in the reanalyses: The role of ENSO. Theor. Appl. Climatol., 87, 4159.

    • Search Google Scholar
    • Export Citation
  • Beringer, J., , N. J. Tapper, , and T. D. Keenan, 2001: Evolution of maritime continent thunderstorms under varying meteorological conditions over the Tiwi Islands. Int. J. Climatol., 21, 10211036.

    • Search Google Scholar
    • Export Citation
  • Chang, C. P., , Z. Wang, , J. Ju, , and T. Li, 2004: On the relationship between western Maritime Continent monsoon rainfall and ENSO during northern winter. J. Climate, 17, 665672.

    • Search Google Scholar
    • Export Citation
  • Chen, S. S., , and R. A. Houze Jr., 1997: Diurnal variation and life-cycle of deep convective systems over the tropical Pacific warm pool. Quart. J. Roy. Meteor. Soc., 123, 357388.

    • Search Google Scholar
    • Export Citation
  • Giorgi, F., , M. R. Marinucci, , and G. T. Bates, 1993: Development of a second-generation regional climate model (RegCM2). Part I: Boundary-layer and radiative transfer processes. Mon. Wea. Rev., 121, 27942813.

    • Search Google Scholar
    • Export Citation
  • Gray, W. M., , and R. W. Jacobson, 1977: Diurnal variation of deep cumulus convection. Mon. Wea. Rev., 105, 11711188.

  • Hamada, J. I., , M. D. Yamanaka, , J. Matsumoto, , S. Fukao, , P. A. Winarso, , and T. Sribimawati, 2002: Spatial and temporal variations of the rainy season over Indonesia and their link to ENSO. J. Meteor. Soc. Japan, 80, 285310.

    • Search Google Scholar
    • Export Citation
  • Hara, M., , T. Yoshikane, , H. G. Takahashi, , F. Kimura, , A. Noda, , and T. Tokioka, 2009: Assessment of the diurnal cycle of precipitation over the Maritime Continent simulated by a 20 km mesh GCM using TRMM PR data. J. Meteor. Soc. Japan, 87A, 413424.

    • Search Google Scholar
    • Export Citation
  • Hartigan, J. A., , and M. A. Wong, 1979: A K-means clustering algorithm. Appl. Stat., 28, 100108.

  • Hasegawa, T., , K. Ando, , K. Mizuno, , R. Lukas, , B. Taguchi, , and H. Sasaki, 2010: Coastal upwelling along the north coast of Papua New Guinea and El Niño events during 1981–2005. Ocean Dyn., 60, 12551269, doi:10.1007/s10236-010-0334-y.

    • Search Google Scholar
    • Export Citation
  • Haylock, M., , and J. McBride, 2001: Spatial coherence and predictability of Indonesian wet season rainfall. J. Climate, 14, 38823887.

  • Hendon, H. H., 2003: Indonesian rainfall variability: Impacts of ENSO and local air–sea interaction. J. Climate, 16, 17751790.

  • Hirose, M., , and K. Nakamura, 2005: Spatial and diurnal variation of precipitation systems over Asia observed by the TRMM precipitation radar. J. Geophys. Res., 110, D05106, doi:10.1029/2004JD004815.

    • Search Google Scholar
    • Export Citation
  • Hirose, M., , R. Oki, , S. Shimizu, , M. Kachi, , and T. Higashiuwatoko, 2008: Finescale diurnal rainfall statistics refined from eight years of TRMM PR data. J. Appl. Meteor. Climatol., 47, 544561.

    • Search Google Scholar
    • Export Citation
  • Houze, R. A., Jr., , S. G. Geotis, , F. D. Marks Jr., , and A. K. West, 1981: Winter monsoon convection in the vicinity of North Borneo. Part I: Structure and time variation of the clouds and precipitation. Mon. Wea. Rev., 109, 15951614.

    • Search Google Scholar
    • Export Citation
  • Hsu, H.-H., , and M.-Y. Lee, 2005: Topographic effects on the eastward propagation and initiation of the Madden–Julian oscillation. J. Climate, 18, 795809.

    • Search Google Scholar
    • Export Citation
  • Hu, Q., 2003: A multidecadal variation in summer season diurnal rainfall in the central United States. J. Climate, 16, 174178.

  • Huffman, G. J., and Coauthors, 2007: The TRMM multisatellite precipitation analysis (TMPA): Quasi-global, multiyear, combined-sensor precipitation estimates at fine scales. J. Hydrometeor., 8, 3855.

    • Search Google Scholar
    • Export Citation
  • Ichikawa, H., , and T. Yasunari, 2006: Time–space characteristics of diurnal rainfall over Borneo and surrounding oceans as observed by TRMM-PR. J. Climate, 19, 12381260.

    • Search Google Scholar
    • Export Citation
  • Ichikawa, H., , and T. Yasunari, 2007: Propagating diurnal disturbances embedded in the Madden–Julian oscillation. J. Geophys. Res., 34, L18811, doi:10.1029/2007GL030480.

    • Search Google Scholar
    • Export Citation
  • Ichikawa, H., , and T. Yasunari, 2008: Intraseasonal variability in diurnal rainfall over New Guinea and the surrounding oceans during austral summer. J. Climate, 21, 28522868.

    • Search Google Scholar
    • Export Citation
  • Johnson, R. H., , T. M. Rickenbach, , S. A. Rutledge, , P. E. Ciesielski, , and W. H. Schubert, 1999: Trimodal characteristics of tropical convection. J. Climate, 12, 23972418.

    • Search Google Scholar
    • Export Citation
  • Kikuchi, K., , and B. Wang, 2008: Diurnal precipitation regimes in the global tropics. J. Climate, 21, 26802696.

  • Kondragunta, C. R., , and A. Gruber, 1996: Seasonal and annual variability of the diurnal cycle of clouds. J. Geophys. Res., 101 (D16), 21 37721 390.

    • Search Google Scholar
    • Export Citation
  • Kraus, E. B., 1963: The diurnal precipitation change over the sea. J. Atmos. Sci., 20, 551556.

  • Kubota, H., , and T. Nitta, 2001: Diurnal variations of tropical convection observed during the TOGA-COARE. J. Meteor. Soc. Japan, 79, 815830.

    • Search Google Scholar
    • Export Citation
  • Kubota, H., , R. Shirooka, , J. I. Hamada, , and F. Syamsudin, 2011: Interannual rainfall variability over the eastern Maritime Continent. J. Meteor. Soc. Japan, 89A, 111122.

    • Search Google Scholar
    • Export Citation
  • Lau, W. K. M., , and D. E. Waliser, 2005: Intraseasonal Variability in the Atmosphere–Ocean Climate System. Praxis, 436 pp.

  • Lau, W. K. M., , H. T. Wu, , and S. Bony, 1997: The role of large-scale atmospheric circulation in the relationship between tropical convection and sea surface temperature. J. Climate, 10, 381392.

    • Search Google Scholar
    • Export Citation
  • Li, W., , C. Luo, , D. Wang, , and T. Lei, 2010: Diurnal variations of precipitation over the South China Sea. Meteor. Atmos. Phys., 109, 3346.

    • Search Google Scholar
    • Export Citation
  • Liberti, G. L., , F. Chéruy, , and M. Desbois, 2001: Land effect on the diurnal cycle of clouds over the TOGA COARE area, as observed from GMS IR data. Mon. Wea. Rev., 129, 15001517.

    • Search Google Scholar
    • Export Citation
  • Liebmann, B., , and C. A. Smith, 1996: Description of a complete (interpolated) outgoing longwave radiation dataset. Bull. Amer. Meteor. Soc., 77, 12751277.

    • Search Google Scholar
    • Export Citation
  • Love, B. S., , A. J. Matthews, , and G. M. S. Lister, 2011: The diurnal cycle of precipitation over the Maritime Continent in a high-resolution atmospheric model. Quart. J. Roy. Meteor. Soc., 137, 934947.

    • Search Google Scholar
    • Export Citation
  • Mapes, B. E., , T. T. Warner, , M. Xu, , and A. J. Negri, 2003: Diurnal patterns of rainfall in northwestern South America. Part I: Observations and context. Mon. Wea. Rev., 131, 799812.

    • Search Google Scholar
    • Export Citation
  • Massie, S. T., , J. Gille, , C. Craig, , R. Khosravi, , J. Barnett, , W. Read, , and D. Winker, 2010: HIRDLS and CALIPSO observations of tropical cirrus. J. Geophys. Res., 115, D00H11, doi:10.1029/2009JD012100.

    • Search Google Scholar
    • Export Citation
  • Milligan, G. W., , and M. C. Cooper, 1988: A study of standardization of variables in cluster analysis. J. Classif., 5, 181204.

  • Misra, V., 2010: Interaction of interannual and diurnal variations over equatorial Africa. J. Geophys. Res., 115, D01111, doi:10.1029/2009JD012512.

    • Search Google Scholar
    • Export Citation
  • Mori, S., and Coauthors, 2004: Diurnal land–sea rainfall peak migration over Sumatera Island, Indonesian Maritime Continent, observed by TRMM satellite and intensive rawinsonde soundings. Mon. Wea. Rev., 132, 20212039.

    • Search Google Scholar
    • Export Citation
  • Moron, V., , A. W. Robertson, , and J.-H. Qian, 2010: Local versus regional-scale characteristics of monsoon onset and post-onset rainfall over Indonesia. Climate Dyn., 34, 281299.

    • Search Google Scholar
    • Export Citation
  • Morrissey, M. L., 1986: A statistical analysis of the relationships among rainfall, outgoing longwave radiation and the moisture budget during January–March 1979. Mon. Wea. Rev., 114, 931942.

    • Search Google Scholar
    • Export Citation
  • Murakami, M., 1983: Analysis of the deep convective activity over the western Pacific and Southeast Asia. Part I: Diurnal cycle. J. Meteor. Soc. Japan, 61, 6075.

    • Search Google Scholar
    • Export Citation
  • Neale, R., , and J. Slingo, 2003: The Maritime Continent and its role in the global climate: A GCM study. J. Climate, 16, 834848.

  • Negri, A. J., , T. L. Bell, , and L. Xu, 2002: Sampling of the diurnal cycle of precipitation using TRMM. J. Atmos. Oceanic Technol., 19, 13331344.

    • Search Google Scholar
    • Export Citation
  • Nitta, T., , and S. Sekine, 1994: Diurnal variation of convective activity over the tropical western Pacific. J. Meteor. Soc. Japan, 72, 627641.

    • Search Google Scholar
    • Export Citation
  • Ohsawa, T., , H. Ueda, , T. Hayashi, , A. Watanabe, , and J. Matsumoto, 2001: Diurnal variations of convective activity and rainfall in tropical Asia. J. Meteor. Soc. Japan, 79, 333352.

    • Search Google Scholar
    • Export Citation
  • Oki, T., , and K. Musiake, 1994: Seasonal change of the diurnal cycle of precipitation over Japan and Malaysia. J. Appl. Meteor., 33, 14451463.

    • Search Google Scholar
    • Export Citation
  • Poveda, G., and Coauthors, 2005: The diurnal cycle of precipitation in the tropical Andes of Colombia. Mon. Wea. Rev., 133, 228240.

  • Qian, J.-H., 2008: Why precipitation is mostly concentrated over islands in the Maritime Continent. J. Atmos. Sci., 65, 14281441.

  • Qian, J.-H., , A. W. Robertson, , and V. Moron, 2010: Interactions among ENSO, the monsoon, and diurnal cycle in rainfall variability over Java, Indonesia. J. Atmos. Sci., 67, 35093524.

    • Search Google Scholar
    • Export Citation
  • Ramage, C. S., 1968: Role of a tropical “maritime continent” in the atmospheric circulation. Mon. Wea. Rev., 96, 365370.

  • Randel, W. J., , F. Wu, , and W. R. Ríos, 2002: Thermal variability of the tropical tropopause region derived from GPS/MET observations. J. Geophys. Res., 108, 4024, doi:10.1029/2002JD002595.159.

    • Search Google Scholar
    • Export Citation
  • Rauniyar, S. P., , and K. J. E. Walsh, 2011: Scale interaction of the diurnal cycle of rainfall over the Maritime Continent and Australia: Influence of the MJO. J. Climate, 24, 325348.

    • 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.

    • Search Google Scholar
    • Export Citation
  • Ropelewski, C. F., , and M. S. Halpert, 1987: Global and regional scale precipitation patterns associated with the El Niño/Southern Oscillation. Mon. Wea. Rev., 115, 16061626.

    • Search Google Scholar
    • Export Citation
  • Sakurai, N., and Coauthors, 2005: Diurnal cycle of cloud system migration over Sumatera Island. J. Meteor. Soc. Japan, 83, 835850.

  • Sato, T., , H. Miura, , M. Satoh, , Y. N. Takayabu, , and Y. Wang, 2009: Diurnal cycle of precipitation in the tropics simulated in a global cloud-resolving model. J. Climate, 22, 48094826.

    • Search Google Scholar
    • Export Citation
  • Sen Roy, S., , and R. C. Balling, 2004: Analysis of Hawaiian diurnal rainfall patterns. Theor. Appl. Climatol., 79, 209214.

  • Slingo, J., , P. Inness, , R. Neale, , S. Woolnough, , and G. Y. Yang, 2003: Scale interactions on diurnal to seasonal timescales and their relevance to model systematic errors. Ann. Geophys., 46, 139155.

    • Search Google Scholar
    • Export Citation
  • Sorooshian, S., , X. Gao, , K. Hsu, , R. A. Maddox, , Y. Hong, , H. V. Gupta, , and B. Imam, 2002: Diurnal variability of tropical rainfall retrieved from combined GOES and TRMM satellite information. J. Climate, 15, 9831001.

    • Search Google Scholar
    • Export Citation
  • Steinley, D., 2004: Standardizing variables in K-means clustering. Classification, Clustering, and Data Mining Applications, D. Banks et al., Eds., Springer, 53–60.

  • Steinley, D., 2006: K-means clustering: A half-century synthesis. Br. J. Math. Stat. Psychol., 59, 134.

  • Sui, C.-H., , K.-M. Lau, , Y. N. Takayabu, , and D. A. Short, 1997: Diurnal variations in tropical oceanic cumulus convection during TOGA COARE. J. Atmos. Sci., 54, 639655.

    • Search Google Scholar
    • Export Citation
  • Tangang, F. T., , and L. Juneng, 2004: Mechanisms of Malaysian rainfall anomalies. J. Climate, 17, 36163622.

  • Trenberth, K. E., 1997: The definition of El Niño. Bull. Amer. Meteor. Soc., 78, 27712777.

  • Varikoden, H., , A. A. Samah, , and C. A. Babu, 2010: Spatial and temporal characteristics of rain intensity in the peninsular Malaysia using TRMM rain rate. J. Hydrol., 387, 312319.

    • Search Google Scholar
    • Export Citation
  • Waliser, D. E., 2002: Tropical meteorology: Intertropical convergence zones (ITCZ). Encyclopedia of Atmospheric Sciences, J. Holton, J. Pyle, and J. Curry, Eds., Academic Press, 2325–2334.

  • Wallace, J. M., 1975: Diurnal variations in precipitation and thunderstorm frequency over the conterminous United States. Mon. Wea. Rev., 103, 406419.

    • Search Google Scholar
    • Export Citation
  • Wheeler, M. C., , H. H. Hendon, , S. Cleland, , H. Meinke, , and A. Donald, 2009: Impacts of the Madden–Julian oscillation on Australian rainfall and circulation. J. Climate, 22, 14821498.

    • Search Google Scholar
    • Export Citation
  • Wu, P., , M. Hara, , J. Hamada, , M. D. Yamanaka, , and F. Kimura, 2009: Why a large amount of rain falls over the sea in the vicinity of western Sumatra Island during nighttime. J. Appl. Meteor. Climatol., 48, 13451361.

    • Search Google Scholar
    • Export Citation
  • Yamamoto, M. K., , F. A. Furuzawa, , A. Higuchi, , and K. Nakamura, 2008: Comparison of diurnal variations in precipitation systems observed by TRMM PR, TMI, and VIRS. J. Climate, 21, 40114028.

    • Search Google Scholar
    • Export Citation
  • Yang, G.-Y., , and J. Slingo, 2001: The diurnal cycle in the tropics. Mon. Wea. Rev., 129, 784801.

  • Zhang, C., 2005: Madden–Julian oscillation. Rev. Geophys., 43, RG2003, doi:10.1029/2004RG000158.

  • Zhou, L., , and Y. Wang, 2006: Tropical Rainfall Measuring Mission observation and regional model study of precipitation diurnal cycle in the New Guinean region. J. Geophys. Res., 111, D17104, doi:10.1029/2006JD007243.

    • Search Google Scholar
    • Export Citation
  • Zuidema, P., 2003: Convective clouds over the Bay of Bengal. Mon. Wea. Rev., 131, 780798.

  • View in gallery

    (a) Location map of study area showing topography and bathymetrical features in shading. Locations referred to in the text are shown in italic font for sea regions and in normal font for land and islands. (b) The climatological mean rainfall (mm day−1) composited from 12 (1998–2010) DJF seasons using 3B42 data.

  • View in gallery

    Rainfall anomalies (mm day−1) for the two ENSO phases, (a) La Niña and (b) El Niño, using 3B42 data. (c) The warm anomaly (El Niño La Niña) of rainfall. The values shown in warm anomaly plot are statistically significant at 95% confidence level according to bootstrap tests. Each phase of ENSO consists of four DJF seasons.

  • View in gallery

    OLR anomalies (W m−2) for the two ENSO phases: (a) La Niña and (b) El Niño derived from National Centers for Environmental prediction (NCEP)–Department of Energy (DOE) Reanalysis II data for 12 (1998–2010) DJF seasons. The dark (light) shade shows regions of positive (negative) OLR anomalies that are statistically significant at 95% confidence level according to the two-tailed Student’s t test. Contour interval is 5 W m−2 and 0 lines are removed.

  • View in gallery

    (left) Frequency of the climatological mean OLR < 170 W m−2 (here interpreted as tropical cirrus–anvil clouds) and anomalies corresponding to the two ENSO phases. (right) As at left, but for OLR values between 170 and 240 W m−2 (convective clouds). Data are for (a),(d) mean OLR, (b),(e) the La Niña phase anomaly, and (c),(f) the El Niño phase.

  • View in gallery

    As in Fig. 3, but for SST (°C), contour interval is 0.25°C and 0 lines are removed; and 850 hPa winds (m s−1).

  • View in gallery

    (a) Gridpoint simultaneous local correlation between the monthly rainfall anomaly and the monthly SSTA. (b) The simultaneous correlations of the monthly rainfall anomaly with the ENSO index (ONI). A contour at 750 m is overlaid in black over the islands.

  • View in gallery

    The warm anomaly (El Niño La Niña) for (a) the moisture divergence ( · qV; 10−8 s−1) at 850 hPa, (b) vertical velocity (ω; Pa s−1) at 500 hPa, and (c) midtropospheric (average of 700–400 hPa) relative humidity (%). The negative–positive values (dark–light shading) represent convergence–divergence, upward–downward motion, and moist–drier atmosphere in (a)–(c), respectively. Areas above the 95% significance level are shaded.

  • View in gallery

    (left) The spatial pattern of the six clusters from the k-means analysis of the climatological average (12 DJFs) diurnal cycle of rainfall composited from TRMM (a) 3B42, (b) 3G68-TMI, and (c) 3G68-PR data. (right) The corresponding diurnal cycles of the six clusters. The percentage of total area covered by each cluster is shown in the parentheses at the bottom of the maps. The contour at 750 m is overlaid in black over the islands.

  • View in gallery

    (a) The spatial pattern of rainfall of the six k-mean clusters and (b) the clusters’ diurnal cycle anomalies from the k-means analysis of the diurnal cycle of rainfall (3B42) composited for the La Niña phase. (c),(d) As in (a),(b), but for the El Niño phase. Also shown is the percentage of total area covered by a cluster in the parentheses at the bottom of the maps: (e) the amplitude of the first harmonic of the diurnal cycle for the warm anomaly and (f) its phase. The contour at 750 m is overlaid in black over the islands.

  • View in gallery

    Composite Hovmöller diagram of 3B42 rainfall averaged for (first row) La Niña and (second row) El Niño along (fourth row) rectangular domains over Sumatra, Borneo, and New Guinea. (third row) The average elevation along the cross section (dotted line) and the daily accumulated rainfall corresponding to the La Niña (El Niño) phase [line with circles (crosses)]. Longitudes of top two rows match that of the third row. Rainfall rates >0.2 mm h−1 are shaded and the contour interval is 0.2 mm h−1.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 112 112 20
PDF Downloads 82 82 14

Influence of ENSO on the Diurnal Cycle of Rainfall over the Maritime Continent and Australia

View More View Less
  • 1 School of Earth Sciences, University of Melbourne, Melbourne, Victoria, Australia
© Get Permissions
Full access

Abstract

This study examines the influence of ENSO on the diurnal cycle of rainfall during boreal winter for the period 1998–2010 over the Maritime Continent (MC) and Australia using Tropical Rainfall Measuring Mission (TRMM) and reanalysis data. The diurnal cycles are composited for the ENSO cold (La Niña) and warm (El Niño) phases. The k-means clustering technique is then applied to group the TRMM data into six clusters, each with a distinct diurnal cycle. Despite the alternating patterns of widespread large-scale subsidence and ascent associated with the Walker circulation, which dominates the climate over the MC during the opposing phases of ENSO, many of the islands of the MC show localized differences in rainfall anomalies that depend on the local geography and orography. While ocean regions mostly experience positive rainfall anomalies during La Niña, some local regions over the islands have more rainfall during El Niño. These local features are also associated with anomalies in the amplitude and characteristics of the diurnal cycle in these regions. These differences are also well depicted in large-scale dynamical fields derived from the interim ECMWF Re-Analysis (ERA-Interim).

Corresponding author address: Surendra Rauniyar, School of Earth Sciences, University of Melbourne, Melbourne VIC 3010, Australia. E-mail: s.rauniyar@student.unimelb.edu.au

Abstract

This study examines the influence of ENSO on the diurnal cycle of rainfall during boreal winter for the period 1998–2010 over the Maritime Continent (MC) and Australia using Tropical Rainfall Measuring Mission (TRMM) and reanalysis data. The diurnal cycles are composited for the ENSO cold (La Niña) and warm (El Niño) phases. The k-means clustering technique is then applied to group the TRMM data into six clusters, each with a distinct diurnal cycle. Despite the alternating patterns of widespread large-scale subsidence and ascent associated with the Walker circulation, which dominates the climate over the MC during the opposing phases of ENSO, many of the islands of the MC show localized differences in rainfall anomalies that depend on the local geography and orography. While ocean regions mostly experience positive rainfall anomalies during La Niña, some local regions over the islands have more rainfall during El Niño. These local features are also associated with anomalies in the amplitude and characteristics of the diurnal cycle in these regions. These differences are also well depicted in large-scale dynamical fields derived from the interim ECMWF Re-Analysis (ERA-Interim).

Corresponding author address: Surendra Rauniyar, School of Earth Sciences, University of Melbourne, Melbourne VIC 3010, Australia. E-mail: s.rauniyar@student.unimelb.edu.au

1. Introduction

The Maritime Continent (MC), which includes the northern portion of Australia and consists of a complex organization of islands with varying size and orography, is an interesting region for exploring multiscale interactions (Ramage 1968; Neale and Slingo 2003; Slingo et al. 2003; Rauniyar and Walsh 2011). The climate of the MC is strongly influenced by the Indo-Pacific circulations because of its unique geographical location near the equator surrounded by shallow seas, including the warmest open ocean waters in the world (Sakurai et al. 2005; Aldrian et al. 2007). A significant amount of uptake and release of latent heat is associated with convection over this region, which influences the energy budget and the diabatic heating of the entire atmosphere (Neale and Slingo 2003; Mori et al. 2004). This region not only shows a strong diurnal cycle in rainfall and convection (Yang and Slingo 2001; Sorooshian et al. 2002; Mori et al. 2004; Kikuchi and Wang 2008), but substantial intraseasonal (Zhang 2005; Lau and Waliser 2005; Wheeler et al. 2009), seasonal, and interannual variations also occur (Ropelewski and Halpert 1987; Hendon 2003; Aldrian and Dwi Susanto 2003; Chang et al. 2004). Many studies have shown a noticeable interaction between the strong diurnal cycle over the MC and the large-scale circulations (Chen and Houze 1997; Johnson et al. 1999; Slingo et al. 2003; Hsu and Lee 2005; Ichikawa and Yasunari 2007; Li et al. 2010; Qian et al. 2010; Rauniyar and Walsh 2011). Hence it is argued that understanding the mechanisms of the diurnal cycle of rainfall and convection over this region is also important in the study of variations in large-scale circulations.

The characteristics of the diurnal cycle of total rainfall and convection over the MC as a whole, over its individual islands, and over northern Australia have been studied by analysis of various forms of observations (Houze et al. 1981; Murakami 1983; Nitta and Sekine 1994; Beringer et al. 2001; Liberti et al. 2001; Ohsawa et al. 2001; Yang and Slingo 2001; Mori et al. 2004; Sakurai et al. 2005; Ichikawa and Yasunari 2006, 2008; Kikuchi and Wang 2008) and also by using numerical models (Neale and Slingo 2003; Zhou and Wang 2006; Qian 2008; Hara et al. 2009; Wu et al. 2009; Sato et al. 2009; Love et al. 2011). These studies show distinct contrasts in the general features of the diurnal cycle among inland, coastal, and adjacent sea regions of the MC. In general, the maximum rainfall occurs during the late afternoon and early evening over land, but during the morning over the surrounding oceans (Murakami 1983; Nitta and Sekine 1994; Yang and Slingo 2001; Mori et al. 2004). More specifically, the peak in the diurnal cycle of rainfall over the smaller islands (e.g., Java, Sulawesi, New Britain, etc.) and also over the “top end” or northernmost portion of Australia occurs during the late afternoon (Oki and Musiake 1994; Qian 2008; Varikoden et al. 2010; Rauniyar and Walsh 2011), whereas over the inland regions of larger islands (e.g., Borneo, Sumatra, New Guinea) it is delayed until midnight (Mori et al. 2004; Zhou and Wang 2006; Ichikawa and Yasunari 2006, 2008; Wu et al. 2009; Rauniyar and Walsh 2011). In addition, the strong diurnal signal over land spreads out several hundred kilometers over adjacent oceans as a gravity wave (Yang and Slingo 2001). Mori et al. (2004), using Tropical Rainfall Measuring Mission (TRMM) precipitation radar (PR) data, found both onshore and offshore propagation of the diurnal signal during daytime and nighttime from the southwestern coastline of Sumatra. They suggested that the eastward propagation of rainfall until late evening might be due to advection by the background westerly wind in the lower troposphere. In contrast, the offshore propagation during nighttime might arise from the generation of new convection by self-replicating and gravity wave mechanisms. Similarly, Liberti et al. (2001) showed that northward propagation of cloud systems toward the open ocean from the northern coast of New Guinea arises from low-level convergence between the large-scale flow and a land breeze. Recent studies show the extent and direction of this diurnal propagation, as well as other features such as its amplitude and phase, are significantly modulated by the intraseasonal oscillation (Ichikawa and Yasunari 2006, 2008; Rauniyar and Walsh 2011). However, less attention has been paid to analyzing the influence of the El Niño–Southern Oscillation (ENSO) phenomenon on the diurnal cycle of rainfall over the MC and northern Australia, due to the limited availability of high-resolution rainfall datasets.

The influence of ENSO on annual and seasonal rainfall variability over the MC region has been studied by many researchers (Haylock and McBride 2001; Hamada et al. 2002; Hendon 2003; Aldrian and Dwi Susanto 2003; Chang et al. 2004). In general, drought conditions prevail over the Indonesian region during the warm (e.g., El Niño) phase of ENSO due to large-scale subsidence associated with an eastward shift of convection into the central Pacific. By simple averaging of rainfall from 43 stations throughout the islands of the MC, Hendon (2003) showed that this inverse relationship is stronger and spatially more coherent and therefore more predictable during the dry half (June–November) rather than during the wet half (December–May) of the year. In contrast, Tangang and Juneng (2004) found a strengthening of the ENSO–Malaysia rainfall relationship during boreal winter. Aldrian and Dwi Susanto (2003) argued that given the large size of the Indonesian archipelago, explaining the impact of ENSO over the entire region by associating it with a single rainfall index disregards the wide range of rainfall variability. Hence, Aldrian and Dwi Susanto (2003) divide an area within 15°S–8°N, 90°–140°E into three climatic regions with distinct characteristics using a double correlation method (see Fig. 1a for a location map). They found that the region encompassing an area south of Halmahera and northern Sulawesi is most strongly influenced by ENSO, where negative rainfall anomalies occur during the summer and autumn [June–November (JJASON)] seasons of El Niño years. In contrast, another region consisting of an area from northern Sumatra to northwestern Borneo does not experience any ENSO impact. Moreover, their results showed that the region located in southern Indonesia, including an area encompassing southern Sumatra to Timor, southern Borneo, Sulawesi, and part of western New Guinea, is somewhat less influenced by ENSO. Chang et al. (2004) found positive (negative) correlation between Niño-3 sea surface temperature (SST) and area-averaged Indonesian station rainfall west (east) of 112°E during boreal winter and suggested that the low correlation between all-Indonesian rainfall and ENSO may be due to the averaging of these opposite polarities.

Fig. 1.
Fig. 1.

(a) Location map of study area showing topography and bathymetrical features in shading. Locations referred to in the text are shown in italic font for sea regions and in normal font for land and islands. (b) The climatological mean rainfall (mm day−1) composited from 12 (1998–2010) DJF seasons using 3B42 data.

Citation: Journal of Climate 26, 4; 10.1175/JCLI-D-12-00124.1

A number of studies have investigated the influence of ENSO on the diurnal cycle of cloud (Kondragunta and Gruber 1996) and rainfall (e.g., Hu 2003; Sen Roy and Balling 2004; Poveda et al. 2005; Misra 2010; Li et al. 2010; Qian et al. 2010) at various locations worldwide. For the region close to the MC, Li et al. (2010) found that both the amplitude and phase of the diurnal cycle are modulated by ENSO over the South China Sea (SCS). In particular, early morning rainfall is much more significant during La Niña years over the ocean area while the amplitude of the diurnal cycle is much higher during El Niño years over the offshore area of Borneo Island. Through a modeling experiment, Qian et al. (2010) found that over the mountains of Java Island above (below) than normal rainfall occurs during El Niño (La Niña) years during the December–February (DJF) season, due to more (less) frequent occurrence of a quiescent monsoon weather type, which amplifies (reduces) the local diurnal cycle.

The above studies have advanced our understanding of the impact of ENSO on the diurnal cycle of rainfall and tropical convection. Nevertheless, these studies were mainly based on a few ENSO events, which can differ significantly from event to event. The metrics used to partition the ENSO phases were also not consistent from study to study. In addition, in previous studies the analysis of the impact of ENSO on the diurnal cycle of rainfall over the MC and northern Australia has been limited to a particular location such as the SCS (Li et al. 2010) and Java (Qian et al. 2010) only. These issues necessitate preparation of a climatology of ENSO impact on the diurnal cycle over a wider region (Fig. 1a), which includes the MC along with tropical and subtropical regions of Australia, using a high spatial and temporal resolution rainfall dataset. The objectives of the present study are to understand better what causes the diurnal cycle variability over the study domain with a focus on different phases of ENSO during boreal winter. This study also intends to provide a foundation for the validation of future results from regional-scale numerical models and high-resolution climate or cloud-resolving models in the above region. The paper is organized as follows: a brief description of data and methodology is shown in section 2. The influence of ENSO on rainfall and other atmospheric variables is described in section 3. The spatial and temporal variations in the diurnal cycle of rainfall under the influence of ENSO are explained in section 4. Section 5 summarizes the major findings of this study, discusses possible explanations for the observed results, and suggests future work.

2. Data and methods

a. TRMM satellite data

The present study utilizes two TRMM rainfall products, namely 3G68_V6 and 3B42_V6 (Huffman et al. 2007; a description of the 3B42 algorithm is provided at http://trmm.gsfc.nasa.gov/3b42.html) for 12 full Australian summer seasons (DJF, 1998–2010). The 3G68_V6 dataset is an hourly gridded product containing TRMM instrumental near surface rain estimates at 0.25° × 0.25° horizontal resolution. It consists of total pixels, rainy pixels, mean rain rate (mm h−1) and the percentage of rainfall calculated to be convective from the 2A12 [TRMM Microwave Imager (TMI)], 2A25 (PR), and 2B31 (TMI-PR combined) algorithms merged into a single daily file. The TMI data have a systematic error over coastal areas in which both land and ocean regions are included in a single pixel (Mori et al. 2004). The PR sensor covers only a small region due to its narrower swath, but it does provide a better detectability of rainfall in this region. Therefore, both the PR and TMI datasets are analyzed in this study. The 3B42 data contains estimated rain rate (mm h−1) created by calibrating the IR brightness temperatures to the high-quality microwave estimates. The 3B42 rainfall product has 3-h temporal resolution with 0.25° × 0.25° spatial resolution globally, extending from 50°S to 50°N latitude and available from 1998 to the present. The diurnal cycle of rainfall is composited at each grid point (local solar time) for both datasets corresponding to each of the two phases of ENSO (see section 2c for ENSO phase definition). A 4-h running mean is applied to the diurnal cycle composited from the 3G68 data to smooth the temporal variation that would otherwise give a checkerboard pattern in the sampling due to the narrow swath of the satellite measurements (Negri et al. 2002). The applied smoothing results in an even distribution of samples at each local time (Hirose and Nakamura 2005; Hirose et al. 2008).

b. Large-scale dataset

The daily optimum interpolated SST on 0.25° spatial resolution (Reynolds et al. 2007) and National Oceanic and Atmospheric Administration (NOAA) interpolated outgoing longwave radiation (OLR) data at horizontal resolution of 2.5° × 2.5° (Liebmann and Smith 1996) are used to understand the relationship of rainfall variability with SST and OLR, respectively. In addition, air temperature, wind (U, V), vertical velocity (omega), and specific and relative humidity at different pressure levels were extracted from the interim European Centre for Medium-Range Weather Forecasts (ECMWF) Reanalysis (ERA-Interim) dataset available at a horizontal resolution of 1.5° × 1.5°. These data are used to explain the links between large-scale dynamic fields and rainfall variations during ENSO phases.

c. Partition into ENSO phases

There are many indices (e.g., Niño-3, Niño-3.4, and Niño-4) developed to categorize ENSO into cold (La Niña) and warm (El Niño) phases when area-averaged SST anomalies (SSTAs) in the eastern Pacific Ocean exceed a given threshold over a span of several months (Trenberth 1997). In this study, we used the Oceanic Niño Index (ONI; see http://www.cpc.ncep.noaa.gov/) to define ENSO phases, defined as a 3-month running mean of SST anomalies (SSTAs) in the Niño-3.4 region (5°S–5°N, 170°–120°W). The warm (cold) phase occurs when the ONI is above (below) +0.5°C (−0.5°C) for a minimum of five consecutive overlapping 3-month seasons [e.g., DJF, January–March (JFM), etc.]. Based upon the above definition, the four DJF seasons of the years 1998/99, 1999/00, 2000/01, and 2007/08 are in the La Niña phase while the four DJF seasons corresponding to the El Niño phase are 2002/03, 2004/05, 2006/07, and 2009/10. (Note that for 2006/07, February of this season has an SSTA of 0.4°C; the result of the subsequent analysis does not change even after excluding this DJF season.) Although the study period includes a strong long-lasting La Niña (1998–2000), and while all four El Niño events are modest ones, the results of this study are comparable with previous several studies in this region, which have used many long-lasting strong El Niño events. Therefore, we believe that separating La Niña and El Niño events is still meaningful based on the limited temporal availability of TRMM rainfall data as it provides finer-scale details of rainfall variation than in previous studies.

d. Methodology

In the present study, the k-means cluster algorithm of Hartigan and Wong (1979) from the stats package (kmean) from the R project (http://www.R-project.org) is used to group the pixels with similar diurnal rainfall patterns into clusters, thus better enabling relationships to be established between the diurnal cycle and physical factors such as land–sea contrast. Although the k-means algorithm used varies between studies (Steinley 2006), nonetheless all algorithms iteratively search for a desired number (k) of cluster centroids (for our case, those representing a specific diurnal pattern) by partitioning the pixels into k groups such that the sum of squares from points to the assigned cluster centers is a minimum (i.e., each pixel is assigned to the cluster whose centroid diurnal pattern is similar to it). There is a debate over the method of standardization (i.e., the process of transforming normal variates to a dimensionless standard score form—an example being the z score, which has a mean = 0 and a standard deviation = 1) when performing k-means clustering [for different types of standardization, see Milligan and Cooper (1988) and Steinley (2004)]. To determine an appropriate value of k and a proper standardization method, a within-group sum of squares (WSS), here an indicator of the within-cluster variation or noise, is calculated for different values of k by performing k-means clustering on nonstandardized and standardized hourly diurnal cycles. In our case, the WSS is a minimum for nonstandardized data and a maximum for the data standardized to the z score. This is consistent with the conclusion of Milligan and Cooper (1988) and Steinley (2004), who found that traditional z-score transformation has a larger WSS and is thus generally less effective. Furthermore, plotting the WSS against the number of clusters (not shown) suggests that based on their different WSS values, there are at least three distinct clusters within the data. However, the optimum number of clusters is selected by further evaluating each cluster against two criteria. To be an independent centroid, either the amplitude (maximum rainfall) or the phase (the time of maximum rainfall) of that cluster should be statistically different at the 95% confidence level from the rest of the clusters. This leads to total of six clusters, which is adopted for this study. This higher number of clusters also gives a better geographical discrimination of the variations in the diurnal cycle.

3. ENSO-induced large-scale anomalies

a. ENSO-induced rainfall anomalies

The rainfall climatology (average of 12 DJF seasons) derived from the 3B42 product exhibits substantial geographical variation (Fig. 1b). In general, the highest rainfall rates (>12 mm day−1) are located over the islands of the MC and the northernmost part (top end) of Australia. Similarly, localized maxima in rainfall appear over the sea in the vicinity of southwestern Sumatra, over the eastern Java Sea between Java, Sulawesi, and Borneo and over the Bismarck Sea. These features are modulated by ENSO, which is shown in the ENSO-related rainfall anomaly maps (Fig. 2). During the La Niña phase (Fig. 2a), rainfall heavier than climatology covers wide areas of the tropical seas on both sides of the equator, except near the northeastern edge of our study domain, near the warm pool (hereafter NEWP) within 5°N–5°S, 150°–160°E, where lower than average rainfall occurs. In contrast, the reverse occurs in the El Niño phase (Fig. 2b). Similarly, large regions of Australia experience dry condition during El Niño, but rainfall above climatology often occurs during La Niña. Although adjacent coastal seas surrounding the top end of Australia receive more (less) rainfall during La Niña (El Niño), the effects of ENSO phases in DJF are minimal on inland areas near Darwin. These results agree with the many previous studies that reported that during a warm ENSO event, low rainfall covers wide areas of the MC as the locus of deep convection and rainfall associated with the rising branch of the Walker circulation shifts eastward toward the central and eastern equatorial Pacific. In contrast, during cold ENSO events, the anomalous Walker circulation exhibits large-scale, low-level convergence over the MC, resulting in widespread heavier rainfall.

Fig. 2.
Fig. 2.

Rainfall anomalies (mm day−1) for the two ENSO phases, (a) La Niña and (b) El Niño, using 3B42 data. (c) The warm anomaly (El Niño La Niña) of rainfall. The values shown in warm anomaly plot are statistically significant at 95% confidence level according to bootstrap tests. Each phase of ENSO consists of four DJF seasons.

Citation: Journal of Climate 26, 4; 10.1175/JCLI-D-12-00124.1

On the other hand, despite this large-scale subsidence or ascent over the MC during the different phases of ENSO, many of the islands of the MC show a localized north–south disparity in rainfall anomaly during boreal winter (Figs. 2a,b). To highlight those areas, a warm anomaly (El Niño La Niña) composite of rainfall is shown in Fig. 2c. The values shown are statistically significant above the 95% confidence level according to bootstrap test. The result shows that more rainfall occurs over the smaller land regions located in the Northern Hemisphere (e.g., Malay Peninsula, Mindanao, Halmahera, etc.) during La Niña whereas less occurs during El Niño. In contrast, over most of the smaller land regions located in the Southern Hemisphere (e.g., Java, Sulawesi, New Britain, Timor, etc.) higher rainfall occurs during El Niño than in La Niña. Similar features of interannual rainfall variability, for off-equatorial islands located in the eastern MC, have been recently reported by Kubota et al. (2011).

Furthermore, most of the larger islands (e.g., Sumatra, Borneo, and New Guinea) exhibit a dipole structure in rainfall distribution during the phases of ENSO. These dipole structures are clearly noticeable over Borneo and New Guinea but are weaker over Sumatra and the smaller islands of Java and Sulawesi. For example, along the coastal mountains of Sumatra and over its central northern plains, more rainfall occurs during El Niño than in La Niña. Similarly, the southwestern coast of western Java shows more rainfall during El Niño than in La Niña. Over Borneo, significantly higher rainfall occurs over the entire southern plains during El Niño than in La Niña whereas the opposite occurs over the northern half of the island. Interestingly, the western coast of Borneo, where highest rainfall occurs (Fig. 1b), shows no remarkable differences in rainfall between the ENSO phases. The regions of positive rainfall anomalies during El Niño over the western MC are similar to those for the Sumatra and Malay Peninsula (SMP) area defined by Chang et al. (2004), who observed a notable positive correlation between Niño-3 SST and rainfall over SMP. Over New Guinea, where the central mountains are elongated in the northwest–southeast direction, there exists a contrast in rainfall between coastal regions, plains (including foothills, defined here as regions of topographic height less than 500 m, adjacent to higher mountains) and mountains during the phases of ENSO. Over both the north and south onshore coasts of this island, negative (positive) rainfall anomalies occur during the DJF season of La Niña (El Niño) years. In contrast, over the southern foothills of New Guinea, positive (negative) rainfall anomalies dominate during La Niña (El Niño). However, over the mountains themselves the difference in rainfall between La Niña and El Niño phases is not significant. These results are consistent in the 3G68 product (not shown) except that the spatial pattern is noisier than that derived from 3B42.

Aspects of this geographical pattern of anomalies have also been noted by others. Qian et al. (2010) has noticed a similar dipole pattern over Java during the peak rainy season (DJF) associated with the El Niño phase in both observations and model output. They found a positive rainfall anomaly (more in El Niño than La Niña) in the mountainous south versus a negative rainfall anomaly over the northern plains. Using the high-resolution second-generation regional climate model (RegCM3; Giorgi et al. 1993), they showed that this spatial heterogeneity in rainfall during DJF seasons of El Niño is related to the increased frequency of weaker monsoonal-type winds. This quiescent regime (Moron et al. 2010) provides less hindrance to the thermally driven diurnal cycle and hence less interference with the convergence of enhanced land–sea breezes and mountain–valley winds, thus producing more rainfall over mountainous regions. Qian et al. (2010) argued that the above explanation might also be valid for other islands with similar climate and topography, such as our “smaller islands” Sulawesi and Timor. However, for the larger islands of the MC (e.g., Sumatra, Borneo, and New Guinea), which have a more complex structure and orientation of their topography, the above arguments have limitations in explaining the observed inhomogeneity in rainfall distribution. Hence, more detailed investigations are required over those islands through climatological and numerical studies.

Previous studies (e.g., Chang et al. 2004; Ichikawa and Yasunari 2006, 2008; Qian 2008; Moron et al. 2010; Rauniyar and Walsh 2011) have shown that the large-scale circulation along with the spatial asymmetry of mountains plays a crucial role in such north–south disparity in rainfall distribution. Furthermore, many studies (Morrissey 1986; Chang et al. 2004; Ichikawa and Yasunari 2008; Rauniyar and Walsh 2011; Kubota et al. 2011) have also shown that mesoscale dynamic and thermodynamic features estimated using coarse grid resolution reanalysis products are capable of explaining the observed spatiotemporal variations in rainfall to some extent over the most of the larger islands of the MC. However, over the smaller islands the reanalyses are too coarse to capture these processes. In the following subsections, we have examined how well large-scale features are able to represent the rainfall heterogeneity during the ENSO phases.

b. ENSO-induced OLR anomalies

OLR is treated as a proxy for rainfall caused by deep convection in several studies (Waliser 2002; Randel et al. 2002; Massie et al. 2010). Lower values of OLR are mostly confined between the latitudinal band of 10°S–5°N and the centers of these deep convection regions are mostly collocated with the general regions of heavy rainfall [see Fig. 2a of Rauniyar and Walsh (2011)]. The locations of deep convection are greatly modified by the phases of ENSO, as shown in Fig. 3. The large negative anomaly (<10 W m−2) in OLR during La Niña (Fig. 3a) is collocated with regions of high positive rainfall anomaly. In contrast, the El Niño phase is associated with widespread suppression of convection over the entire study domain except over the NEWP region (Fig. 3b). However, the use of OLR values at all atmospheric heights as shown in Fig. 3 may not be the best representation of deep convection in this region. Given the coarse resolution of OLR, it is highly likely that during the El Niño phase, patches of cold OLR from isolated deep convection (cumulonimbi) over parts of the islands would be surrounded by warm OLR from the surface, which would then average together into an intermediate OLR value in the composite mean. Furthermore, during La Niña, lower OLR values are observed over the parts of islands where negative rainfall anomalies exist. This suggests that the absence of a positive rainfall anomaly over the islands during La Niña is related to the higher frequency of nonprecipitating high clouds with low OLR (e.g., cirrus), rather than precipitating clouds. Therefore, we examined the frequency of two levels of clouds as described below.

Fig. 3.
Fig. 3.

OLR anomalies (W m−2) for the two ENSO phases: (a) La Niña and (b) El Niño derived from National Centers for Environmental prediction (NCEP)–Department of Energy (DOE) Reanalysis II data for 12 (1998–2010) DJF seasons. The dark (light) shade shows regions of positive (negative) OLR anomalies that are statistically significant at 95% confidence level according to the two-tailed Student’s t test. Contour interval is 5 W m−2 and 0 lines are removed.

Citation: Journal of Climate 26, 4; 10.1175/JCLI-D-12-00124.1

Figure 4 shows the climatological mean frequency and anomalies during the ENSO phases for very cold OLRs (OLR < 170 W m−2) and intermediate OLRs (170 W m−2 < OLR < 240 W m−2), respectively. The climatological mean of 12 DJF seasons (Figs. 4a,d) shows that on average the Australian monsoonal area (over the Australian continent north of 18°S) is covered with clouds for more than 60% (54 days) of a DJF season (90 days). This percentage increases to above 80% over the islands of the MC. However, a close resemblance to the ENSO-related rainfall anomalies occurs only for the intermediate OLRs (cf. Figs. 2c and 4f). The frequency of very cold OLRs—in other words, extensive high clouds—increases (decreases) over the islands of Sumatra, Java, Borneo, and Sulawesi during the La Niña (El Niño) phase, but the rainfall is below (above) normal (cf. Figs. 2c and 4c). The better agreement between the ENSO-related rainfall anomalies and the intermediate OLRs signifies an enhancement in isolated deep convection during El Niño, resulting in above normal rainfall. Hence, given the present resolution of OLR, this shows that the OLR frequency of intermediate OLRs, interpreted here as scattered deep clouds, may be a better measure of rainfall–convection than the composite OLR mean or anomaly.

Fig. 4.
Fig. 4.

(left) Frequency of the climatological mean OLR < 170 W m−2 (here interpreted as tropical cirrus–anvil clouds) and anomalies corresponding to the two ENSO phases. (right) As at left, but for OLR values between 170 and 240 W m−2 (convective clouds). Data are for (a),(d) mean OLR, (b),(e) the La Niña phase anomaly, and (c),(f) the El Niño phase.

Citation: Journal of Climate 26, 4; 10.1175/JCLI-D-12-00124.1

c. Local and remote SST responses

The relative contribution of local versus remote SST forcing to local rainfall is of interest in this equatorial region of strong diurnal variation and ENSO forcing. The anomaly maps for SST show that local variations in SST are less than 1°C during the phases of ENSO over most of the regions, except over the NEWP (Fig. 5). Consistent with the findings of Hendon (2003) and others, SST anomalies during the opposite phases of ENSO encompassing the western regions of the MC have the same sign as those in the equatorial regions of the eastern edge of the study domain, continuing into the central Pacific (not shown in map). Cold SSTs surround the western regions of the MC during La Niña whereas warm SSTs surround the same region during El Niño. However, the relationship between SST and rainfall over these regions is not linear (Lau et al. 1997). Over the warm pool and NEWP regions, an increase in rainfall is associated with positive SSTA (cf. Figs. 2a and 5a) while a decrease in rainfall is strongly coupled with negative SSTA (cf. Figs. 2b and 5b). In contrast, the larger maritime areas off the western and southern coast of Sumatra and Java show anomalous positive SSTs during El Niño, but rainfall less than normal occurs. This increase in local SST in this region during El Niño is primarily related to the increase in the frequency of clear sky days, as depicted in OLR frequency (Figs. 4c,f). Nevertheless, in some locations, the relationship between local SST and wind speed will also be substantial. For example, Hasegawa et al. (2010) showed the impact of wind-driven coastal upwelling on SST off the northern coast of New Guinea during El Niño (see Fig. 5b).

Fig. 5.
Fig. 5.

As in Fig. 3, but for SST (°C), contour interval is 0.25°C and 0 lines are removed; and 850 hPa winds (m s−1).

Citation: Journal of Climate 26, 4; 10.1175/JCLI-D-12-00124.1

The rainfall responses to local and remote SST are further explored by correlating gridpoint values of mean monthly rainfall anomaly with the gridpoint values of mean monthly SSTA and with the monthly ONI, respectively for 12 DJF seasons (Fig. 6). In general, over oceans surrounding the MC, the local SST exhibits an anticorrelation with rainfall but a positive correlation with the ONI. More specifically, a negative correlation with SST exists over the sea off the western coast of Sumatra, the Bismarck Sea, and also inside the Gulf of Carpentaria (Fig. 6a). In contrast, positive correlations with SST exist over the western Java Sea, the Sulu Sea, the warm pool area, and the NEWP region. On the other hand, the rainfall is negatively correlated with ONI over oceans except over the NEWP (Fig. 6b). Over the islands, the correlation map between ONI and rainfall shows a strong relationship to the ENSO rainfall anomalies (cf. Figs. 2c and 6b). Furthermore, a partial correlation between rainfall and ONI, after removing the effect of the local SST (not shown), gives no significant reduction in the amplitude of correlations, except over the warm pool and NEWP. Over these two regions, the rainfall is sensitive to variations in both the remote and the local SST. In contrast, elsewhere rainfall is more sensitive to variations in the remote SST. However, further study is required to understand the impact of the variations in local SST on rainfall over the islands.

Fig. 6.
Fig. 6.

(a) Gridpoint simultaneous local correlation between the monthly rainfall anomaly and the monthly SSTA. (b) The simultaneous correlations of the monthly rainfall anomaly with the ENSO index (ONI). A contour at 750 m is overlaid in black over the islands.

Citation: Journal of Climate 26, 4; 10.1175/JCLI-D-12-00124.1

d. ENSO-induced anomalies in large-scale dynamics

Similar to many previous studies, we found that the lower atmospheric westerly winds dominate just south of equator of the MC region, with reduced magnitudes during El Niño but enhanced during La Niña events (Fig. 5). During La Niña over southern Borneo, the winds at 850 hPa appear to be much stronger, zonal, and less convergent. In contrast, during El Niño the band where the winds turn from northeast to northwest (a zone of convergence) is located slightly farther south (Fig. 5b). Figure 7 shows some physical factors known to be related to tropical precipitation, including the warm anomaly (El Niño La Niña) composite for horizontal moisture divergence at 850 hPa, vertical velocity (omega) at 500 hPa, and midtropospheric relative humidity (average of 700 to 400 hPa) to indicate the potential regions of large-scale deep convection and rainfall during the ENSO phases. The regions of strong moisture convergence and divergence anomalies are clearly collocated with the regions of positive and negative rainfall anomalies during the phase of ENSO (cf. Figs. 2c and 7a). Similar spatial structure is evident in the midlevel vertical velocity (Fig. 7b). Interestingly, dipole patterns similar to the anomalous rainfall patterns in Fig. 2c over the larger islands of the MC are also observed in these two large-scale variables. However, a dipole pattern in midtropospheric humidity (Fig. 7c) is only evident over New Guinea. During La Niña, the atmosphere is relatively more humid with anomalous convergence and ascents, resulting in widespread rainfall over the MC, except over the parts of some islands (e.g., southern Borneo). Over these islands, although the midtroposphere remains humid, low-level divergence and midlevel descent are the primary causes for these small regions of below normal rainfall during La Niña. In contrast, the atmosphere is drier over the same island regions during El Niño, but low-level moisture convergence and midlevel ascents may be providing a conducive environment for isolated deep convection and hence above normal rainfall. Alternatively, the presence of positive rainfall anomalies may be inducing areas of moisture convergence. It is not possible to determine the causal relationship from the present analysis. Either way, it appears that the two are linked. Nevertheless, the environment remains different over New Guinea and appears mainly affected by the large-scale conditions over the NEWP region, rather than being associated with the conditions in the western MC. These results show that the finer details of rainfall variations over many islands of the MC are successfully replicated in large-scale variables taken from the ERA-Interim data. However, the ERA-Interim data are still coarse resolution compared with the processes occurring on zonally elongated small islands such as Java and Timor.

Fig. 7.
Fig. 7.

The warm anomaly (El Niño La Niña) for (a) the moisture divergence ( · qV; 10−8 s−1) at 850 hPa, (b) vertical velocity (ω; Pa s−1) at 500 hPa, and (c) midtropospheric (average of 700–400 hPa) relative humidity (%). The negative–positive values (dark–light shading) represent convergence–divergence, upward–downward motion, and moist–drier atmosphere in (a)–(c), respectively. Areas above the 95% significance level are shaded.

Citation: Journal of Climate 26, 4; 10.1175/JCLI-D-12-00124.1

4. Influence of ENSO on the diurnal cycle

To identify common temporally varying pattern of diurnal variations, the results of k-means analysis on the climatological (total) mean diurnal cycles of rainfall composited from 3B42, 3G68-TMI, and 3G68-PR are shown in Fig. 8. There is no noticeable difference in the spatial pattern of clusters composited from 3B42 and 3G68-TMI, but the temporal pattern of clusters 4 and 5 of TMI leads 3B42 by 1–2 h (cf. Figs. 8b and 8d). Although the spatial pattern of clusters is noisier in 3G68-PR (Fig. 8e), this lead in the temporal pattern is even more noticeable (cf. Figs. 8b and 8f) and is consistent with previous studies (e.g., Kubota and Nitta 2001; Yamamoto et al. 2008; Kikuchi and Wang 2008; Rauniyar and Walsh 2011). These studies showed that the IR estimated rainfall (3B42) lags behind the in situ station data and TRMM PR observations of rainfall by approximately 3 h. However, the present study shows that this difference is only noticeable over the tropical land and coastal areas where the contribution of convective rainfall to the total is higher than in subtropical regions (Mori et al. 2004). To avoid any confusion and maintain consistency, hereafter we have interpreted only the results from the 3B42 dataset. But we strongly suggest that the time series of clusters 4 and 5 be moved forward by 2–3 h when comparing the findings of this research with model simulations (Sato et al. 2009) or with in situ observations, especially over land.

Fig. 8.
Fig. 8.

(left) The spatial pattern of the six clusters from the k-means analysis of the climatological average (12 DJFs) diurnal cycle of rainfall composited from TRMM (a) 3B42, (b) 3G68-TMI, and (c) 3G68-PR data. (right) The corresponding diurnal cycles of the six clusters. The percentage of total area covered by each cluster is shown in the parentheses at the bottom of the maps. The contour at 750 m is overlaid in black over the islands.

Citation: Journal of Climate 26, 4; 10.1175/JCLI-D-12-00124.1

For the 3B42 results, there is a clear distinction among the temporal (i.e., diurnal) pattern of clusters (Fig. 8b). Clusters 1 and 2 (purple and blue) cover about 50% of the study domain and comprise the region where the daily rainfall rate is less than 1 mm day−1 and 4 mm day−1, respectively. These two clusters have very weak diurnal variations. In addition, this shows that the mean diurnal rainfall variations of the subtropical region (south of 22°S) of Australia and of the ocean south of 10°S are indeed very similar during the DJF season. Cluster 3 (dark green) mostly consists of the region of the tropical ocean confined between 10°N and 10°S that covers about 30% of the study domain. This cluster shows slightly larger amplitude than in clusters 1 and 2. The rainfall over the pixels of this cluster shows a broad peak from 0600 to 1500 local time (LT) with slight dip at about 1000 LT. The afternoon maximum over this cluster is related to the diurnal heating of ocean surface (Sui et al. 1997), but there are several mechanisms proposed to explain the early morning maximum (Kraus 1963; Gray and Jacobson 1977). Cluster 4 (light green) mostly comprises the coastal offshore regions of the study domain and has a clear diurnal cycle with rainfall maximum between 0500 and 0700 LT. These early morning maxima in rainfall off the coastline are partially due to convergence of land breezes with the prevailing large-scale winds (Houze et al. 1981; Ohsawa et al. 2001), and it has been suggested that they are also due to the propagation of convection embedded in gravity waves generated over the mountainous regions (Mapes et al. 2003; Zuidema 2003; Mori et al. 2004). Cluster 5 (orange) contains the tropical land regions where a strong diurnal variation in rainfall with a maximum in the evening (1700–1900 LT) occurs due to the diurnally regulated surface solar radiative heating and destabilization of the boundary layer (Wallace 1975). Cluster 6 (brown) comprises the tropical land regions where the general features of diurnal rainfall are modified because of the size of an island, the presence of significant orography, and the resulting advection of rainfall under the influence of background synoptic-scale winds (Ichikawa and Yasunari 2008; Rauniyar and Walsh 2011). For example, the rainfall mostly peaks around 1700–1900 LT over a smaller island like Java and also over the top end of Australia, whereas it peaks after 2200–2400 LT over most of the inland regions of Borneo and New Guinea.

To identify the locations where the characteristics of the diurnal cycle are influenced by the ENSO phase, the k-means analysis is then performed on the mean diurnal cycle composited according to the ENSO phases (Fig. 9). However, here the respective cluster mean is subtracted from the temporal pattern of individual clusters to highlight the diurnal cycle. Comparing Figs. 9a and 9c, it is difficult to determine large differences in cluster assignment over the oceans, except for isolated areas such as the NEWP, the region northwest of New Guinea, and small areas to the north of Borneo and Sulawesi. The k-means algorithm is not able to detect the dipole pattern variation in amplitude over the islands, as explained below. Therefore, we perform a Fourier decomposition of the diurnal cycle for both El Niño and La Niña, and the warm anomaly (El Niño La Niña) of the first harmonic is shown in Fig. 9e. This shows that the amplitude is higher over the southern foothills of New Guinea during La Niña but lower over the northern foothills, but the diurnal cycles both peak around the same time and hence these two regions are grouped into the same cluster 6. Similarly during El Niño, the amplitude is enhanced over mountainous regions of Java, but is reduced over the top end of Australia. As the diurnal pattern and the peak time of maximum rainfall are similar at both locations, these regions are grouped into cluster 5.

Fig. 9.
Fig. 9.

(a) The spatial pattern of rainfall of the six k-mean clusters and (b) the clusters’ diurnal cycle anomalies from the k-means analysis of the diurnal cycle of rainfall (3B42) composited for the La Niña phase. (c),(d) As in (a),(b), but for the El Niño phase. Also shown is the percentage of total area covered by a cluster in the parentheses at the bottom of the maps: (e) the amplitude of the first harmonic of the diurnal cycle for the warm anomaly and (f) its phase. The contour at 750 m is overlaid in black over the islands.

Citation: Journal of Climate 26, 4; 10.1175/JCLI-D-12-00124.1

There exists a good match between the mean rainfall changes and differences in the diurnal amplitude of rainfall during ENSO phases over many locations of the study domain (cf. Figs. 2c and 9e). This is evident over parts of Borneo and New Guinea where both the mean rainfall and the amplitude of diurnal rainfall are larger during El Niño compared to La Niña. The same is observed over many smaller islands such as Java, Sulawesi and New Britain. This is obvious as a higher mean rainfall often leads to higher diurnal amplitude. However, there are several exceptions to this rule. For example, over Halmahera Island, the amplitude of diurnal rainfall is larger during El Niño (red color in Fig. 9e), but the mean rainfall is smaller (Fig. 2c). This may be possible as break type monsoons are more common during El Niño (Moron et al. 2010). In this type of environment, storms tend to be more vigorous with a stronger influence from the diurnal cycle in afternoon, which results in a lower mean rainfall but a higher diurnal amplitude (Rauniyar and Walsh 2011). In contrast, over the islands with significant topography, monsoonal systems interact with topographically generated storms, resulting in both higher mean rainfall and higher diurnal amplitude (Qian et al. 2010).

In addition to local development of rainfall, propagation of rainfall also modulates the amplitude of the diurnal cycle. Figure 10 shows the time–distance cross section of rainfall for regions of Sumatra, Borneo, and New Guinea for the two phases of ENSO. Differences between the phases are obvious over New Guinea and to a lesser extent over Borneo. For Sumatra, differences are more subtle: off the southwest coast (6°S, 100°E), the mean rainfall is higher during La Niña (Fig. 2c), and Figs. 10a and 10b show higher rainfall at this location in the early morning hours during La Niña. In contrast, westward propagation from the east coast (at about 104°E) in the afternoon appears weaker during El Niño.

Fig. 10.
Fig. 10.

Composite Hovmöller diagram of 3B42 rainfall averaged for (first row) La Niña and (second row) El Niño along (fourth row) rectangular domains over Sumatra, Borneo, and New Guinea. (third row) The average elevation along the cross section (dotted line) and the daily accumulated rainfall corresponding to the La Niña (El Niño) phase [line with circles (crosses)]. Longitudes of top two rows match that of the third row. Rainfall rates >0.2 mm h−1 are shaded and the contour interval is 0.2 mm h−1.

Citation: Journal of Climate 26, 4; 10.1175/JCLI-D-12-00124.1

Over southern (northern) Borneo, Fig. 9e shows that the diurnal amplitude is mostly stronger (weaker) during El Niño. This is also shown in Figs. 10e and 10f, which show a swath across central and southern Borneo and indicate stronger rainfall development during the late evening and early morning hours in these locations during El Niño. Finally, over New Guinea there are substantial differences between the phases, with late evening and early morning convection immediately north (south) of the central mountain range being stronger during El Niño (La Niña). Propagation of these anomalies occurs in both directions away from the central mountain range. In contrast, there are no substantial differences between the ENSO phases in the characteristics of the diurnal cycle over the central mountain range itself (142°E). In both ENSO phases, there is also some propagation in the afternoon from the southern coast toward the interior. Possible mechanisms for this propagation are discussed below.

5. Discussion and summary

The results shown in this study indicate that the geographical distribution of DJF rainfall pattern in the MC and northern Australia is greatly modified by ENSO. This modification is consistent with the resulting changes in large-scale atmospheric dynamics. In general, both the mean rainfall and amplitude of diurnal rainfall increase (decrease) over wide areas of the oceans, except over the NEWP region, during La Niña (El Niño). This is due to an anomalous positive (negative) frequency of precipitating clouds associated with an increase (decrease) in large-scale vertical motion, moisture convergence, and midtropospheric relative humidity during La Niña (El Niño). In contrast, over the NEWP region, during El Niño rainfall is more than normal and hence the higher amplitude of diurnal rainfall is due to a local warm anomaly in SST. A similar rainfall response to local SSTA is also observed over the warm pool region itself, but instead during La Niña. In contrast, despite the western MC sea regions being relatively warmer during El Niño than in La Niña, less rainfall is observed. This implies that the rainfall over these regions is more sensitive to variations in remote SST than local SST. The primary cause of higher SSTs during El Niño over these regions is less convection due to the eastward shift of the Walker circulation (Aldrian and Dwi Susanto 2003; Chang et al. 2004). In addition, the local subsidence resulting from isolated deep convection over the nearby land, as seen in the increase of intermediate OLR frequency (Fig. 4f), may partially also be responsible for the rise in adjacent SST being accompanied by less rainfall. Kubota et al. (2011) proposed a similar mechanism to explain the drier conditions over the Banda and Arafura seas during October of El Niño years.

In contrast, unlike previous studies, we found that the impact of ENSO over the islands of the MC varies systematically in different ways depending on local geography and orography. Localized positive and negative rainfall anomalies exist over parts of several islands of the MC during the different phases of ENSO. The rainfall of these regions also shows a significant correlation with ENSO, which contradicts the DJF season result of Hendon (2003), who found a weak correlation between area-averaged Indonesian rainfall index and ENSO. We note that the low correlation detected by Hendon (2003) and others is in part due to the averaging of rainfall across the islands that have opposite correlation with ENSO. Chang et al. (2004) suggested the same for the existence of a low positive correlation in the vicinity of Sumatra and the Malay Peninsula. Nevertheless, we found that the rainfall during El Niño is negatively correlated over Malaysia, but mostly positively correlated over Sumatra. The positively correlated parts of islands during El Niño are found to be associated with regions of strong moisture convergence due to weak monsoonal winds and an unstable midtroposphere, even though the atmosphere is drier. In contrast, the areas over the islands with negative rainfall anomalies during La Niña are associated with low-level moisture divergence and a relatively stable middle atmosphere, even though the atmosphere remains humid. However, given the complex orography of the islands of the MC, other mechanisms may also act to modify the rainfall variations. For example, Tangang and Juneng (2004) suggested the existence of an anomalous cyclonic circulation over the SCS and the Sulu Sea during DJF of La Niña as a primary cause for existence of moisture convergence and hence positive anomalous rainfall over the Malaysian Peninsula, northern Borneo, and Mindanao.

We found an increase in the diurnal amplitude of rainfall over various locations of Australia during La Niña, but a decrease during El Niño. However, over the islands of MC, the amplitude of diurnal rainfall changes with different ENSO phases in ways different from one region to another. Over the small islands of the southern MC with significant topography (e.g., Java, Sulawesi, New Britain), both the mean rainfall and the amplitude of the first harmonic are higher during El Niño than in La Niña. Qian et al. (2010) found that the local-scale sea-breeze and valley-breeze convergence are enhanced (suppressed) over southwestern Java during DJF of El Niño (La Niña) years due to less (more) disturbance from the weaker (stronger) monsoonal westerly winds, producing both the higher mean rainfall and the higher diurnal amplitude over the mountains. A similar mechanism may be at work over other smaller islands. However, interpretation of the results of this study is limited over several smaller islands because at the current resolution of the TRMM dataset, the same pixel can contain both ocean and land (Mori et al. 2004), which may produce different results.

Over the larger islands, both the initiation and propagation of diurnal rainfall are modified as local-scale circulations interact with ENSO-induced large-scale anomalies. We found that the higher amplitude of diurnal rainfall at night over the southern parts of Borneo during El Niño is related to the development of deeper storms along the west and east coasts of Borneo in the afternoon, which then propagate inward, leading to a strong merging over southern Borneo. A zone of convergence where winds turn from northeast to northwest is mainly located over southern Borneo during El Niño. This is also a region of weak winds where strong sea breezes and therefore late afternoon convection are more likely to occur, resulting more rainfall and larger diurnal amplitude. In contrast, during La Niña the same region is accompanied by strong westerly winds while the zone of convergence is shifted farther north. The effect may suppress storms on the east coast and result in more limited westward propagation, leading to weaker storms over the southern plains of Borneo. This is quite noticeable in the warm anomaly plots for the moisture convergence and midlevel vertical velocity.

No significant difference in the characteristics of rainfall is noticed between the ENSO phases over the central elongated mountainous regions of New Guinea. Nevertheless, the rainfall on either side of mountain responds quite oppositely between ENSO phases. Both the mean rainfall and diurnal amplitude are higher over the southern plains during La Niña. In contrast, the opposite occurs over the northern plains. During La Niña, Fig. 7a shows more large-scale convergence (divergence) over southern (northern) plains whereas the reverse is observed during El Niño. This is manifested also in large-scale anomalies of moisture convergence and vertical velocity between ENSO phases (Figs. 7b,c). In addition, we speculate that the higher amplitude of diurnal rainfall in the night hours over the southern plains of New Guinea during La Niña may also be due to in situ development of rainfall by the convergence of downslope winds and inland penetrating low-level prevailing westerly winds. A similar mechanism may be in operation but with stronger northwesterly winds and over the northern plains during El Niño, resulting in higher amplitude of rainfall overnight. We also found strong development of morning (0600 LT) rainfall over the southwestern offshore regions of Sumatra, over the southeastern offshore regions of Borneo, and for both the southwestern and northeastern offshore regions of New Guinea. This localized rainfall development may be partially attributed to the boundary layer destabilization due to gravity wave propagation from the mountainous regions and also due to the convergence of land breezes with the large-scale background winds (Mapes et al. 2003; Mori et al. 2004; Zhou and Wang 2006). Further investigation of these mechanisms may be possible through numerical modeling to understand the role of orography in rainfall-building processes between the ENSO phases.

In conclusion, we found that the characteristics of rainfall over the MC and northern Australia are modified by the ENSO phases. Consistent with previous studies, ocean areas typically show an increase in mean rainfall during La Niña which also generally leads to increase in the diurnal amplitude of rainfall. This is due to the rising branch of Walker circulation being located over the MC during La Niña, which results in an increase in large-scale ascent, moisture convergence, and midtropospheric relative humidity. In contrast, unlike previous studies, we found significant variations in the correlation of mean rainfall with ENSO over the islands of the MC, with many regions of the islands having more rainfall during El Niño. These subregional anomalies are more associated with enhanced local moisture convergence and ascent rather than with increased relative humidity although the exact causal relationship between rainfall and moisture convergence anomalies is not established here. In these locations, the amplitude of the diurnal cycle of rainfall is also mostly larger during El Niño, and a number of systematic differences between ENSO phases in the diurnal cycle over these regions have been documented. We have provided some hypotheses to explain these observations, but further work is planned using high-resolution numerical models to investigate the mechanisms for local variations in the characteristics of the diurnal cycle in this region.

Acknowledgments

The authors wish to thank Prof. Ian Simmonds and Dr. Matthew Wheeler for their valuable suggestions and comments. We are very grateful for constructive comments from three anonymous reviewers, who greatly improved the content of this manuscript. In this study the TRMM 3B42 product was obtained from the TRMM Science Data and Information System (TSDIS), distributed by the National Aeronautics and Space Administration (NASA) Goddard Distributed Active Archive Centre (DAAC) while ERA-Interim data were obtained from the European Centre for Medium-Range Weather Forecasts.

REFERENCES

  • Aldrian, E., , and R. Dwi Susanto, 2003: Identification of three dominant rainfall regions within Indonesia and their relationship to sea surface temperature. Int. J. Climatol., 23, 14351452.

    • Search Google Scholar
    • Export Citation
  • Aldrian, E., , L. Dümenil Gates, , and F. H. Widodo, 2007: Seasonal variability of Indonesian rainfall in ECHAM4 simulations and in the reanalyses: The role of ENSO. Theor. Appl. Climatol., 87, 4159.

    • Search Google Scholar
    • Export Citation
  • Beringer, J., , N. J. Tapper, , and T. D. Keenan, 2001: Evolution of maritime continent thunderstorms under varying meteorological conditions over the Tiwi Islands. Int. J. Climatol., 21, 10211036.

    • Search Google Scholar
    • Export Citation
  • Chang, C. P., , Z. Wang, , J. Ju, , and T. Li, 2004: On the relationship between western Maritime Continent monsoon rainfall and ENSO during northern winter. J. Climate, 17, 665672.

    • Search Google Scholar
    • Export Citation
  • Chen, S. S., , and R. A. Houze Jr., 1997: Diurnal variation and life-cycle of deep convective systems over the tropical Pacific warm pool. Quart. J. Roy. Meteor. Soc., 123, 357388.

    • Search Google Scholar
    • Export Citation
  • Giorgi, F., , M. R. Marinucci, , and G. T. Bates, 1993: Development of a second-generation regional climate model (RegCM2). Part I: Boundary-layer and radiative transfer processes. Mon. Wea. Rev., 121, 27942813.

    • Search Google Scholar
    • Export Citation
  • Gray, W. M., , and R. W. Jacobson, 1977: Diurnal variation of deep cumulus convection. Mon. Wea. Rev., 105, 11711188.

  • Hamada, J. I., , M. D. Yamanaka, , J. Matsumoto, , S. Fukao, , P. A. Winarso, , and T. Sribimawati, 2002: Spatial and temporal variations of the rainy season over Indonesia and their link to ENSO. J. Meteor. Soc. Japan, 80, 285310.

    • Search Google Scholar
    • Export Citation
  • Hara, M., , T. Yoshikane, , H. G. Takahashi, , F. Kimura, , A. Noda, , and T. Tokioka, 2009: Assessment of the diurnal cycle of precipitation over the Maritime Continent simulated by a 20 km mesh GCM using TRMM PR data. J. Meteor. Soc. Japan, 87A, 413424.

    • Search Google Scholar
    • Export Citation
  • Hartigan, J. A., , and M. A. Wong, 1979: A K-means clustering algorithm. Appl. Stat., 28, 100108.

  • Hasegawa, T., , K. Ando, , K. Mizuno, , R. Lukas, , B. Taguchi, , and H. Sasaki, 2010: Coastal upwelling along the north coast of Papua New Guinea and El Niño events during 1981–2005. Ocean Dyn., 60, 12551269, doi:10.1007/s10236-010-0334-y.

    • Search Google Scholar
    • Export Citation
  • Haylock, M., , and J. McBride, 2001: Spatial coherence and predictability of Indonesian wet season rainfall. J. Climate, 14, 38823887.

  • Hendon, H. H., 2003: Indonesian rainfall variability: Impacts of ENSO and local air–sea interaction. J. Climate, 16, 17751790.

  • Hirose, M., , and K. Nakamura, 2005: Spatial and diurnal variation of precipitation systems over Asia observed by the TRMM precipitation radar. J. Geophys. Res., 110, D05106, doi:10.1029/2004JD004815.

    • Search Google Scholar
    • Export Citation
  • Hirose, M., , R. Oki, , S. Shimizu, , M. Kachi, , and T. Higashiuwatoko, 2008: Finescale diurnal rainfall statistics refined from eight years of TRMM PR data. J. Appl. Meteor. Climatol., 47, 544561.

    • Search Google Scholar
    • Export Citation
  • Houze, R. A., Jr., , S. G. Geotis, , F. D. Marks Jr., , and A. K. West, 1981: Winter monsoon convection in the vicinity of North Borneo. Part I: Structure and time variation of the clouds and precipitation. Mon. Wea. Rev., 109, 15951614.

    • Search Google Scholar
    • Export Citation
  • Hsu, H.-H., , and M.-Y. Lee, 2005: Topographic effects on the eastward propagation and initiation of the Madden–Julian oscillation. J. Climate, 18, 795809.

    • Search Google Scholar
    • Export Citation
  • Hu, Q., 2003: A multidecadal variation in summer season diurnal rainfall in the central United States. J. Climate, 16, 174178.

  • Huffman, G. J., and Coauthors, 2007: The TRMM multisatellite precipitation analysis (TMPA): Quasi-global, multiyear, combined-sensor precipitation estimates at fine scales. J. Hydrometeor., 8, 3855.

    • Search Google Scholar
    • Export Citation
  • Ichikawa, H., , and T. Yasunari, 2006: Time–space characteristics of diurnal rainfall over Borneo and surrounding oceans as observed by TRMM-PR. J. Climate, 19, 12381260.

    • Search Google Scholar
    • Export Citation
  • Ichikawa, H., , and T. Yasunari, 2007: Propagating diurnal disturbances embedded in the Madden–Julian oscillation. J. Geophys. Res., 34, L18811, doi:10.1029/2007GL030480.

    • Search Google Scholar
    • Export Citation
  • Ichikawa, H., , and T. Yasunari, 2008: Intraseasonal variability in diurnal rainfall over New Guinea and the surrounding oceans during austral summer. J. Climate, 21, 28522868.

    • Search Google Scholar
    • Export Citation
  • Johnson, R. H., , T. M. Rickenbach, , S. A. Rutledge, , P. E. Ciesielski, , and W. H. Schubert, 1999: Trimodal characteristics of tropical convection. J. Climate, 12, 23972418.

    • Search Google Scholar
    • Export Citation
  • Kikuchi, K., , and B. Wang, 2008: Diurnal precipitation regimes in the global tropics. J. Climate, 21, 26802696.

  • Kondragunta, C. R., , and A. Gruber, 1996: Seasonal and annual variability of the diurnal cycle of clouds. J. Geophys. Res., 101 (D16), 21 37721 390.

    • Search Google Scholar
    • Export Citation
  • Kraus, E. B., 1963: The diurnal precipitation change over the sea. J. Atmos. Sci., 20, 551556.

  • Kubota, H., , and T. Nitta, 2001: Diurnal variations of tropical convection observed during the TOGA-COARE. J. Meteor. Soc. Japan, 79, 815830.

    • Search Google Scholar
    • Export Citation
  • Kubota, H., , R. Shirooka, , J. I. Hamada, , and F. Syamsudin, 2011: Interannual rainfall variability over the eastern Maritime Continent. J. Meteor. Soc. Japan, 89A, 111122.

    • Search Google Scholar
    • Export Citation
  • Lau, W. K. M., , and D. E. Waliser, 2005: Intraseasonal Variability in the Atmosphere–Ocean Climate System. Praxis, 436 pp.

  • Lau, W. K. M., , H. T. Wu, , and S. Bony, 1997: The role of large-scale atmospheric circulation in the relationship between tropical convection and sea surface temperature. J. Climate, 10, 381392.

    • Search Google Scholar
    • Export Citation
  • Li, W., , C. Luo, , D. Wang, , and T. Lei, 2010: Diurnal variations of precipitation over the South China Sea. Meteor. Atmos. Phys., 109, 3346.

    • Search Google Scholar
    • Export Citation
  • Liberti, G. L., , F. Chéruy, , and M. Desbois, 2001: Land effect on the diurnal cycle of clouds over the TOGA COARE area, as observed from GMS IR data. Mon. Wea. Rev., 129, 15001517.

    • Search Google Scholar
    • Export Citation
  • Liebmann, B., , and C. A. Smith, 1996: Description of a complete (interpolated) outgoing longwave radiation dataset. Bull. Amer. Meteor. Soc., 77, 12751277.

    • Search Google Scholar
    • Export Citation
  • Love, B. S., , A. J. Matthews, , and G. M. S. Lister, 2011: The diurnal cycle of precipitation over the Maritime Continent in a high-resolution atmospheric model. Quart. J. Roy. Meteor. Soc., 137, 934947.

    • Search Google Scholar
    • Export Citation
  • Mapes, B. E., , T. T. Warner, , M. Xu, , and A. J. Negri, 2003: Diurnal patterns of rainfall in northwestern South America. Part I: Observations and context. Mon. Wea. Rev., 131, 799812.

    • Search Google Scholar
    • Export Citation
  • Massie, S. T., , J. Gille, , C. Craig, , R. Khosravi, , J. Barnett, , W. Read, , and D. Winker, 2010: HIRDLS and CALIPSO observations of tropical cirrus. J. Geophys. Res., 115, D00H11, doi:10.1029/2009JD012100.

    • Search Google Scholar
    • Export Citation
  • Milligan, G. W., , and M. C. Cooper, 1988: A study of standardization of variables in cluster analysis. J. Classif., 5, 181204.

  • Misra, V., 2010: Interaction of interannual and diurnal variations over equatorial Africa. J. Geophys. Res., 115, D01111, doi:10.1029/2009JD012512.

    • Search Google Scholar
    • Export Citation
  • Mori, S., and Coauthors, 2004: Diurnal land–sea rainfall peak migration over Sumatera Island, Indonesian Maritime Continent, observed by TRMM satellite and intensive rawinsonde soundings. Mon. Wea. Rev., 132, 20212039.

    • Search Google Scholar
    • Export Citation
  • Moron, V., , A. W. Robertson, , and J.-H. Qian, 2010: Local versus regional-scale characteristics of monsoon onset and post-onset rainfall over Indonesia. Climate Dyn., 34, 281299.

    • Search Google Scholar
    • Export Citation
  • Morrissey, M. L., 1986: A statistical analysis of the relationships among rainfall, outgoing longwave radiation and the moisture budget during January–March 1979. Mon. Wea. Rev., 114, 931942.

    • Search Google Scholar
    • Export Citation
  • Murakami, M., 1983: Analysis of the deep convective activity over the western Pacific and Southeast Asia. Part I: Diurnal cycle. J. Meteor. Soc. Japan, 61, 6075.

    • Search Google Scholar
    • Export Citation
  • Neale, R., , and J. Slingo, 2003: The Maritime Continent and its role in the global climate: A GCM study. J. Climate, 16, 834848.

  • Negri, A. J., , T. L. Bell, , and L. Xu, 2002: Sampling of the diurnal cycle of precipitation using TRMM. J. Atmos. Oceanic Technol., 19, 13331344.

    • Search Google Scholar
    • Export Citation
  • Nitta, T., , and S. Sekine, 1994: Diurnal variation of convective activity over the tropical western Pacific. J. Meteor. Soc. Japan, 72, 627641.

    • Search Google Scholar
    • Export Citation
  • Ohsawa, T., , H. Ueda, , T. Hayashi, , A. Watanabe, , and J. Matsumoto, 2001: Diurnal variations of convective activity and rainfall in tropical Asia. J. Meteor. Soc. Japan, 79, 333352.

    • Search Google Scholar
    • Export Citation
  • Oki, T., , and K. Musiake, 1994: Seasonal change of the diurnal cycle of precipitation over Japan and Malaysia. J. Appl. Meteor., 33, 14451463.

    • Search Google Scholar
    • Export Citation
  • Poveda, G., and Coauthors, 2005: The diurnal cycle of precipitation in the tropical Andes of Colombia. Mon. Wea. Rev., 133, 228240.

  • Qian, J.-H., 2008: Why precipitation is mostly concentrated over islands in the Maritime Continent. J. Atmos. Sci., 65, 14281441.

  • Qian, J.-H., , A. W. Robertson, , and V. Moron, 2010: Interactions among ENSO, the monsoon, and diurnal cycle in rainfall variability over Java, Indonesia. J. Atmos. Sci., 67, 35093524.

    • Search Google Scholar
    • Export Citation
  • Ramage, C. S., 1968: Role of a tropical “maritime continent” in the atmospheric circulation. Mon. Wea. Rev., 96, 365370.

  • Randel, W. J., , F. Wu, , and W. R. Ríos, 2002: Thermal variability of the tropical tropopause region derived from GPS/MET observations. J. Geophys. Res., 108, 4024, doi:10.1029/2002JD002595.159.

    • Search Google Scholar
    • Export Citation
  • Rauniyar, S. P., , and K. J. E. Walsh, 2011: Scale interaction of the diurnal cycle of rainfall over the Maritime Continent and Australia: Influence of the MJO. J. Climate, 24, 325348.

    • 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.

    • Search Google Scholar
    • Export Citation
  • Ropelewski, C. F., , and M. S. Halpert, 1987: Global and regional scale precipitation patterns associated with the El Niño/Southern Oscillation. Mon. Wea. Rev., 115, 16061626.

    • Search Google Scholar
    • Export Citation
  • Sakurai, N., and Coauthors, 2005: Diurnal cycle of cloud system migration over Sumatera Island. J. Meteor. Soc. Japan, 83, 835850.

  • Sato, T., , H. Miura, , M. Satoh, , Y. N. Takayabu, , and Y. Wang, 2009: Diurnal cycle of precipitation in the tropics simulated in a global cloud-resolving model. J. Climate, 22, 48094826.

    • Search Google Scholar
    • Export Citation
  • Sen Roy, S., , and R. C. Balling, 2004: Analysis of Hawaiian diurnal rainfall patterns. Theor. Appl. Climatol., 79, 209214.

  • Slingo, J., , P. Inness, , R. Neale, , S. Woolnough, , and G. Y. Yang, 2003: Scale interactions on diurnal to seasonal timescales and their relevance to model systematic errors. Ann. Geophys., 46, 139155.

    • Search Google Scholar
    • Export Citation
  • Sorooshian, S., , X. Gao, , K. Hsu, , R. A. Maddox, , Y. Hong, , H. V. Gupta, , and B. Imam, 2002: Diurnal variability of tropical rainfall retrieved from combined GOES and TRMM satellite information. J. Climate, 15, 9831001.

    • Search Google Scholar
    • Export Citation
  • Steinley, D., 2004: Standardizing variables in K-means clustering. Classification, Clustering, and Data Mining Applications, D. Banks et al., Eds., Springer, 53–60.

  • Steinley, D., 2006: K-means clustering: A half-century synthesis. Br. J. Math. Stat. Psychol., 59, 134.

  • Sui, C.-H., , K.-M. Lau, , Y. N. Takayabu, , and D. A. Short, 1997: Diurnal variations in tropical oceanic cumulus convection during TOGA COARE. J. Atmos. Sci., 54, 639655.

    • Search Google Scholar
    • Export Citation
  • Tangang, F. T., , and L. Juneng, 2004: Mechanisms of Malaysian rainfall anomalies. J. Climate, 17, 36163622.

  • Trenberth, K. E., 1997: The definition of El Niño. Bull. Amer. Meteor. Soc., 78, 27712777.

  • Varikoden, H., , A. A. Samah, , and C. A. Babu, 2010: Spatial and temporal characteristics of rain intensity in the peninsular Malaysia using TRMM rain rate. J. Hydrol., 387, 312319.

    • Search Google Scholar
    • Export Citation
  • Waliser, D. E., 2002: Tropical meteorology: Intertropical convergence zones (ITCZ). Encyclopedia of Atmospheric Sciences, J. Holton, J. Pyle, and J. Curry, Eds., Academic Press, 2325–2334.

  • Wallace, J. M., 1975: Diurnal variations in precipitation and thunderstorm frequency over the conterminous United States. Mon. Wea. Rev., 103, 406419.

    • Search Google Scholar
    • Export Citation
  • Wheeler, M. C., , H. H. Hendon, , S. Cleland, , H. Meinke, , and A. Donald, 2009: Impacts of the Madden–Julian oscillation on Australian rainfall and circulation. J. Climate, 22, 14821498.

    • Search Google Scholar
    • Export Citation
  • Wu, P., , M. Hara, , J. Hamada, , M. D. Yamanaka, , and F. Kimura, 2009: Why a large amount of rain falls over the sea in the vicinity of western Sumatra Island during nighttime. J. Appl. Meteor. Climatol., 48, 13451361.

    • Search Google Scholar
    • Export Citation
  • Yamamoto, M. K., , F. A. Furuzawa, , A. Higuchi, , and K. Nakamura, 2008: Comparison of diurnal variations in precipitation systems observed by TRMM PR, TMI, and VIRS. J. Climate, 21, 40114028.

    • Search Google Scholar
    • Export Citation
  • Yang, G.-Y., , and J. Slingo, 2001: The diurnal cycle in the tropics. Mon. Wea. Rev., 129, 784801.

  • Zhang, C., 2005: Madden–Julian oscillation. Rev. Geophys., 43, RG2003, doi:10.1029/2004RG000158.

  • Zhou, L., , and Y. Wang, 2006: Tropical Rainfall Measuring Mission observation and regional model study of precipitation diurnal cycle in the New Guinean region. J. Geophys. Res., 111, D17104, doi:10.1029/2006JD007243.

    • Search Google Scholar
    • Export Citation
  • Zuidema, P., 2003: Convective clouds over the Bay of Bengal. Mon. Wea. Rev., 131, 780798.

Save