Climate Variability over the Maritime Continent and Its Role in Global Climate Variation: A Review

+ Author Affiliations + Find other works by these authors
  • Corresponding author: Tuantuan ZHANG, tuantuanzhang@cuhk.edu.hk
  • Funds:

    Supported by the Vice-Chancellor’s Discretionary Fund of the Chinese University of Hong Kong (4930744), and National Natural Science Foundation of China (41661144019, 91637208, 41690123, and 41690120)

  • doi: 10.1007/s13351-019-9025-x

PDF

  • The Maritime Continent (MC) consists of multiple islands with varying sizes and topography, and surrounding seas. It is characterized by rainfall (convection) variability on multiple spatial and temporal scales. Various large-scale atmospheric, oceanic, and coupled climate systems, such as the El Niño–Southern Oscillation (ENSO), Indian Ocean Dipole (IOD), Madden–Julian Oscillation (MJO), and cold surge, exert significant influences on the spatio-temporal complexity of the MC climate and climate variability. As a major tropical heat source located within the warmest oceanic area (the western Pacific warm pool), the MC has been identified as a region of great importance for climate variation on the global scale. However, prediction of climate variability over the MC and its surrounding areas and the relationships to large-scale atmospheric circulation patterns are big challenges, even for state-of-the-art climate models. In this paper, we provide a thorough review on current understanding of the spatiotemporal complexity and prediction of climate variability over this important region, and its influence on global climate variation.
  • 加载中
  • Fig. 1.  Topography of the MC. Major islands are labeled. Topography height data are from the Global Multi-resolution Terrain Elevation Dataset 2010 (GMTED2010).

    Fig. 2.  (a–l) Spatial patterns of monthly climatological rainfall (shading; mm day–1) and 850-hPa wind (vector; m s–1) from January to Decenmber over 1982–2017. The reference vector of the 850-hPa wind is indicated on the top right of each panel. Rainfall data are from the Global Precipitation Climatology Project version 2.3 Combined Precipitation Data Set (GPCP; Adler et al., 2003), and wind data are from the ERA-Interim (Dee et al., 2011).

    Fig. 3.  Regressions of (a, b) SST (shading; K)/850-hPa wind (vectors; m s−1) and (c, d) rainfall (shading; mm day−1)/850-hPa wind (vectors; m s−1) against the MC rainfall index in (a, c) the wet season (December–March) and (b, d) the dry season (July–October) for the period of 1982–2017. The reference vector of the 850-hPa wind is indicated on the top right in each panel. Stippling (denoted by purple dots) indicates regression coefficients for SST and rainfall (shading) exceeding the 95% confidence level. Only the regressed winds exceeding the 95% confidence level are drawn. The domain of the MC is outlined by a red box. Modified from Zhang et al. (2016c).

    Fig. 4.  Daily-mean precipitation anomaly (shading; mm day−1) from Tropical Rainfall Measuring Mission (TRMM) 3B42HQ for each phase of the MJO (MJO 1–8) as defined in Wheeler and Hendon (2004). Phases move forward with time in the anti-clockwise direction. After Peatman et al. (2014).

    Fig. 5.  (a–f) Diurnal cycle of rainfall (mm day–1) from 0200 LST (local standard time; UTC + 8 h) to 0200 LST next day over the MC from TRMM 3B42 version 7 data for the borel winter of 1998–2014.

    Fig. 6.  (a) Correlation between the MC (15°S–10°N, 95°–145°E) rainfall in the Climate Prediction Center Merged Analysis of Rainfall (CMAP; Xie and Arkin, 1997) and that predicted by the NCEP CFSv2 hindcast (solid lines with circles), observed Niño3.4 (dashed lines), and predicted Niño3.4 (dotted lines) for different lead time (month) in dry season (red lines) and wet season (black lines). The horizonal dotted line in (a) denotes the 95% confidence level. (b) Monthly correlation coefficients between the MC rainfall in CMAP and that predicted by CFSv2 for different time lead (month). Shadings in (b) denote values exceeding the 95% confidence level. After Zhang et al. (2016c).

  • [1]

    Abdillah, M. R., Y. Kanno, and T. Iwasaki, 2017: Tropical–extratropical interactions associated with East Asian cold air outbreaks. Part I: Interannual variability. J. Climate, 30, 2989–3007. doi: 10.1175/jcli-d-16-0152.1.
    [2]

    Adames, Á. F., and J. M. Wallace, 2014: Three-dimensional structure and evolution of the MJO and its relation to the mean flow. J. Atmos. Sci., 71, 2007–2026. doi: 10.1175/JAS-D-13-0254.1.
    [3]

    Adler, R. F., G. J. Huffman, A. Chang, et al., 2003: The version-2 Global Precipitation Climatology Project (GPCP) monthly precipitation analysis (1979–present). J. Hydrometeor., 4, 1147–1167. doi: 10.1175/1525-7541(2003)004<1147:TVGPCP>2.0.CO;2.
    [4]

    Aldrian, E., and R. D. Susanto, 2003: Identification of three dominant rainfall regions within Indonesia and their relationship to sea surface temperature. Int. J. Climatol., 23, 1435–1452. doi: 10.1002/joc.950.
    [5]

    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, 41–59. doi: 10.1007/s00704-006-0218-8.
    [6]

    Allen, M. R., and L. A. Smith, 1997: Optimal filtering in singular spectrum analysis. Phys. Lett., 234, 419–428. doi: 10.1016/S0375-9601(97)00559-8.
    [7]

    Arakawa, O., and A. Kitoh, 2005: Rainfall diurnal variation over the Indonesian Maritime Continent simulated by 20 km-mesh GCM. SOLA, 1, 109–112. doi: 10.2151/sola.2005-029.
    [8]

    Argüeso, D., A. Di Luca, and J. P. Evans, 2016: Precipitation over urban areas in the western Maritime Continent using a convection-permitting model. Climate Dyn., 47, 1143–1159. doi: 10.1007/s00382-015-2893-6.
    [9]

    Ashok, K., Z. Y. Guan, and T. Yamagata, 2003: Influence of the Indian Ocean Dipole on the Australian winter rainfall. Geophys. Res. Lett., 30, 1821. doi: 10.1029/2003GL017926.
    [10]

    Ashok, K., S. K. Behera, S. A. Rao, et al., 2007: El Niño Modoki and its possible teleconnection. J. Geophys. Res. Oceans, 112, C11007. doi: 10.1029/2006JC003798.
    [11]

    As-syakur, A. R., I. W. S. Adnyana, M. S. Mahendra, et al., 2014: Observation of spatial patterns on the rainfall response to ENSO and IOD over Indonesia using TRMM Multisatellite Precipitation Analysis (TMPA). Int. J. Climatol., 34, 3825–3839. doi: 10.1002/joc.3939.
    [12]

    As-syakur, A. R., T. Osawa, F. Miura, et al., 2016: Maritime Continent rainfall variability during the TRMM era: The role of monsoon, topography and El Niño Modoki. Dyn. Atms. Ocea., 75, 58–77. doi: 10.1016/j.dynatmoce.2016.05.004.
    [13]

    Bao, M., and D. L. Hartmann, 2014: The response to MJO-like forcing in a nonlinear shallow-water model. Geophys. Res. Lett., 41, 1322–1328. doi: 10.1002/2013GL057683.
    [14]

    Birch, C. E., S. Webster, S. C. Peatman, et al., 2016: Scale interactions between the MJO and the western Maritime Continent. J. Climate, 29, 2471–2492. doi: 10.1175/jcli-d-15-0557.1.
    [15]

    Burleyson, C. D., S. M. Hagos, Z. Feng, et al., 2018: Large-scale environmental characteristics of MJOs that strengthen and weaken over the Maritime Continent. J. Climate, 31, 5731–5748. doi: 10.1175/jcli-d-17-0576.1.
    [16]

    Chang, C.-P., and K. M. W. Lau, 1980: Northeasterly cold surges and near-equatorial disturbances over the winter MONEX area during December 1974. Part II: Planetary-scale aspects. Mon. Wea. Rev., 108, 298–312. doi: 10.1175/1520-0493(1980)108<0298:Ncsane>2.0.Co;2.
    [17]

    Chang, C.-P., and K. M. Lau, 1982: Short-term planetary-scale interactions over the tropics and midlatitudes during northern winter. Part I: Contrasts between active and inactive periods. Mon. Wea. Rev., 110, 933–946. doi: 10.1175/1520-0493(1982)110<0933:Stpsio>2.0.Co;2.
    [18]

    Chang, C.-P., J. E. Erickson, and K. M. Lau, 1979: Northeasterly cold surges and near-equatorial disturbances over the winter MONEX area during December 1974. Part I: Synoptic aspects. Mon. Wea. Rev., 107, 812–829. doi: 10.1175/1520-0493(1979)107<0812:Ncsane>2.0.Co;2.
    [19]

    Chang, C.-P., Z. Wang, J. H. Ju, et al., 2004: On the relationship between western Maritime Continent monsoon rainfall and ENSO during northern winter. J. Climate, 17, 665–672. doi: 10.1175/1520-0442(2004)017<0665:Otrbwm>2.0.Co;2.
    [20]

    Chang, C.-P., P. A. Harr, and H.-J. Chen, 2005a: Synoptic disturbances over the equatorial South China Sea and western Maritime Continent during boreal winter. Mon. Wea. Rev., 133, 489–503. doi: 10.1175/mwr-2868.1.
    [21]

    Chang, C.-P., Z. Wang, J. McBride, et al., 2005b: Annual cycle of Southeast Asia—Maritime Continent rainfall and the asymmetric monsoon transition. J. Climate, 18, 287–301. doi: 10.1175/jcli-3257.1.
    [22]

    Chang, P., R. Saravanan, L. Ji, et al., 2000: The effect of local sea surface temperatures on atmospheric circulation over the tropical Atlantic sector. J. Climate, 13, 2195–2216. doi: 10.1175/1520-0442(2000)013<2195:TEOLSS>2.0.CO;2.
    [23]

    Charney, J. G., and J. Shukla, 1981: Predictability of monsoons. Monsoon Dynamics. Sir J. Lighthill, and R. P. Pearce, Cambridge University Press, Cambridge, UK, 99–110 pp.
    [24]

    Chen, J. W., Y. Deng, J. F. Wang, et al., 2017: Hindcasting the Madden–Julian Oscillation with a new parameterization of surface heat fluxes. J. Adv. Model. Earth Syst., 9, 2696–2709. doi: 10.1002/2017MS001175.
    [25]

    Chen, T.-C., J.-D. Tsay, J. Matsumoto, et al., 2015a: Development and formation mechanism of the Southeast Asian winter heavy rainfall events around the South China Sea. Part I: Formation and propagation of cold surge vortex. J. Climate, 28, 1417–1443. doi: 10.1175/jcli-d-14-00170.1.
    [26]

    Chen, T.-C., J.-D. Tsay, and J. Matsumoto, 2015b: Development and formation mechanism of the Southeast Asian winter heavy rainfall events around the South China Sea. Part II: Multiple interactions. J. Climate, 28, 1444–1464. doi: 10.1175/jcli-d-14-00171.1.
    [27]

    Ciesielski, P. E., and R. H. Johnson, 2006: Contrasting characteristics of convection over the northern and southern South China Sea during SCSMEX. Mon. Wea. Rev., 134, 1041–1062. doi: 10.1175/mwr3113.1.
    [28]

    Collier, J. C., and K. P. Bowman, 2004: Diurnal cycle of tropical precipitation in a general circulation model. J. Geophys. Res. Atmos., 109, D17105. doi: 10.1029/2004JD004818.
    [29]

    Compo, G. P., G. N. Kiladis, and P. J. Webster, 1999: The horizontal and vertical structure of East Asian winter monsoon pressure surges. Quart. J. Roy. Meteor. Soc., 125, 29–54. doi: 10.1002/qj.49712555304.
    [30]

    Cronin, T. W., K. A. Emanuel, and P. Molnar, 2015: Island precipitation enhancement and the diurnal cycle in radiative–convective equilibrium. Quart. J. Roy. Meteor. Soc., 141, 1017–1034. doi: 10.1002/qj.2443.
    [31]

    Crueger, T., B. Stevens, and R. Brokopf, 2013: The Madden–Julian Oscillation in ECHAM6 and the introduction of an objective MJO metric. J. Climate, 26, 3241–3257. doi: 10.1175/JCLI-D-12-00413.1.
    [32]

    Dee, D. P., S. M. Uppala, A. J. Simmon, et al., 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553–597. doi: 10.1002/qj.828.
    [33]

    Emanuel, K. A., 1987: An air–sea interaction model of intraseasonal oscillations in the tropics. J. Atmos. Sci., 44, 2324–2340. doi: 10.1175/1520-0469(1987)044<2324:Aasimo>2.0.Co;2.
    [34]

    Feng, J., L. Wang, W. Chen, et al., 2010: Different impacts of two types of Pacific Ocean warming on Southeast Asian rainfall during boreal winter. J. Geophys. Res. Atmos., 115, D24122. doi: 10.1029/2010JD014761.
    [35]

    Feng, J., T. Li, and W. J. Zhu, 2015: Propagating and nonpropagating MJO events over Maritime Continent. J. Climate, 28, 8430–8449. doi: 10.1175/jcli-d-15-0085.1.
    [36]

    Gianotti, R. L., D. F. Zhang, and E. A. B. Eltahir, 2012: Assessment of the regional climate model version 3 over the Maritime Continent using different cumulus parameterization and land surface schemes. J. Climate, 25, 638–656. doi: 10.1175/jcli-d-11-00025.1.
    [37]

    Hagos, S. M., C. D. Zhang, Z. Feng, et al., 2016: The impact of the diurnal cycle on the propagation of Madden–Julian Oscillation convection across the Maritime Continent. J. Adv. Model. Earth Syst., 8, 1552–1564. doi: 10.1002/2016MS000725.
    [38]

    Hamada, J.-I., M. D. Yamanaka, J. Matsumoto, et al., 2002: Spatial and temporal variations of the rainy season over Indonesia and their link to ENSO. J. Meteor. Soc. Japan, 80, 285–310. doi: 10.2151/jmsj.80.285.
    [39]

    Hamada, J.-I., S. Mori, H. Kubota, et al., 2012: Interannual rainfall variability over northwestern Jawa and its relation to the Indian Ocean Dipole and El Niño–Southern Oscillation events. SOLA, 8, 69–72. doi: 10.2151/sola.2012-018.
    [40]

    Hara, M., T. Yoshikane, H. G. Takahashi, et al., 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, 413–424. doi: 10.2151/jmsj.87A.413.
    [41]

    Haylock, M., and J. McBride, 2001: Spatial coherence and predictability of Indonesian wet season rainfall. J. Climate, 14, 3882–3887. doi: 10.1175/1520-0442(2001)014<3882:Scapoi>2.0.Co;2.
    [42]

    Hendon, H. H., 2003: Indonesian rainfall variability: Impacts of ENSO and local air–sea interaction. J. Climate, 16, 1775–1790. doi: 10.1175/1520-0442(2003)016<1775:Irvioe>2.0.Co;2.
    [43]

    Hidayat, R., and S. Kizu, 2010: Influence of the Madden–Julian Oscillation on Indonesian rainfall variability in austral summer. Int. J. Climatol., 30, 1816–1825. doi: 10.1002/joc.2005.
    [44]

    Holland, G. J., and T. D. Keenan, 1980: Diurnal variations of convection over the " Maritime Continent”. Mon. Wea. Rev., 108, 223–225. doi: 10.1175/1520-0493(1980)108<0223:Dvocot>2.0.Co;2.
    [45]

    Hou, A. Y., R. K. Kakar, S. Neeck, et al., 2014: The global precipitation measurement mission. Bull. Amer. Meteor. Soc., 95, 701–722. doi: 10.1175/bams-d-13-00164.1.
    [46]

    Hsu, H.-H., and M.-Y. Lee, 2005: Topographic effects on the eastward propagation and initiation of the Madden–Julian Oscillation. J. Climate, 18, 795–809. doi: 10.1175/jcli-3292.1.
    [47]

    Huang, B., 2004: Remotely forced variability in the tropical Atlantic Ocean. Climate Dyn., 23, 133–152. doi: 10.1007/s00382-004-0443-8.
    [48]

    Hung, C.-W., X. D. Liu, and M. Yanai, 2004: Symmetry and asymmetry of the Asian and Australian summer monsoons. J. Climate, 17, 2413–2426. doi: 10.1175/1520-0442(2004)017<2413:Saaota>2.0.Co;2.
    [49]

    Im, E.-S., and E. A. B. Eltahir, 2018: Simulation of the diurnal variation of rainfall over the western Maritime Continent using a regional climate model. Climate Dyn., 51, 73–88. doi: 10.1007/s00382-017-3907-3.
    [50]

    Inness, P. M., and J. M. Slingo, 2006: The interaction of the Madden–Julian Oscillation with the Maritime Continent in a GCM. Quart. J. Roy. Meteor. Soc., 132, 1645–1667. doi: 10.1256/qj.05.102.
    [51]

    Jia, X. J., J. W. Ge, and S. Wang, 2016: Diverse impacts of ENSO on wintertime rainfall over the Maritime Continent. Int. J. Climatol., 36, 3384–3397. doi: 10.1002/joc.4562.
    [52]

    Jiang, L. S., and T. Li, 2018: Why rainfall response to El Niño over Maritime Continent is weaker and non-uniform in boreal winter than in boreal summer. Climate Dyn., 51, 1465–1483. doi: 10.1007/s00382-017-3965-6.
    [53]

    Jiang, X. W., S. Yang, Y. Q. Li, et al., 2013: Dynamical prediction of the East Asian winter monsoon by the NCEP Climate Forecast System. J. Geophys. Res. Atmos., 118, 1312–1328. doi: 10.1002/jgrd.50193.
    [54]

    Jiang, X. W., J. C. Shu, X. Wang, et al., 2017: The roles of convection over the western Maritime Continent and the Philippine Sea in interannual variability of summer rainfall over southwest China. J. Hydrometeor., 18, 2043–2056. doi: 10.1175/jhm-d-16-0292.1.
    [55]

    Johnson, S. J., R. C. Levine, A. G. Turner, et al., 2016: The resolution sensitivity of the South Asian monsoon and Indo–Pacific in a global 0.35° AGCM. Climate Dyn., 46, 807–831. doi: 10.1007/s00382-015-2614-1.
    [56]

    Kao, H. Y., and J. Y. Yu, 2009: Contrasting eastern-Pacific and central-Pacific types of ENSO. J. Climate, 22, 615–632. doi: 10.1175/2008JCLI2309.1.
    [57]

    Keenan, T., S. Rutledge, R. Carbone, et al., 2000: The Maritime Continent—Thunderstorm Experiment (MCTEX): Overview and some results. Bull. Amer. Meteor. Soc., 81, 2433–2456. doi: 10.1175/1520-0477(2000)081<2433:Tmctem>2.3.Co;2.
    [58]

    Kerns, B. W., and S. S. Chen, 2016: Large-scale precipitation tracking and the MJO over the Maritime Continent and Indo–Pacific warm pool. J. Geophys. Res. Atmos., 121, 8755–8776. doi: 10.1002/2015JD024661.
    [59]

    Kida, S., and K. J. Richards, 2009: Seasonal sea surface temperature variability in the Indonesian Seas. J. Geophys. Res. Oceans, 114, C06016. doi: 10.1029/2008JC005150.
    [60]

    Kim, D., H. Kim, and M.-I. Lee, 2017: Why does the MJO detour the Maritime Continent during austral summer? Geophys. Res. Lett., 44, 2579–2587. doi: 10.1002/2017GL072643.
    [61]

    Kim, H.-M., P. J. Webster, V. E. Toma, et al., 2014: Predictability and prediction skill of the MJO in two operational forecasting systems. J. Climate, 27, 5364–5378. doi: 10.1175/jcli-d-13-00480.1.
    [62]

    Kim, H.-M., D. Kim, F. Vitart, et al., 2016: MJO propagation across the Maritime Continent in the ECMWF ensemble prediction system. J. Climate, 29, 3973–3988. doi: 10.1175/jcli-d-15-0862.1.
    [63]

    Kubota, H., R. Shirooka, J.-I. Hamada, et al., 2011: Interannual rainfall variability over the eastern Maritime Continent. J. Meteor. Soc. Japan, 89A, 111–122. doi: 10.2151/jmsj.2011-A07.
    [64]

    Kumar, A., M. Y. Chen, and W. Q. Wang, 2013: Understanding prediction skill of seasonal mean precipitation over the tropics. J. Climate, 26, 5674–5681. doi: 10.1175/jcli-d-12-00731.1.
    [65]

    Lau, K.-M., and P. H. Chan, 1983a: Short-term climate variability and atmospheric teleconnections from satellite-observed outgoing longwave radiation. Part I: Simultaneous relationships. J. Atmos. Sci., 40, 2735–2750. doi: 10.1175/1520-0469(1983)040<2735:Stcvaa>2.0.Co;2.
    [66]

    Lau, K.-M., and P. H. Chan, 1983b: Short-term climate variability and atmospheric teleconnections from satellite-observed outgoing longwave radiation. Part II: Lagged correlations. J. Atmos. Sci., 40, 2751–2767. doi: 10.1175/1520-0469(1983)040<2751:Stcvaa>2.0.Co;2.
    [67]

    Lau, K.-M., C.-P. Chang, and P. H. Chan, 1983: Short-term planetary-scale interactions over the tropics and midlatitudes. Part II: Winter-MONEX period. Mon. Wea. Rev., 111, 1372–1388. doi: 10.1175/1520-0493(1983)111<1372:Stpsio>2.0.Co;2.
    [68]

    Lau, K. M., and J. S. Boyle, 1987: Tropical and extratropical forcing of the large-scale circulation: A diagnostic study. Mon. Wea. Rev., 115, 400–428. doi: 10.1175/1520-0493(1987)115<0400:Taefot>2.0.Co;2.
    [69]

    Lau, N.-C., and K.-M. Lau, 1984: The structure and energetics of midlatitude disturbances accompanying cold-air outbreaks over East Asia. Mon. Wea. Rev., 112, 1309–1327. doi: 10.1175/1520-0493(1984)112<1309:Tsaeom>2.0.Co;2.
    [70]

    Li, Q. P., S. Yang, T. W. Wu, et al., 2017: Subseasonal dynamical prediction of East Asian cold surges. Wea. Forecasting, 32, 1675–1694. doi: 10.1175/WAF-D-16-0209.1.
    [71]

    Li, Y., N. C. Jourdain, A. S. Taschetto, et al., 2017: Resolution dependence of the simulated precipitation and diurnal cycle over the Maritime Continent. Climate Dyn., 48, 4009–4028. doi: 10.1007/s00382-016-3317-y.
    [72]

    Li, Z. N., S. Yang, X. M. Hu, et al., 2018: Charge in long-lasting El Niño events by convection-induced wind anomalies over the western Pacific in boreal spring. J. Climate, 31, 3755–3763. doi: 10.1175/JCLI-D-17-0558.1.
    [73]

    Liang, J. Y., S. Yang, Z.-Z. Hu, et al., 2009: Predictable patterns of the Asian and Indo–Pacific summer precipitation in the NCEP CFS. Climate Dyn., 32, 989–1001. doi: 10.1007/s00382-008-0420-8.
    [74]

    Lin, H., G. Brunet, and J. Derome, 2008: Forecast skill of the Madden–Julian Oscillation in two canadian atmospheric models. Mon. Wea. Rev., 136, 4130–4149. doi: 10.1175/2008mwr2459.1.
    [75]

    LinHo, and B. Wang, 2002: The time–space structure of the Asian–Pacific summer monsoon: A fast annual cycle view. J. Climate, 15, 2001–2019. doi: 10.1175/1520-0442(2002)015<2001:Ttssot>2.0.Co;2.
    [76]

    Liu, X. W., S. Yang, J. L. Li, et al., 2015: Subseasonal predictions of regional summer monsoon rainfall over tropical Asian oceans and land. J. Climate, 28, 9583–9605. doi: 10.1175/JCLI-D-14-00853.1.
    [77]

    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, 934–947. doi: 10.1002/qj.809.
    [78]

    Madden, R. A., and P. R. Julian, 1972: Description of global-scale circulation cells in the tropics with a 40–50 day period. J. Atmos. Sci., 29, 1109–1123. doi: 10.1175/1520-0469(1972)029<1109:Dogscc>2.0.Co;2.
    [79]

    Madden, R. A., and P. R. Julian, 1994: Observations of the 40–50-day tropical oscillation—A review. Mon. Wea. Rev., 122, 814–837. doi: 10.1175/1520-0493(1994)122<0814:Ootdto>2.0.Co;2.
    [80]

    Maloney, E. D., and A. H. Sobel, 2004: Surface fluxes and ocean coupling in the tropical intraseasonal oscillation. J. Climate, 17, 4368–4386. doi: 10.1175/jcli-3212.1.
    [81]

    Mapes, B., and R. A. Jr. Houze, 1992: An integrated view of the 1987 Australian monsoon and its mesoscale convective systems. I: Horizontal structure. Quart. J. Roy. Meteor. Soc., 118, 927–963. doi: 10.1002/qj.49711850706.
    [82]

    Mapes, B. E., T. T. Warner, and M. Xu, 2003: Diurnal patterns of rainfall in northwestern south America. Part III: Diurnal gravity waves and nocturnal convection offshore. Mon. Wea. Rev., 131, 830–844. doi: 10.1175/1520-0493(2003)131<0830:Dporin>2.0.Co;2.
    [83]

    Matsumoto, J., 1992: The seasonal changes in Asian and Australian Monsoon regions. J. Meteor. Soc. Japan, 70, 257–273. doi: 10.2151/jmsj1965.70.1B_257.
    [84]

    Matsumoto, J., and T. Murakami, 2000: Annual changes of tropical convective activities as revealed from equatorially symmetric OLR data. J. Meteor. Soc. Japan, 78, 543–561. doi: 10.2151/jmsj1965.78.5_543.
    [85]

    Matsumoto, J., and T. Murakami, 2002: Seasonal migration of monsoons between the Northern and Southern Hemisphere as revealed from equatorially symmetric and asymmetric OLR data. J. Meteor. Soc. Japan, 80, 419–437. doi: 10.2151/jmsj.80.419.
    [86]

    McBride, J. L., and N. Nicholls, 1983: Seasonal relationships between Australian rainfall and the Southern Oscillation. Mon. Wea. Rev., 111, 1998–2004. doi: 10.1175/1520-0493(1983)111<1998:Srbara>2.0.Co;2.
    [87]

    McBride, J. L., M. R. Haylock, and N. Nicholls, 2003: Relationships between the Maritime Continent heat source and the El Niño–Southern Oscillation phenomenon. J. Climate, 16, 2905–2914. doi: 10.1175/1520-0442(2003)016<2905:Rbtmch>2.0.Co;2.
    [88]

    Meehl, G. A., 1987: The annual cycle and interannual variability in the tropical Pacific and Indian Ocean regions. Mon. Wea. Rev., 115, 27–50. doi: 10.1175/1520-0493(1987)115<0027:Tacaiv>2.0.Co;2.
    [89]

    Miura, H., M. Satoh, T. Nasuno, et al., 2007: A Madden–Julian Oscillation event realistically simulated by a global cloud-resolving model. Science, 318, 1763–1765. doi: 10.1126/science.1148443.
    [90]

    Moorthi, S., and M. J. Suarez, 1999: Documentation of Version 2 of Relaxed Arakawa–Schubert Cumulus Parameterization with Convective Downdrafts. NOAA Office Note 99–01, NOAA, Washington DC, 44 pp.
    [91]

    Mori, S., H. Jun-Ichi, Y. I. Tauhid, et al., 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, 2021–2039. doi: 10.1175/1520-0493(2004)132<2021:Dlrpmo>2.0.Co;2.
    [92]

    Neale, R., and J. Slingo, 2003: The Maritime Continent and its role in the global climate: A GCM study. J. Climate, 16, 834–848. doi: 10.1175/1520-0442(2003)016<0834:Tmcair>2.0.Co;2.
    [93]

    Neelin, J. D., I. M. Held, and K. H. Cook, 1987: Evaporation-wind feedback and low-frequency variability in the tropical atmosphere. J. Atmos. Sci., 44, 2341–2348. doi: 10.1175/1520-0469(1987)044<2341:Ewfalf>2.0.Co;2.
    [94]

    Neelin, J. D., D. S. Battisti, A. C. Hirst, et al., 1998: ENSO theory. J. Geophys. Res. Oceans, 103, 14261–14290. doi: 10.1029/97JC03424.
    [95]

    Nesbitt, S. W., and E. J. Zipser, 2003: The diurnal cycle of rainfall and convective intensity according to three years of TRMM measurements. J. Climate, 16, 1456–1475. doi: 10.1175/1520-0442-16.10.1456.
    [96]

    Nguyen, H., C. Franklin, and A. Protat, 2017: Understanding the ACCESS model errors over the Maritime Continent using CloudSat and CALIPSO simulators. Quart. J. Roy. Meteor. Soc., 143, 3136–3152. doi: 10.1002/qj.3168.
    [97]

    Oh, J.-H., K.-Y. Kim, and G.-H. Lim, 2012: Impact of MJO on the diurnal cycle of rainfall over the western Maritime Continent in the austral summer. Climate Dyn., 38, 1167–1180. doi: 10.1007/s00382-011-1237-4.
    [98]

    Oh, J.-H., B.-M. Kim, K.-Y. Kim, et al., 2013: The impact of the diurnal cycle on the MJO over the Maritime Continent: A modeling study assimilating TRMM rain rate into global analysis. Climate Dyn., 40, 893–911. doi: 10.1007/s00382-012-1419-8.
    [99]

    Pan, H.-L., and W.-S. Wu, 1995: Implementing a Mass Flux Convection Parameterization Package for the NMC Medium-Range Forecast Model. NMC Office Note No. 409, NOAA NWS, Silver Spring, MD, USA, 42 pp.
    [100]

    Park, T.-W., C.-H. Ho, and S. Yang, 2011: Relationship between the Arctic Oscillation and cold surges over East Asia. J. Climate, 24, 68–83. doi: 10.1175/2010JCLI3529.1.
    [101]

    Peatman, S. C., A. J. Matthews, and D. P. Stevens, 2014: Propagation of the Madden–Julian Oscillation through the Maritime Continent and scale interaction with the diurnal cycle of precipitation. Quart. J. Roy. Meteor. Soc., 140, 814–825. doi: 10.1002/qj.2161.
    [102]

    Peatman, S. C., A. J. Matthews, and D. P. Stevens, 2015: Propagation of the Madden–Julian Oscillation and scale interaction with the diurnal cycle in a high-resolution GCM. Climate Dyn., 45, 2901–2918. doi: 10.1007/s00382-015-2513-5.
    [103]

    Qian, J.-H., 2008: Why precipitation is mostly concentrated over islands in the Maritime Continent. J. Atmos. Sci., 65, 1428–1441. doi: 10.1175/2007jas2422.1.
    [104]

    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, 3509–3524. doi: 10.1175/2010jas3348.1.
    [105]

    Qian, J.-H., A. W. Robertson, and V. Moron, 2013: Diurnal cycle in different weather regimes and rainfall variability over Borneo associated with ENSO. J. Climate, 26, 1772–1790. doi: 10.1175/jcli-d-12-00178.1.
    [106]

    Ramage, C. S., 1968: Role of a tropical " Maritime Continent” in the atmospheric circulation. Mon. Wea. Rev., 96, 365–370. doi: 10.1175/1520-0493(1968)096<0365:Roatmc>2.0.Co;2.
    [107]

    Rashid, H. A., and A. C. Hirst, 2017: Mechanisms of improved rainfall simulation over the Maritime Continent due to increased horizontal resolution in an AGCM. Climate Dyn., 49, 1747–1764. doi: 10.1007/s00382-016-3413-z.
    [108]

    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, 325–348. doi: 10.1175/2010jcli3673.1.
    [109]

    Rauniyar, S. P., and K. J. E. Walsh, 2013: Influence of ENSO on the diurnal cycle of rainfall over the Maritime Continent and Australia. J. Climate, 26, 1304–1321. doi: 10.1175/jcli-d-12-00124.1.
    [110]

    Ruti, P. M., D. Di Rocco, and S. Gualdi, 2006: Impact of increased vertical resolution on simulation of tropical climate. Theor. Appl. Climatol., 85, 61–80. doi: 10.1007/s00704-005-0174-8.
    [111]

    Saji, N. H., B. N. Goswami, P. N. Vinayachandran, et al., 1999: A dipole mode in the tropical Indian Ocean. Nature, 401, 360–363. doi: 10.1038/43854.
    [112]

    Salimun, E., F. Tangang, L. Juneng, et al., 2014: Differential impacts of conventional El Niño versus El Niño Modoki on Malaysian rainfall anomaly during winter monsoon. Int. J. Climatol., 34, 2763–2774. doi: 10.1002/joc.3873.
    [113]

    Schiemann, R., M.-E. Demory, M. S. Mizielinski, et al., 2014: The sensitivity of the tropical circulation and Maritime Continent precipitation to climate model resolution. Climate Dyn., 42, 2455–2468. doi: 10.1007/s00382-013-1997-0.
    [114]

    Seo, K.-H., and W. Q. Wang, 2010: The Madden–Julian Oscillation simulated in the NCEP Climate Forecast System model: The importance of stratiform heating. J. Climate, 23, 4770–4793. doi: 10.1175/2010jcli2983.1.
    [115]

    Seo, K.-H., W. Q. Wang, J. Gottschalck, et al., 2009: Evaluation of MJO forecast skill from several statistical and dynamical forecast models. J. Climate, 22, 2372–2388. doi: 10.1175/2008jcli2421.1.
    [116]

    Simpson, J., N. E. Westcott, R. J. Clerman, et al., 1980: On cumulus mergers. Arch. Meteor. Geophys. Bioklimatol. Ser. A, 29, 1–40. doi: 10.1007/bf02247731.
    [117]

    Simpson, J., T. D. Keenan, B. Ferrier, et al., 1993: Cumulus mergers in the Maritime Continent region. Meteor. Atmos. Phys., 51, 73–99. doi: 10.1007/bf01080881.
    [118]

    Slingo, J. M., 1998: Extratropical forcing of tropical convection in a northern winter simulation with the UGAMP GCM. Quart. J. Roy. Meteor. Soc., 124, 27–51. doi: 10.1002/qj.49712454503.
    [119]

    Sobel, A. H., E. D. Maloney, G. Bellon, et al., 2008: The role of surface heat fluxes in tropical intraseasonal oscillations. Nat. Geosci., 1, 653–657. doi: 10.1038/ngeo312.
    [120]

    Sobel, A. H., C. D. Burleyson, and S. E. Yuter, 2011: Rain on small tropical islands. J. Geophys. Res. Atmos., 116, D08102. doi: 10.1029/2010JD014695.
    [121]

    Stuecker, M. F., A. Timmermann, F.-F. Jin, et al., 2013: A combination mode of the annual cycle and the El Niño/Southern Oscillation. Nat. Geosci., 6, 540–544. doi: 10.1038/ngeo1826.
    [122]

    Suarez, M. J., and P. S. Schopf, 1988: A delayed action oscillator for ENSO. J. Atmos. Sci., 45, 3283–3287. doi: 10.1175/1520-0469(1988)045<3283:ADAOFE>2.0.CO;2.
    [123]

    Sutton, R. T., S. P. Jewson, and D. P. Rowell, 2000: The elements of climate variability in the tropical Atlantic region. J. Climate, 13, 3261–3284. doi: 10.1175/1520-0442(2000)013<3261:TEOCVI>2.0.CO;2.
    [124]

    Tan, H. C., P. Ray, B. S. Barrett, et al., 2018: Role of topography on the MJO in the Maritime Continent: A numerical case study. Climate Dyn. . doi: 10.1007/s00382-018-4275-3.
    [125]

    Tanaka, M., 1994: The onset and retreat dates of the Austral summer monsoon over Indonesia, Australia and New Guinea. J. Meteor. Soc. Japan, 72, 255–267. doi: 10.2151/jmsj1965.72.2_255.
    [126]

    Teo, C.-K., T.-Y. Koh, J. C.-F. Lo, et al., 2011: Principal component analysis of observed and modeled diurnal rainfall in the Maritime Continent. J. Climate, 24, 4662–4675. doi: 10.1175/2011jcli4047.1.
    [127]

    Toh, Y. Y., A. G. Turner, S. J. Johnson, et al., 2018: Maritime Continent seasonal climate biases in AMIP experiments of the CMIP5 multimodel ensemble. Climate Dyn., 50, 777–800. doi: 10.1007/s00382-017-3641-x.
    [128]

    Tseng, W.-L., H.-H. Hsu, N. Keenlyside, et al., 2017: Effects of surface orography and land–sea contrast on the Madden–Julian Oscillation in the Maritime Continent: A numerical study using ECHAM5-SIT. J. Climate, 30, 9725–9741. doi: 10.1175/jcli-d-17-0051.1.
    [129]

    Tung, W.-W., D. Giannakis, and A. J. Majda, 2014: Symmetric and antisymmetric convection signals in the Madden–Julian Oscillation. Part I: Basic modes in infrared brightness temperature. J. Atmos. Sci., 71, 3302–3326. doi: 10.1175/jas-d-13-0122.1.
    [130]

    Venzke, S., M. R. Allen, R. T. Sutton, et al., 1999: The atmospheric response over the North Atlantic to decadal changes in sea surface temperature. J. Climate, 12, 2562–2584. doi: 10.1175/1520-0442(1999)012<2562:TAROTN>2.0.CO;2.
    [131]

    Vincent, C. L., and T. P. Lane, 2016: Evolution of the diurnal precipitation cycle with the passage of a Madden–Julian Oscillation event through the Maritime Continent. Mon. Wea. Rev., 144, 1983–2005. doi: 10.1175/mwr-d-15-0326.1.
    [132]

    Vitart, F., and F. Molteni, 2010: Simulation of the Madden–Julian Oscillation and its teleconnections in the ECMWF forecast system. Quart. J. Roy. Meteor. Soc., 136, 842–855. doi: 10.1002/qj.623.
    [133]

    Vitart, F., S. Woolnough, M. A. Balmaseda, et al., 2007: Monthly forecast of the Madden–Julian Oscillation using a coupled GCM. Mon. Wea. Rev., 135, 2700–2715. doi: 10.1175/MWR3415.1.
    [134]

    Wang, B., R. G. Wu, and X. H. Fu, 2000: Pacific–East Asian Teleconnection: How does ENSO affect East Asian climate? J Climate, 13, 1517–1536. doi: 10.1175/1520-0442(2000)013<1517:Peathd>2.0.Co;2.
    [135]

    Wang, B., B. Q. Xiang, and J.-Y. Lee, 2013: Subtropical high predictability establishes a promising way for monsoon and tropical storm predictions. Proc. Natl. Acad. Sci. USA, 110, 2718–2722. doi: 10.1073/pnas.1214626110.
    [136]

    Wang, B., J.-Y. Lee, and B. Q. Xiang, 2015: Asian summer monsoon rainfall predictability: A predictable mode analysis. Climate Dyn., 44, 61–74. doi: 10.1007/s00382-014-2218-1.
    [137]

    Wang, M., Z. Y. Guan, and D. C. Jin, 2018: Two new sea surface temperature anomalies indices for capturing the eastern and central equatorial Pacific type El Niño–Southern Oscillation events during boreal summer. Int. J. Climatol., 38, 4066–4076. doi: 10.1002/joc.5552.
    [138]

    Wang, W. Q., M.-P. Hung, S. J. Weaver, et al., 2014: MJO prediction in the NCEP Climate Forecast System version 2. Climate Dyn., 42, 2509–2520. doi: 10.1007/s00382-013-1806-9.
    [139]

    Wang, W. Q., A. Kumar, J. X. Fu, et al., 2015: What is the role of the sea surface temperature uncertainty in the prediction of tropical convection associated with the MJO? Mon. Wea. Rev., 143, 3156–3175. doi: 10.1175/mwr-d-14-00385.1.
    [140]

    Wang, Y. Q., L. Zhou, and K. Hamilton, 2007: Effect of convective entrainment/detrainment on the simulation of the tropical precipitation diurnal cycle. Mon. Wea. Rev., 135, 567–585. doi: 10.1175/mwr3308.1.
    [141]

    Weaver, S. J., W. Q. Wang, M. Y. Chen, et al., 2011: Representation of MJO variability in the NCEP climate forecast system. J. Climate, 24, 4676–4694. doi: 10.1175/2011jcli4188.1.
    [142]

    Wheeler, M. C., and H. H. Hendon, 2004: An all-season real-time multivariate MJO Index: Development of an index for monitoring and prediction. Mon. Wea. Rev., 132, 1917–1932. doi: 10.1175/1520-0493(2004)132<1917:Aarmmi>2.0.Co;2.
    [143]

    Wu, C.-H., and H.-H. Hsu, 2009: Topographic influence on the MJO in the Maritime Continent. J. Climate, 22, 5433–5448. doi: 10.1175/2009jcli2825.1.
    [144]

    Wu, M. C., and J. C. L. Chan, 1995: Surface features of winter monsoon surges over South China. Mon. Wea. Rev., 123, 662–680. doi: 10.1175/1520-0493(1995)123<0662:Sfowms>2.0.Co;2.
    [145]

    Xie, P. P., and P. A. Arkin, 1997: Global precipitation: A 17-year monthly analysis based on gauge observations, satellite estimates, and numerical model outputs. Bull. Amer. Meteor. Soc., 78, 2539–2558. doi: 10.1175/1520-0477(1997)078<2539:GPAYMA>2.0.CO;2.
    [146]

    Xu, Q., and Z. Y. Guan, 2017: Interannual variability of summertime outgoing longwave radiation over the Maritime Continent in relation to East Asian summer monsoon anomalies. J. Meteor. Res., 31, 665–677. doi: 10.1007/s13351-017-6178-3.
    [147]

    Xu, Q., Z. Y. Guan, D. C. Jin, et al., 2019: Regional characteristics of interannual variability of summer rainfall in the Maritime Continent and their related anomalous circulation patterns. J. Climate, 32, 4179–4192. doi: 10.1175/JCLI-D-18-0480.1.
    [148]

    Yang, G.-Y., and J. Slingo, 2001: The diurnal cycle in the tropics. Mon. Wea. Rev., 129, 784–801. doi: 10.1175/1520-0493(2001)129<0784:Tdcitt>2.0.Co;2.
    [149]

    Yasunari, T., 1991: The monsoon year — A new concept of the climatic year in the tropics. Bull. Amer. Meteor. Soc., 72, 1331–1338. doi: 10.1175/1520-0477(1991)072<1331:Tmynco>2.0.Co;2.
    [150]

    Zhang, C. D., and J. Ling, 2017: Barrier effect of the Indo–Pacific Maritime Continent on the MJO: Perspectives from tracking MJO precipitation. J. Climate, 30, 3439–3459. doi: 10.1175/jcli-d-16-0614.1.
    [151]

    Zhang, C. D., M. Dong, S. Gualdi, et al., 2006: Simulations of the Madden–Julian oscillation in four pairs of coupled and uncoupled global models. Climate Dyn., 27, 573–592. doi: 10.1007/s00382-006-0148-2.
    [152]

    Zhang, C. D., J. Gottschalck, E. D. Maloney, et al., 2013: Cracking the MJO nut. Geophys. Res. Lett., 40, 1223–1230. doi: 10.1002/grl.50244.
    [153]

    Zhang, T. T., S. Yang, X. W. Jiang, et al., 2016a: Roles of remote and local forcings in the variation and prediction of regional Maritime Continent rainfall in wet and dry seasons. J. Climate, 29, 8871–8879. doi: 10.1175/jcli-d-16-0417.1.
    [154]

    Zhang, T. T., S. Yang, X. W. Jiang, et al., 2016b: Sub-seasonal prediction of the Maritime Continent rainfall of wet–dry transitional seasons in the NCEP climate forecast version 2. Atmosphere, 7, 28. doi: 10.3390/atmos7020028.
    [155]

    Zhang, T. T., S. Yang, X. W. Jiang, et al., 2016c: Seasonal–interannual variation and prediction of wet and dry season rainfall over the Maritime Continent: Roles of ENSO and monsoon circulation. J. Climate, 29, 3675–3695. doi: 10.1175/jcli-d-15-0222.1.
    [156]

    Zhang, T. T., B. H. Huang, S. Yang, et al., 2018a: Predictable patterns of the atmospheric low-level circulation over the Indo–Pacific region in project Minerva: Seasonal dependence and intraensemble variability. J. Climate, 31, 8351–8379. doi: 10.1175/jcli-d-17-0577.1.
    [157]

    Zhang, T. T., B. H. Huang, S. Yang, et al., 2018b: Seasonal dependence of the predictable low-level circulation patterns over the tropical Indo–Pacific domain. Climate Dyn., 50, 4263–4284. doi: 10.1007/s00382-017-3874-8.
    [158]

    Zhang, T. T., B. H. Huang, S. Yang, et al., 2018c: Dynamical and thermodynamical influences of the Maritime Continent on ENSO evolution. Sci. Rep., 8, 15352. doi: 10.1038/s41598-018-33436-5.
    [159]

    Zhang, Y., K. R. Sperber, and J. S. Boyle, 1997: Climatology and interannual variation of the East Asian winter monsoon: Results from the 1979–95 NCEP/NCAR reanalysis. Mon. Wea. Rev., 125, 2605–2619. doi: 10.1175/1520-0493(1997)125<2605:Caivot>2.0.Co;2.
    [160]

    Zhao, S. Y., S. Yang, Y. Deng, et al., 2015: Skills of yearly prediction of the early-season rainfall over southern China by the NCEP Climate Forecast System. Theor. Appl. Climatol., 122, 743–754. doi: 10.1007/s00704-014-1333-6.
    [161]

    Zhu, J. S., W. Q. Wang, and A. Kumar, 2017: Simulations of MJO propagation across the Maritime Continent: Impacts of SST feedback. J. Climate, 30, 1689–1704. doi: 10.1175/jcli-d-16-0367.1.
  • 加载中
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Climate Variability over the Maritime Continent and Its Role in Global Climate Variation: A Review

    Corresponding author: Tuantuan ZHANG, tuantuanzhang@cuhk.edu.hk
  • 1. School of Atmospheric Sciences, Sun Yat-sen University, Guangzhou 510275
  • 2. Guangdong Province Key Laboratory for Climate Change and Natural Disaster Studies, Sun Yat-sen University, Guangzhou 510275
  • 3. Southern Laboratory of Ocean Science and Engineering (Zhuhai, Guangdong), Zhuhai 519082
  • 4. Institute of Environment, Energy and Sustainability, The Chinese University of Hong Kong, Hong Kong 999077
  • 5. Climate Center of Guangdong Province, Guangzhou 510080
Funds: Supported by the Vice-Chancellor’s Discretionary Fund of the Chinese University of Hong Kong (4930744), and National Natural Science Foundation of China (41661144019, 91637208, 41690123, and 41690120)

Abstract: The Maritime Continent (MC) consists of multiple islands with varying sizes and topography, and surrounding seas. It is characterized by rainfall (convection) variability on multiple spatial and temporal scales. Various large-scale atmospheric, oceanic, and coupled climate systems, such as the El Niño–Southern Oscillation (ENSO), Indian Ocean Dipole (IOD), Madden–Julian Oscillation (MJO), and cold surge, exert significant influences on the spatio-temporal complexity of the MC climate and climate variability. As a major tropical heat source located within the warmest oceanic area (the western Pacific warm pool), the MC has been identified as a region of great importance for climate variation on the global scale. However, prediction of climate variability over the MC and its surrounding areas and the relationships to large-scale atmospheric circulation patterns are big challenges, even for state-of-the-art climate models. In this paper, we provide a thorough review on current understanding of the spatiotemporal complexity and prediction of climate variability over this important region, and its influence on global climate variation.

1.   Introduction
  • The Maritime Continent (MC) in Southeast Asia consists of a group of islands with varying sizes and topography, including Sumatra, Java, Borneo, New Guinea, and their surrounding seas, making up the largest archipelago on earth (see Fig. 1; Ramage, 1968). It is a “land bridge” connecting the continental areas of East/Southeast Asia and Australia, and is located in the core of the world’s strongest monsoon region (Lau et al., 1983; Matsumoto, 1992; Chang et al., 2005b). Surrounded by the warmest seas (as part of the western Pacific warm pool), the climate of the MC is characterized by abundant supply of moisture, large amount of latent heat release, and atmospheric deep convection, which fuel the ascending branches of both the Walker cell and the local Hadley cell. This fundamental role of the MC as a “boiler box” in the atmospheric circulation was outlined in Ramage (1968).

    Figure 1.  Topography of the MC. Major islands are labeled. Topography height data are from the Global Multi-resolution Terrain Elevation Dataset 2010 (GMTED2010).

    Climate variation over the MC is of great importance for agriculture productivity of this densely populated region, and hence socio-economic wellbeing of the inhabitants. The climate of the MC is determined by the impacts of various atmospheric systems on a range of temporal and spatial scales, such as the El Niño–Southern Oscillation (ENSO; planetary scale and mainly interannual variation), Indian Ocean Dipole (IOD; mainly interannual variation), Madden–Julian Oscillation (MJO; intraseasonal variation), and cold surge (synoptic scale; Chang and Lau, 1980; Chang et al., 1979, 2004, 2005a; Hendon, 2003; Peatman et al., 2014; Zhang et al., 2016c). For example, the 1997/98 El Niño event created a drought condition over the MC and a weak, late arrival of monsoonal rainfall, which caused large and long-lasting forest and peat fires, resulting in severe environmental and societal consequences. In addition, interactions between systems on different spatial and temporal scales have been considered particularly important to the mean climate over the MC. For example, as a dominant mode of intraseasonal variability over the tropics, the MJO can modulate the intraseasonal rainfall variability over the MC, and affect synoptic-scale cold surge and vortex circulations, as well as the diurnal cycle over the region (Chang et al., 2005a). On the other hand, the MJO tends to be blocked by the diurnal cycle when it propagates across the MC (Oh et al., 2013).

    To improve prediction of climate variation on various timescales, climate forecast systems have been established at major climate prediction centers worldwide. However, predicting the MC rainfall variation is always a huge challenge even for state-of-the-art climate models including the U.S. NCEP Climate Forecast System version 2 (NCEP CFSv2), due to the complex topography over this region and among others (Jiang et al., 2013; Zhang et al., 2016c). These models generally show large systematic bias, particularly over the western MC, which can be a driver for other systematic errors through Rossby waves (Neale and Slingo, 2003). Increases in horizontal and vertical resolutions, which contribute to more realistic simulations of terrains and surface types and better representation of physical processes, are of great importance for improving the prediction of MC rainfall (Yang and Slingo, 2001; Neale and Slingo, 2003; Wang et al., 2007; Schiemann et al., 2014; Johnson et al., 2016; Rashid and Hirst, 2017).

    This paper is a review of our understanding of the variation and prediction skills of the climate over the MC and the influence of the regional climate on global climate variation. In Section 2, we describe the general features of MC climate. Impacts of various atmospheric systems on MC rainfall variation are summarized in Section 3. We review prediction and predictability of the MC climate on various timescales in Section 4, and discuss the influences of the MC on large-scale climate variation in Section 5. A summary is given in Section 6.

2.   Climatological features of rainfall over the MC
  • Figure 2 presents the monthly-mean rainfall and 850-hPa wind from January to December over 1982–2017. We can see that the MC is dominated by a convergence of low-level winds with heavy rainfall from November to April. In May, the southwesterly monsoonal winds start to build up over the northern Indian Ocean (IO) and East Asia, with rainfall decreasing over the MC and increasing to the north. The intertropical convergence zone (ITCZ) shifts northward in the subsequent months. Heavy rainfall gradually returns to the MC region during the boreal fall, along with a reversal of the low-level monsoonal wind (Fig. 2). The annual evolution of MC rainfall is clearly characterized by a dry season and a wet season, with an asymmetric seasonal transition between the two (Zhang et al., 2016c).

    Figure 2.  (a–l) Spatial patterns of monthly climatological rainfall (shading; mm day–1) and 850-hPa wind (vector; m s–1) from January to Decenmber over 1982–2017. The reference vector of the 850-hPa wind is indicated on the top right of each panel. Rainfall data are from the Global Precipitation Climatology Project version 2.3 Combined Precipitation Data Set (GPCP; Adler et al., 2003), and wind data are from the ERA-Interim (Dee et al., 2011).

    The MC wet season and dry season are November–March and May–September, respectively, according to Aldrian and Susanto (2003). Using detailed monthly evolution of rainfall, Zhang et al. (2016c) defined December–March (July–October) as the wet (dry) season. For most of the MC region, the major wet (dry) season coincides with the boreal winter (summer) monsoon (McBride and Nicholls, 1983; Haylock and McBride, 2001; Chang et al., 2004, 2005b). In the wet season, heavy rainfall with warm sea surface temperature (SST) occurs in the MC region, accompanied by northeasterlies to the north of the equator and northwesterlies to the south of the equator, with the development of the Asian winter monsoon (Chang et al., 2005b; Kida and Richards, 2009). Rainfall in the heart of the wet season tends to be spatially incoherent (Hendon, 2003). On the other hand, in the dry season only moderate rainfall occurs over the MC, accompanied by southwesterlies to the north of the equator and southeasterlies to the south of the equator, with development of the Asian summer monsoon (Chang et al., 2005b; Zhang et al., 2016c). Due to the intensified cross-equatorial southerly flow, the local SST becomes lower, especially around the Banda and Arafura seas, suppressing convection over the MC during the dry season (Aldrian and Susanto, 2003; Kida and Richards, 2009).

    Typically, the MC experiences a sudden transition from the wet season to the dry season during the boreal spring, and a gradual transition from the dry season to the wet season during the boreal fall (Zhang et al., 2016b). This asymmetric seasonal march of rainfall is consistent with the migration of convection (e.g., Meehl, 1987; Yasunari, 1991; Matsumoto, 1992; Matsumoto and Murakami, 2000, 2002). During the boreal spring, maximum convection and rainfall remain mostly near and to the south of the equator, and jump northward to 5°N after May (e.g., Tanaka, 1994; LinHo and Wang, 2002). During the boreal fall, the maximum convection does not retrace the process, but experiences a gradual southeastward progression from the north of the equator to the south, in a track that roughly follows the Southeast Asian “land bridge” (e.g., Lau and Chan, 1983a, b; Meehl, 1987; Yasunari, 1991; Matsumoto, 1992).

    The cause of this asymmetric seasonal march during the transitional seasons (boreal spring and fall) has not been fully understood. Several studies suggested that the cold surge originated from East Asia during the boreal spring is stronger than that originated from Australia during the boreal fall, and that the respective annual variations of the Kelvin-type basic flow over the western Pacific and the Rossby-type mean flow over the IO help convection to move southeastward during the boreal fall but inhibit its return journey during the boreal spring (Matsumoto and Murakami, 2000, 2002). Hung et al. (2004) proposed that the discontinuity of the ITCZ movement during the boreal spring and the seasonal heating over the Asian continent can result in this asymmetry. The ITCZ is confined to south of 5°N until the onset of the Asian summer monsoon, compared to its continuous southward propagation during the boreal fall (Hung et al., 2004). Moreover, Chang et al. (2005b) demonstrated that at least part of this asymmetry was related to the difference in sea level pressure between spring and fall, as a result of the global-scale mass redistribution driven by different thermal memories between the land and ocean. It should be pointed out that previous studies mainly focused on the wet season or dry season rainfall variability over the MC; a clear theoretical perspective for the asymmetric seasonal march during the transitional seasons has been largely ignored. The exact roles of the ITCZ, the landmass, the low-level wind and terrain interaction, or other processes and systems in the seasonal evolution over the MC warrant further investigations.

  • Rainfall variation shows distinct characteristics in different areas of the MC, due to the complex terrains and land–sea distribution. The modes of annual and semi-annual cycles are the first two harmonics of the variation of climatologically averaged annual rainfall (Hamada et al., 2002; Aldrian and Susanto, 2003; Chang et al., 2005b). The annual cycle mode of the rainfall dominates over most of the MC region, which is influenced largely by the interaction between complex terrains and a simple annual reversal of surface monsoonal winds (Chang et al., 2005b). On the other hand, the semi-annual cycle mode, which represents the southward and northward movements of the ITCZ, is influenced by both summer and winter monsoons, and the twice yearly crossing of the sun over the region (Chang et al., 2005b). Details of the relationship between MC rainfall and monsoons will be presented in Section 3.

    Based on the amplitude and phase of annual rainfall oscillation at 46 stations during 1961–90, Hamada et al. (2002) classified the MC into four climatological regions. (1) The annual cycle with rainfall maximum in the boreal fall/winter is dominant in most of the MC region, which mainly appears on the south side of 5°S (type A-I). (2) The semi-annual cycle with maximum rainfall in the boreal fall mainly appears in western Kalimantan and equatorial Sumatra (type A-II). (3) The annual cycle with rainfall maximum in the boreal spring/summer is distributed in eastern Kalimantan and Maluku (type B-I). (4) Rainfall over the eastern equatorial region is abundant or poor throughout the year (no clear wet and dry seasons; type C). By applying “double correlation method,” Aldrian and Susanto (2003) identified three clearly distinctive sub-regions, which were roughly consistent with the type A-I, type A-II, and type B-I in Hamada et al. (2002) (see Fig. 2 in Aldrian and Susanto, 2003). Chang et al. (2005b) reported similar results, except that the semi-annual cycle mode also appeared over northeastern Borneo between 114° and 120°E, and over the equatorial region between 120° and 130°E. Overall, the annual and semi-annual cycle modes are the two major harmonics for climatological rainfall variation in the MC region, and the occurrence of maximum rainfall mainly depends on the direction of monsoonal wind, local terrain, and seasonal march of the ITCZ.

3.   Influences of large-scale climate systems on MC rainfall
  • Considerable interannual variability of the MC rainfall and its close relationship with the ENSO have been well recognized (e.g., McBride and Nicholls, 1983; Hendon, 2003; Chang et al., 2004). For example, the regional rainfall decreases during El Niño years while increases during La Niña years (e.g., Aldrian and Susanto, 2003; Chang et al., 2004; Kubota et al., 2011).

    Figure 3 shows the patterns of regressions of SST/rainfall/850-hPa wind against the MC rainfall in wet and dry seasons, respectively (Zhang et al., 2016c). In the wet season, increased MC rainfall is associated with negative SST anomalies over the equatorial IO–western MC and the central–eastern Pacific (EP), with V-shaped positive SST anomalies in between (Figs. 3a, c). This SST anomaly pattern enhances the SST gradient in the tropical Pacific and IO, resulting in strengthened Walker circulations (Figs. 3a, c). These features bear a large resemblance to those for the peak-decaying phase of the ENSO. In the dry season, increased MC rainfall is associated with negative SST anomalies over the central–eastern equatorial Pacific, with positive SST anomalies from the whole MC region to the western South Pacific, which strengthen the Walker circulations, resembling the features of ENSO developing phase (Figs. 3b, d). An opposite condition tends to occur for the decreased MC rainfall. The correlation coefficients between the MC rainfall index and Niño3.4 index in wet and dry seasons are −0.83 and −0.85, respectively. This significant relationship between interannual variation of MC rainfall and ENSO has been well demonstrated (Hamada et al., 2002; Hendon, 2003; McBride et al., 2003; Chang et al., 2004; Kubota et al., 2011).

    Figure 3.  Regressions of (a, b) SST (shading; K)/850-hPa wind (vectors; m s−1) and (c, d) rainfall (shading; mm day−1)/850-hPa wind (vectors; m s−1) against the MC rainfall index in (a, c) the wet season (December–March) and (b, d) the dry season (July–October) for the period of 1982–2017. The reference vector of the 850-hPa wind is indicated on the top right in each panel. Stippling (denoted by purple dots) indicates regression coefficients for SST and rainfall (shading) exceeding the 95% confidence level. Only the regressed winds exceeding the 95% confidence level are drawn. The domain of the MC is outlined by a red box. Modified from Zhang et al. (2016c).

    According to Figs. 3b, d, the rainfall variation in the dry season is spatially coherent and is positively correlated with the regional SST over the entire MC region. During the dry season of El Niño developing years, the MC region is dominated by anomalous descending branches of the Pacific and IO Walker circulations (Aldrian and Susanto, 2003; Hendon, 2003). In the meantime, the El Niño-induced low-level cyclonic vorticity over the western Pacific leads to strengthened western North Pacific monsoon trough (Kubota et al., 2011; Jiang and Li, 2018). As a result, the local Hadley circulation strengthens, leading to anomalous descending motion over the MC and anomalous ascending motion to the north (Jiang and Li, 2018). In addition, the cross-equatorial southeasterly wind north of Australia becomes stronger, enhancing evaporation and hence inducing cold SST anomalies over the MC region (Hendon, 2003; Kubota et al., 2011). The cold SST anomalies then increase the anomalous SST gradient across the Pacific, and further intensify the anomalous Walker circulation in the Pacific (Hendon, 2003). A nearly mirrored image of the anomalies of the Walker circulation and local Hadley circulation occurs during the dry season of developing La Niña years (Hendon, 2003; Jiang and Li, 2018). The combined effect of the anomalous Walker circulation and local Hadley circulation driven by local and remote SSTs leads to a homogeneous response of the whole MC region to ENSO (Hendon, 2003; Zhang et al., 2016a; Jiang and Li, 2018).

    In the wet season, the rainfall variation over the MC is spatially incoherent, especially for the western MC (see the insignificant correlation around Sumatra and Malay in Fig. 3c). These seasonal and regional features have been reported previously (e.g., Hendon, 2003; Chang et al., 2004; Jiang and Li, 2018). In particular, the eastern MC rainfall variation was significantly modulated by both ENSO and local SST in the wet season, while the western MC rainfall variation was only moderately correlated with the ENSO (Aldrian and Susanto, 2003; Chang et al., 2004; Zhang et al., 2016a). Strong atmospheric forcing may play a role over the western MC (Aldrian and Susanto, 2003; Zhang et al., 2016a). Hendon (2003) suggested that during the wet season of El Niño (La Niña) years, the climatological mean northwesterlies north of Australia become weaker (stronger) and the SST near the MC region increases (decreases) due to reduced (enhanced) evaporation. The local SST anomalies tend to reduce the anomalous SST gradient across the Pacific and hence weaken the anomalous Walker circulation there, which is opposite to the effect produced by remote SST (i.e., the SST anomalies in the central–eastern equatorial Pacific) and contributes to a weaker, non-uniform response of MC rainfall to the ENSO (Hendon, 2003). However, several studies argued that, instead of the local SST effect, the topographic lifting effect may play a major role (Chang et al., 2004; Rauniyar and Walsh, 2013; Jiang and Li, 2018). Jiang and Li (2018) proposed a physical mechanism for the weaker and non-uniform response of rainfall over the MC to the ENSO during the wet season: the anomalous northeasterlies from the southern flank of anomalous western North Pacific anticyclone established during October–November of El Niño developing years and the anomalous southeasterlies from the Java Sea associated with anomalous Walker circulation tend to converge over western Borneo, leading to anomalous ascending motion. Meanwhile, the easterly anomalies (consisting of anomalous northeasterlies and southeasterlies) are lifted by the high mountains over Sumatra and western Borneo, inducing anomalous ascending motion (Chang et al., 2004). The anomalous ascending motion advects the climatological mean moisture upward, moistening the lower troposphere in situ, and eventually setting up a convectively unstable stratification. As a consequence, rainfall increases over the western MC in the wet season. In contrast, negative rainfall anomalies always appear over the eastern MC during the wet season because of the anomalous decent throughout the troposphere. An opposite condition in the wet season tends to occur during La Niña.

    There appear diverse impacts of ENSO on MC rainfall variation. Jia et al. (2016) found that significant negative rainfall anomalies occurred during the boreal winter of strong El Niño years, whereas positive anomalies occurred during the boreal winter of both strong and moderate La Niña years. However, there existed case-to-case variance of SST anomalies over the tropical Pacific associated with moderate El Niño events, leading to a weak MC rainfall response. Aldrian et al. (2007) reported that except for the northwestern region, the dry-season MC rainfall was more sensitive to El Niño events than to La Niña events. In addition to the asymmetric impact of ENSO, conventional ENSO, or EP ENSO, and ENSO-Modoki (Ashok et al., 2007), or by different names, such as the central Pacific (CP) ENSO by Kao and Yu (2009), exert distinctive impacts on the rainfall variability over the MC, due to difference in anomalous Walker circulation over the tropical Pacific and regional SST anomalies induced by the two types of ENSO (Feng et al., 2010; Salimun et al., 2014; As-syakur et al., 2016; Wang et al., 2018). However, to what extent the conventional ENSO and ENSO-Modoki can affect the rainfall variability over the MC is still being debated. As-syakur et al. (2016) indicated a stronger and wider impact of conventional ENSO than of ENSO-Modoki on MC rainfall during the boreal winter and spring, but similar impacts by the two during the boreal fall and summer seasons. Feng et al. (2010) also found a much stronger influence of conventional ENSO than of ENSO-Modoki on the MC during the boreal winter (see Figs. 2, 3). Other studies, however, argued that the rainfall over central–eastern Indonesia during summer and over Malaysia during winter was more closely linked to ENSO-Modoki than to conventional ENSO (Salimun et al., 2014; Wang et al., 2018; Xu et al., 2019). The inconsistency of these results might be attributed to different datasets and analysis methods applied, which warrants further investigations.

  • The IOD, which refers to strong zonal SST gradient in the equatorial IO region, is also believed to exert a great impact on the interannual variability of MC rainfall (Saji et al., 1999; Ashok et al., 2003; Hamada et al., 2012; As-syakur et al., 2014). Generally, IOD events are phase-locked to boreal summer and autumn, and closely linked to ENSO events (Saji et al., 1999).

    During the dry season, cold (warm) SST anomalies induced by positive (negative) IOD events prevail over the west of the Indonesian archipelago, causing anomalous subsidence (ascent) and large-scale divergence (convergence), and lower (higher) atmospheric moisture content, hence suppressing (inducing) rainfall over the MC (Ashok et al., 2003; Hamada et al., 2012). On the other hand, the response of MC rainfall to the IOD during the wet season is only moderate. The IOD effect on rainfall fluctuation was found only over the northern and northeastern parts of Indonesia during the boreal winter, with an even smaller effect during the boreal spring (As-syakur et al., 2014). Hamada et al. (2012) found that the interannual variation of rainfall over northwestern Java during the wet season was not closely related to the IOD, and that rainfall tended to be abundant in non-ENSO/IOD years instead.

    As-syakur et al. (2014) applied a partial correlation to remove ENSO influence, based on the fact that many IOD and ENSO events occur together. When the ENSO influence is removed, the IOD only affects the southern part of Indonesia (i.e., southeastern part of Sumatra) during the boreal summer and fall (As-syakur et al., 2014). On the other hand, during the boreal winter, the negative correlation in the northern part of Maluku Island weakens, but significantly positive correlation appears in the southeastern part of Indonesia (As-syakur et al., 2014). There is still generally insignificant correlation of MC rainfall with IOD during the boreal spring when the effect of ENSO is removed (As-syakur et al., 2014). It seems that the influence of the IOD on the interannual variation of MC rainfall is larger in the dry season than in the wet season. This may be linked to different air–sea interactions and fluctuations of the ITCZ during the two seasons; however, the underlying mechanisms need to be further investigated.

  • It is widely believed that the intraseasonal variability of MC rainfall is effectively controlled by deep convection associated with the MJO, a dominant component of the intraseasonal (20–90-day) variability over the tropics (Madden and Julian, 1972; Hidayat and Kizu, 2010; Kerns and Chen, 2016; Zhang and Ling, 2017). The initiation of the MJO generally involves the triggering of the envelope of convection over the western IO. It propagates eastward to the MC with a speed around 5 m s–1 (Madden and Julian, 1972, 1994). About half of the MJO events that formed over the IO propagated across the MC to the western Pacific (MJO-C), while the other half was blocked by the MC due to its barrier effect (MJO-B). Note that there is slightly higher percentage of MJO-C during the boreal winter than during the boreal summer (Hsu and Lee, 2005; Inness and Slingo, 2006; Wu and Hsu, 2009; Feng et al., 2015; Kerns and Chen, 2016; Zhang and Ling, 2017; Tan et al., 2018).

    An example for the canonical MJO-C event is given here, which is formed over the IO and propagates across the MC to the western Pacific during the boreal winter, with the eight MJO phases defined according to an MJO index by Wheeler and Hendon (2004). During phase 1, the center of the convective activity is located near Africa while the MC region is dominated by reduced convection, with easterly anomalies over the IO and possibly westerly anomalies over the Pacific (Wheeler and Hendon, 2004; Hidayat and Kizu, 2010; Zhang and Ling, 2017). Sequentially in phase 2, enhanced convection grows over the IO and moves eastward to the west of the MC, a region still dominated by suppressed convection. Easterly anomalies now prevail over the MC region. From phases 1–2, negative rainfall anomalies appear over most of the MC region, but near-zero and even positive rainfall anomalies are observed over Sumatra, western Borneo, Sulawesi, and New Guinea (Fig. 4; Peatman et al., 2014). During phase 3, active MJO (enhanced convection) reaches the western MC, and positive rainfall anomalies tend to reach their maxima over large land masses of the MC and the oceanic area of the western MC, whereas negative rainfall anomalies are observed over the seas of the eastern MC (Fig. 4). The organized convection over the MC reaches its maximum during phases 4–5, when strong westerly and easterly anomalies converge (Fig. 4; Burleyson et al., 2018). Most of the MJO-C events tend to detour southward when encountering the MC, and hence large positive rainfall anomalies dominate over the seas of the MC region especially over the Southern Hemisphere (Fig. 4). However, the rainfall over land eventually decreases during phases 4–5, and shows negative anomalies in phase 5 (Fig. 4). The convectional center further migrates to the eastern MC during phase 6, and rainfall anomalies show nearly opposite signs compared to those in phase 3 (Fig. 4). During phases 7–8, the center of the convective activity moves further to the Pacific, and the MC is mostly a convection-suppressed area (Fig. 4; Wheeler and Hendon, 2004; Hidayat and Kizu, 2010; Zhang and Ling, 2017). Moreover, negative rainfall anomalies cover most of the MC, with only slight anomalies over the land masses (Fig. 4; Peatman et al., 2014).

    Figure 4.  Daily-mean precipitation anomaly (shading; mm day−1) from Tropical Rainfall Measuring Mission (TRMM) 3B42HQ for each phase of the MJO (MJO 1–8) as defined in Wheeler and Hendon (2004). Phases move forward with time in the anti-clockwise direction. After Peatman et al. (2014).

    The impact of MJO-C is largely inhomogeneous between land and sea areas over the MC. Rainfall variability over the surrounding seas is more clearly controlled by MJO-C, while the impact of MJO-C on rainfall variability over the land is small and sometimes even negative (Fig. 4; Peatman et al., 2014). This land–sea contrasted rainfall variability related to the MJO-C life cycle is associated with the distinctive influence of the MJO on the diurnal cycle of rainfalls over land and seas (Fig. 4; Oh et al., 2012; Peatman et al., 2014, 2015; Birch et al., 2016). The impact of the MJO on the diurnal cycle of rainfall over the MC will be discussed in Section 3.4. Note that there exist differences in propagation patterns of MJO-C events between boreal winter and summer, with main MJO activities over the region south of the equator (i.e., the Java Sea, the Banda Sea, and the Timor Sea) in winter but over the region north of the equator (i.e., the South China Sea) in summer (Wu and Hsu, 2009; Zhang and Ling, 2017). Previous studies mainly focused on boreal-winter cases, while the modulation of the MJO on MC rainfall during summer warrants further investigations.

    The evolution of anomalous rainfall pattern is similar for MJO-C and MJO-B events before they move across Sumatra. The remarkable difference is that MJO-C maintains its amplitude in rainfall anomalies and propagates eastward, whereas MJO-B vanishes after propagated across the MC and leads to a much smaller impact on the intraseasonal variability of MC rainfall (Zhang and Ling, 2017). In comparison to MJO-C events, there is inhibiting convective development over the seas where rainfall never becomes dominant for MJO-B (Zhang and Ling, 2017). In addition, a “vanguard of rainfall” that increases over land more than over the seas prior to the arrival of MJO active phase is less pronounced for MJO-B than for MJO-C (Peatman et al., 2014; Zhang and Ling, 2017).

  • On the synoptic timescale, northeasterly cold surge often determines the low-level circulation pattern during the boreal winter over East Asia including the South China Sea, and accounts for a certain proportion of rainfall variability over the MC during the boreal winter (Chang et al., 1979; Chang and Lau, 1980, 1982; Lau et al., 1983; Wu and Chan, 1995; Chang et al., 2005a). Cold surge is due to the development of an anticyclone over the Siberia–Mongolia region (or the Siberia high), which results in midlatitude cold air penetrating deep into the tropics (including the MC region; Lau and Lau, 1984; Zhang et al., 1997; Slingo, 1998; Park et al., 2011; Li Q. P. et al., 2017). Typically, it is characterized by a rapid acceleration of low-level northeasterly wind and a drop of surface temperature, and lasts around 5–14 days.

    Following the onset of cold surge triggered by midlatitude baroclinic disturbances, strengthening of low-level northeasterly wind progresses rapidly equatorward around the eastern edge of the low-level East-Asian anticyclone. Prior to the temperature drop, increased northeasterly monsoon flow dominates over the South China Sea, contributing to an increased moisture convergence near the coastal area of Borneo (Chang et al., 1979; Chang and Lau, 1980, 1982; Lau et al., 1983; Zhang et al., 1997; Chang et al., 2005a). In addition, stronger northeasterly wind tends to enhance shear vorticity, resulting in a stronger Borneo vortex and a shift of the vortex center from the southern South China Sea to near the Borneo landmass (Chang et al., 2005a). Hence, as a response to the cold surge, organized deep cumulus convection and low-level convergence are enhanced near the southern South China Sea (including the MC), especially Borneo (e.g., Lau et al., 1983; Zhang et al., 1997). Chang et al. (2005a) reported that during cold surge days, low-level convergence took in a V-shaped pattern on the windward side of the Malay Peninsula and Borneo due to the blocking of surge winds by the terrain. This intensified low-level convection causes strengthening of the East-Asian local Hadley circulation and the Walker circulation (Chang and Lau, 1980, 1982; Lau et al., 1983). Then, rainfall over the MC tends to increase as a result of intensified low-level convection and rising motion (Zhang et al., 2016a; Abdillah et al., 2017).

    Cold surge also modulates rainfall variability over the MC by effecting the formation, development, and propagation of the parent cold surge vortex (CSV) of a heavy rainfall/flood (HRF) event (hereafter CSV(HRF); Chen et al., 2015a, b). According to several studies, cold surge flow has been classified into two types: the Philippine Sea type (the South China Sea type), which is identified by the 925-hPa meridional wind being weaker (stronger) than the 925-hPa zonal wind at the location of maximum isotach in the South China Sea (Compo et al., 1999; Chen et al., 2015a, b). The CSV(HRF) primarily forms over the vicinity of the Philippines and Borneo (Chen et al., 2015a, b). The Philippine CSV(HRF) is formed by the interaction of an easterly wave with the Philippine Sea type of cold surge flow and the surface island-chain trough, while the Borneo CSV(HRF) is formed by the interaction among the South China Sea type of cold surge flow, the near-equator trough, and the terrain of Borneo (Chen et al., 2015a, b). Westward propagation of the CSV(HRF) is facilitated by the Philippine Sea type (i.e., a Philippine CSV(HRF) developed into an HRF event over the Malaysia Peninsular) but hindered by the South China Sea type (i.e., a trapped Borneo CSV(HRF) developed into a Borneo HRF event; Chen et al., 2015a, b). After multiple interactions with sequential cold surges, a CSV(HRF) develops into an HRF event if the following two requirements are met: (1) simultaneous occurrence of the northwestern Pacific explosive cyclone with the HRF event and (2) simultaneous occurrence of the HRF event with the maximum speed of the westerlies (easterlies) of the northwestern Pacific explosive cyclone (tropical trade winds; Chen et al., 2015a, b).

  • Climatologically, heavy rainfall is concentrated over the MC islands, especially over the mountainous range surrounded by a relatively dry ring (Qian, 2008; Qian et al., 2010). Secondary heavy precipitation belts are observed in the middle of the seas between large islands. The diurnal cycle, which is pronounced over the tropics, may play a major role in maintaining these spatial distributions (Qian, 2008).

    Land heats up more quickly than the sea during the day time, due to different heat capacities between the two. Sea breezes develop as a response to the pressure gradient between land and sea. Similarly, upslope valley winds are initiated over the mountainous regions during day time. Sea breeze fronts reach the center of each island around afternoon to early morning, and act as a trigger to generate initial convection (Mori et al., 2004; Qian, 2008). The sea breezes are reinforced by the valley winds when they meet the mountain range. Once the sea breezes and valley winds converge over the island to initiate rainfall, self-amplifying cumulus merger process takes over to reinforce and maintain rainfall until the midnight (Holland and Keenan, 1980; Simpson et al., 1980, 1993; Keenan et al., 2000; Yang and Slingo, 2001; Qian, 2008). Thus, rainfall tends to peak around the afternoon or early evening over the islands. Note that rainfall is generally enhanced on the lee side of mountains and islands where sea breezes head against off-shore synoptic-scale low-level winds, referred to as the “wake effect” in Qian et al. (2013). On the other hand, land (mountain slope) cools off faster at night than the sea (valley), inducing land breezes (mountain winds). At the deep night or early morning, off-shore winds converge at the middle of the seas between two large islands, which generates secondary heavy rainfall (Qian, 2008). Land breezes (mountain winds) are generally weaker than sea breezes (valley winds) due to weaker land–sea temperature gradient during night time than during day time (night time cooling is weaker than day time heating), contributing to less heavy nocturnal/morning rainfall over the seas. In addition to the land–sea breezes and mountain valley winds, the gravity wave propagating from coastal mountains, and the cold pool generated from afternoon convection over the land, may also play an important role in the nocturnal or morning peak of rainfall over the seas (Mapes and Houze, 1992; Yang and Slingo, 2001; Mapes et al., 2003; Ciesielski and Johnson, 2006; Qian, 2008).

    There are also large seasonal and regional variations in the diurnal cycle. The magnitude of the diurnal cycle is the largest in the boreal winter (the wet season), followed by that in the boreal spring and fall (the transitional seasons), and is the smallest in the boreal summer (the dry season; Qian, 2008). The magnitude, phase speed, and timing of rainfall maximum are largely affected by the size of island (Nesbitt and Zipser, 2003; Qian, 2008; Sobel et al., 2011; Cronin et al., 2015). Figure 5 shows the diurnal cycle of rainfall over the MC region during the boreal winter. On small or narrow islands (e.g., Java and Timor), rainfall peaks around early afternoon; on medium-sized islands (e.g., Sumatra), rainfall peaks during late afternoon and early night; and on large islands (e.g., Borneo), it mainly peaks at the deep night [Fig. 5; also see Fig. 4 in Qian (2008)]. Using the data from the Tropical Rainfall Measuring Mission (TRMM) satellite precipitation radar, regional variation of the diurnal rainfall cycle over Sumatra was examined by Mori et al. (2004), which found that most convective clouds developed along the southwestern coastline of Sumatra between 1300 and 1600 LST (local standard time) and migrated to inland regions from 1600 to 2200 LST. The migration speed of convection/rainfall was approximately 10 m s–1. By applying principal component analysis, Teo et al. (2011) diagnosed the diurnal cycle of MC rainfall and found that the first two modes can explain the diurnal variability of MC rainfall. The first mode represents the overall differential peak of the stationary rainfall over land and seas, resulted from different responses of potential instability to surface heat flux and radiation. The second mode represents the propagating nocturnal rainfall peak, resulted from the interaction between gravity waves and mesoscale convection, and from density current or local thermally-induced circulation. The estimated phase speed of propagating rainfall signals is approximately 8–15 m s–1 over the open sea, but is only 4–6 m s–1 over flat land (Teo et al., 2011).

    Figure 5.  (a–f) Diurnal cycle of rainfall (mm day–1) from 0200 LST (local standard time; UTC + 8 h) to 0200 LST next day over the MC from TRMM 3B42 version 7 data for the borel winter of 1998–2014.

  • Previous studies documented the influence of MJO on the diurnal cycle of MC rainfall (e.g., Rauniyar and Walsh, 2011; Oh et al., 2012; Vincent and Lane, 2016). Peatman et al. (2014) found that when the mean diurnal cycle of rainfall was strong, the change in the amplitude of diurnal cycle was the dominant contributor to the daily-mean MJO rainfall. The amplitude of the diurnal cycle of rainfall anomalies is of the same order of the magnitude of the daily-mean rainfall anomalies with the respect to MJO phases.

    Consistent with the change in daily-mean rainfall, the diurnal cycle of rainfall over land reaches its maximum shortly before the large-scale active MJO envelope arrives at the MC, with a lead time of six days (one-eighth of an MJO cycle). At this MJO lead-up period (i.e., phases 1–2), the enhancement of downwelling shortwave radiation due to a relatively clear sky (destabilizing the atmosphere), the high land–sea temperature contrast maintained by high solar insolation (a strong onshore flow), and the strong frictional moisture convergence effect related to equatorial wave dynamics promote favorable conditions for the diurnal cycle of rainfall over the MC land, even though the large-scale convective environment is only moderately favorable for convection (Peatman et al., 2014; Hagos et al., 2016; Vincent and Lane, 2016). On the other hand, the diurnal cycle is weak over the seas, which is mainly controlled by large-scale environment and atmospheric stability to the east of the main MJO envelope at this stage (Birch et al., 2016).

    When MJO crosses the MC (i.e., phase 3 for the western MC; phases 4–6 for the eastern MC), cloudiness increases at all time of the day, resulting in a smaller amplitude of the diurnal cycle of surface heating and decreased surface insolation over the land, hence weakening the strength of the mesoscale circulation and the onshore flow (i.e., sea breeze; Birch et al., 2016; Hagos et al., 2016). Oh et al. (2012) suggested that during the boreal winter, strong westerlies contributed by the anomalous wind accompanying the MJO and the seasonal flow during the active phase of the MJO tended to interrupt convergence over the islands. Therefore, the diurnal cycle of rainfall over land is reduced, even though the large-scale environment is favorable for convection at the time (Birch et al., 2016). On the other hand, over the ocean, large-scale environment is the dominant control of convection, which favors morning rainfall and hence increases the diurnal cycle in situ. In the boreal winter, during the late active phase of the MJO (i.e., phase 5), latent heat flux is also a dominant control over the ocean (Wu and Hsu, 2009; Oh et al., 2012). Due to the blocking effect of orography, the prevailing westerlies (to the west of convective MJO) superimposed by the monsoonal winds increase mainly over Java (Oh et al., 2012). In addition, the near-surface winds induced by the flow bifurcation near mountainous islands converge mainly over the oceanic channels between two islands (Wu and Hsu, 2009; Oh et al., 2012). As a result, convection is fed through enhanced heat flux release induced by increased surface wind (the wind-induced surface heat exchange mechanism) and may contribute to heavy rainfall over the ocean (i.e., the Java Sea) in the morning (Emanuel, 1987; Neelin et al., 1987; Wu and Hsu, 2009; Oh et al., 2012).

    When the MJO is over the MC (i.e., phases 7–8), the large-scale environment becomes unfavorable (i.e., moisture flux divergence, negative equivalent potential temperature anomaly, and stable profile) for convection (Birch et al., 2016). At this stage, the diurnal cycle of rainfall over the ocean is significantly weakened (Oh et al., 2012; Peatman et al., 2014; Birch et al., 2016). However, solar insolation and downwelling shortwave radiation increase due to decreased cloudiness, again leading to enhanced onshore flow. Thus, the diurnal cycle of rainfall over the islands is only slightly weakened (Oh et al., 2012).

4.   Prediction of MC rainfall variation
  • Predicting MC climate is always a huge challenge for climate modelers due to the complex system of islands and shallow seas. In spite of great efforts, state-of-the-art climate models generally have large biases over the MC (Neale and Slingo, 2003; Jiang et al., 2013; Zhang et al., 2016a, c; Nguyen et al., 2017).

    The prediction skills of MC rainfall vary with seasons and regions (e.g., Neale and Slingo, 2003; Jiang et al., 2013; Argüeso et al., 2016). Figure 6a shows the correlations between MC rainfall in observations and that by the NCEP CFSv2 hindcast, observed Niño3.4 SST, and predicted Niño3.4 SST for different time leads (Zhang et al., 2016c). The interannual variability of MC rainfall in the dry season is highly predicted by the CFSv2, with correlation coefficients exceeding the 99% confidence level for all lead months (solid red line in Fig. 6a). The skill using the Niño3.4 index (both in observation and CFSv2) to predict dry-season MC rainfall is lower than that in the CFSv2 hindcast (red lines in Fig. 6a). In comparison, the prediction skill of CFSv2 is relatively lower for wet-season MC rainfall, which decreases rapidly from 0-month lead to 2-month lead (black solid line in Fig. 6a). The skill of the model in predicting wet-season MC rainfall is even lower than that of using the Niño3.4 index (both in observation and CFSv2) when lead time is longer than 0 months, suggesting a large bias of the CFSv2 in simulating the response of MC rainfall to ENSO during that season. Figure 6b depicts the monthly correlation coefficients between MC rainfall in observation and that predicted by the CFSv2 for different leads; it shows higher prediction skills during the dry-season months and lower skills during the wet-season months. This relatively lower prediction skill of MC rainfall in the wet season was reported by other investigators (Aldrian et al., 2007; Jiang et al., 2013). Aldrian et al. (2007) indicated that the prediction skill of MC rainfall in the ECHAM4 model was the highest in the boreal summer, followed by that in the boreal fall, boreal winter, and was the lowest in the boreal spring (the spring barrier).

    Figure 6.  (a) Correlation between the MC (15°S–10°N, 95°–145°E) rainfall in the Climate Prediction Center Merged Analysis of Rainfall (CMAP; Xie and Arkin, 1997) and that predicted by the NCEP CFSv2 hindcast (solid lines with circles), observed Niño3.4 (dashed lines), and predicted Niño3.4 (dotted lines) for different lead time (month) in dry season (red lines) and wet season (black lines). The horizonal dotted line in (a) denotes the 95% confidence level. (b) Monthly correlation coefficients between the MC rainfall in CMAP and that predicted by CFSv2 for different time lead (month). Shadings in (b) denote values exceeding the 95% confidence level. After Zhang et al. (2016c).

    Zhang et al. (2016a) reported distinctive prediction skills of rainfall over the western MC and eastern MC in the NCEP CFSv2. The result of 1-month lead shows high skill for the entire MC rainfall in the dry season but lower skill over parts of the MC, especially New Guinea and the western MC, in the wet season. The prediction skill of rainfall over the western MC during December–June is much lower than that during July–November. On the other hand, for the eastern MC the difference in prediction skill between wet- and dry-season months is not as remarkable as that for the western MC. The roles of remote forcing (i.e., ENSO) and local forcing (i.e., local SST) have been explored to explain these differences in prediction skills of rainfall over the western and eastern MC regions during wet and dry seasons. The results indicate that the relationships of the western MC rainfall with both local and remote forcing are unrealistically predicted by the CFSv2, contributing to lower prediction skill for the western MC rainfall in the wet season. On the other hand, the relationship of the entire MC rainfall with both local SST and remote ENSO is well captured by the CFSv2 for the dry season; hence, the model shows higher prediction skill for both western and eastern MC in that season. By evaluating the atmospheric model results of the Australian Community Climate and Earth System Simulator (ACCESS1.3), Nguyen et al. (2017) found that the model rainfall errors in December–February were characterized by a wet bias over the eastern MC and a dry bias over the western MC. They also revealed that the ACCESS1.3 overestimated the rainfall over the MC islands and underestimated the rainfall over the surrounding seas. The distinctive prediction skills over the islands and the seas were also shown in Zhang et al. (2016a) (Fig. 1).

    It is hypothesized that the systematic bias of MC rainfall prediction, which has been found in most general circulation models (GCMs), is mainly due to the inaccurate representation of the region’s geographical detail (i.e., due to coarse model resolution) and physical processes (i.e., convection; Neale and Slingo, 2003; Ruti et al., 2006; Schiemann et al., 2014; Johnson et al., 2016; Nguyen et al., 2017; Rashid and Hirst, 2017; Toh et al., 2018). Conducting experiments using a climate version of the Met Office model (HadAM3), Neale and Slingo (2003) showed that even a threefold increase in horizontal resolution could not reduce the dry bias over the MC. They argued that the deficiency in representing the physical system was the main reason. However, recent studies using the newer version of the Met Office Unified Model suggested an improvement in simulating MC rainfall with increased horizontal resolutions (Schiemann et al., 2014; Johnson et al., 2016; Rashid and Hirst, 2017). Rashid and Hirst (2017) indicated that a better representation of the orography due to increased horizontal resolution (from ~135 to 60 km) provided a stronger topographic lifting effect, leading to increased rainfall over the MC. Schiemann et al. (2014), however, attributed the improvement to the specification of better resolved surface boundary conditions (land fraction, soil and vegetation parameters) at higher resolutions (~110 km). Ruti et al. (2006) found a decrease in dry bias over the MC with increasing vertical resolution, which had a beneficial effect on the convective scheme performance and on related dynamic fields.

  • In general, the variance associated with slowly-evolving boundary conditions represents the predictable component whereas that with internal flow instabilities represents the chaotic part (Charney and Shukla, 1981). Therefore, Haylock and McBride (2001) suggested that for the interannual rainfall variation of a region to be predictable, it must be associated with the slowly-evolving boundary conditions, which are inherently large scale in nature. The authors indicated that the wet-season rainfall over Indonesia was inherently unpredictable, consistent with the low prediction skill of the wet-season regional rainfall. Kumar et al. (2013) reported that the prediction skill of rainfall was generally high only over the regions where slow ocean variability drove the atmosphere (i.e., large positive SST–precipitation correlation) and was low where atmospheric variability drove the ocean variability (i.e., weakly negative SST–precipitation correlation). Hence, there is low prediction skill of western MC rainfall during the wet season when atmospheric variability drives the ocean variability.

    The ENSO is a major source of seasonal predictability of tropical climate variation, and a better prediction and simulation of MC rainfall is found during ENSO years (Aldrian et al., 2007; Zhang et al., 2016c). The empirical orthogonal function (EOF) analysis with maximized signal-to-noise ratio (MSN EOF hereafter) has been applied to identify the most predictable patterns, and was shown as an effective statistical technique to extract predictable signals from an ensemble of hindcasts (Allen and Smith, 1997; Venzke et al., 1999; Chang et al., 2000; Sutton et al., 2000; Huang, 2004; Liang et al., 2009). By applying this method, the most predictable patterns of 850-hPa wind surrounding the MC have been analyzed by Zhang et al. (2018a, b). They showed that the most predictable patterns of the interannual variation of low-level winds were associated with ENSO developing and maturing phases, and could be well predicted by the NCEP CSFv2 and Project Minerva [a collaborative pro-ject between the Center for Ocean–Land–Atmosphere Studies (COLA) and the ECMWF] models. Furthermore, this result is not sensitive to models, at least for the NCEP and ECMWF models (Zhang et al., 2018a, b).

    The second most predictable patterns, on the other hand, show large seasonality, which is mainly related to the ENSO decaying phase from winter to spring, to the western North Pacific monsoon (ENSO) in shorter (longer) leads of summer, and to the IOD in fall (Zhang et al., 2018b). The model shows the highest skills in depicting the predictable patterns related to ENSO and the lowest skills for those related to the IOD, which is partly due to the seasonal nature (i.e., large seasonality) of the IOD (Zhang et al., 2018b).

  • While most studies have focused on the prediction of MC rainfall (or convection) in wet and dry seasons, Zhang et al. (2016b) investigated how well the rainfall during the transitional seasons be predicted by dynamic models. The date of transition from dry to wet seasons (DTW) or wet to dry seasons (WTD) was defined as the pentad that satisfies: (1) the first pentad when MC rainfall is higher (lower) than the annual mean during the 49th–69th (17th–37th) pentads; (2) the MC rainfall in the three consecutive pentads following the first pentad is higher (lower) than the annual mean; and (3) the averaged MC rainfall for the first pentad and following five pentads is higher (lower) than the annual mean. According to the definition, the DTW (WTD) generally occurs around the 59th (27th) pentad, and the MC experiences asymmetric transitions between wet and dry seasons.

    The WTD is well captured by the NCEP CFSv2 in 0-day lead, although the magnitude of rainfall decrease is underestimated (Zhang et al., 2016b). On the other hand, the DTW in 0-day lead is later and weaker in the model than in the observation. Overall, the CFSv2 well predicts the major features of MC rainfall and related atmospheric circulation surrounding the region (e.g., western Pacific subtropical high) when the forecast lead time is less than two weeks for DTW and less than three weeks for WTD. The model predicts smaller amplitude of the change in rainfall and related atmospheric circulation for both WTD and DTW as lead time increases. The different skills for WTD and DTW and the underestimation of the change in rainfall and related atmospheric circulation require further investigations. It should be noted that only the sub-seasonal prediction of area-averaged MC rainfall was evaluated in Zhang et al. (2016b), and that the regional features need to be further verified. For example, what is the respective performance of the model for land and oceans over the MC? In addition, further evaluations on sub-seasonal prediction of MC rainfall by different state-of-the-art models are needed.

  • As a dominant mode of sub-seasonal variability in the tropics, the MJO offers prospects for improving sub-seasonal climate prediction. The prediction skill by a numerical model depends strongly on the amplitude and phase of the MJO in the model. The past decades have witnessed marked improvement in dynamic MJO prediction, whose skill is up to 3–4 weeks in current operational forecasting systems (Vitart et al., 2007; Lin et al., 2008; Seo et al., 2009; Vitart and Molteni, 2010; Weaver et al., 2011; Zhang et al., 2013; Kim et al., 2014, 2016; Wang et al., 2014). However, many models show difficulties in predicting the propagation of the MJO across the MC, widely known as the MJO “MC prediction barrier” (Vitart et al., 2007; Seo et al., 2009; Vitart and Molteni, 2010; Kim et al., 2014; Wang et al., 2014). Such prediction barrier is sensitive to models. For example, the NCEP CFSv2 shows lower prediction skill in the forecasts targeting MJO in phases 1 and 5, during which enhanced or suppressed convective signal associated with MJO dominates over the MC, due to its large deficiency in predicting the convective anomalies associated with the MJO over the region (Kim et al., 2014). On the other hand, a better representation of convective signals over the MC is found in the ECMWF Variable Resolution Ensemble Prediction System (VarEPS), and its skills of predicting MJO do not show a clear difference between various phases, indicating nonexistence of the “MC prediction barrier” in the model (Kim et al., 2014).

    Recent studies showed that the predictability of the MJO is not sensitive to its phases, suggesting that the model-dependent “MC prediction barrier” is actually a modeling technique problem related to model configuration (Inness and Slingo, 2006; Zhang et al., 2006; Seo and Wang, 2010; Kim et al., 2016; Zhu et al., 2017). Seo and Wang (2010) found that compared to the model simulations using the simplified Arakawa–Schubert (SAS) cumulus parameterization (Pan and Wu, 1995), those with the relaxed Arakawa–Schubert (RAS; Moorthi and Suarez, 1999) produced a significantly better representation of the MJO, particularly the MJO convection signal that penetrates into the MC and the western Pacific. This superiority of the RAS over the SAS in MJO simulation was confirmed by Wang W. Q. et al. (2015) and Zhu et al. (2017). Previous studies also suggested that higher model resolutions led to better representation of the topography over the MC and the response of MJO to air–sea coupling, resulting in better simulations of MJO propagation across the MC (Miura et al., 2007; Crueger et al., 2013; Tan et al., 2018). However, based on the NCEP CFSv2, Zhu et al. (2017) found that the change in model horizontal resolution in both atmosphere and ocean components only had negligible effects on the representation of the MJO, particularly for its eastward propagation. They argued that MJO propagation simulation was more sensitive to convection parameterizations, and that favorable SST conditions to the east of the MC must be realistically developed for an improved simulation of MJO propagation. A recently developed parameterization of surface heat fluxes (i.e., maximum entropy production model), which automatically closes surface energy budgets at all space–time scales, can lead to a more realistic water cycle in the model, resulting in reduction of rainfall bias and more coherent propagation of the MJO over the MC (Chen et al., 2017).

  • The diurnal cycle over the MC islands and the mesoscale circulations generated by land–sea contrast play important roles in energy and hydrological cycles of the MC and in determining its mean climate (Neale and Slingo, 2003). However, simulation of the diurnal cycle over the MC is notoriously problematic for GCMs, which can affect seasonal mean climate (Neale and Slingo, 2003; Love et al., 2011; Gianotti et al., 2012). Significant improvements in simulating diurnal cycle have a potential for reducing the models’ large-scale systematic errors (Neale and Slingo, 2003).

    For those GCMs with coarse horizontal resolutions, it is difficult to represent properly the land–sea distribution and the terrain over the MC, leading to large bias in both amplitude and phase of the diurnal cycle (Neale and Slingo, 2003; Collier and Bowman, 2004; Arakawa and Kitoh, 2005; Qian, 2008). Simulations of the diurnal cycle of rainfall and its phase distribution in a 20-km-grid Meteorological Research Institute General Circulation Model (MRI-GCM) were evaluated in Hara et al. (2009). Improvements were found in the high-resolution model; for instance, the features of observed rainfall were well reproduced over the islands and straits with model horizontal grids smaller than 200 km. Similar improvements have been found in other GCMs and regional climate models (Love et al., 2011; Li Y. et al., 2017; Im and Eltahir, 2018). However, models still have limited accuracy in reproducing the observed timing of diurnal rainfall peak, with the most common error being the earlier occurrence of daily peak than that in the observation, particularly on the larger islands (Hara et al., 2009; Love et al., 2011; Li Y. et al., 2017; Im and Eltahir, 2018).

    The above results suggest that the earlier rainfall (convection) peak over the MC islands must be related to other processes (i.e., physical processes) in the models rather than being simply a problem of resolution (Yang and Slingo, 2001; Love et al., 2011; Gianotti et al., 2012; Li Y. et al., 2017). Wang et al. (2007) demonstrated that the enhancement of fractional convective entrainment/detrainment rate could prolong the development of deep convection and delay the time of rainfall peak, thus contributing to improvement of simulated rainfall diurnal cycle over the MC to some degree. Hara et al. (2009) reported that simulations of the migration of rain belt and the daily maximum at night in the inland areas were substantially improved by using a non-hydrostatic model with a higher resolution and without convective parameterizations. Love et al. (2011) indicated that model sensitivity to gravity wave forcing should be a factor in future convective parametrization schemes.

5.   Role of the MC in global climate variation and prediction
  • Climatologically, the easterly trade wind transports near-surface water from the eastern tropical Pacific to the western tropical Pacific, which in turn forms the underlying sea surface height/temperature differences and the overlying Walker circulation. In the context of east–west asymmetry in the tropical Pacific, ENSO phenomenon emerges from the large-scale dynamics of tropical air–sea interaction (Neelin et al., 1998). Geographically, the MC forms the western boundary of the Pacific, where the westward-propagating oceanic Rossby waves are reflected into eastward-propagating equatorial Kelvin waves, a crucial process of the ENSO delayed oscillator (Suarez and Schopf, 1988). However, the MC is not a solid western boundary compared to the eastern boundary off central and South America. The Indonesian throughflow transports warm and fresh Pacific water into the IO.

    In the atmosphere, ascending branches of both the meridional cell and the Walker circulation are anchored around the MC, leading to strong atmospheric convection. During the ENSO warm episode, convection is enhanced over the central–EP but suppressed over the MC. Zhang et al. (2018c) used coupled model experiments to reveal that the absence of the MC islands (replaced by 10-m-deep water) contributes to a significant asymmetry of ENSO and a weaker nonlinear interaction between the annual cycle of the warm pool and the low-frequency ENSO signals (i.e., the combination mode; Stuecker et al., 2013), which prolongs the warm event, causing a reduction in ENSO frequency. Meanwhile, their results suggested that the NCAR Community Earth System Model (CESM) overestimates the MC topographic uplifting effect on ENSO simulation. Focusing on the role of atmospheric convection, Li et al. (2018) found that long El Niño events tended to start earlier in the boreal spring compared to the short events. Experiments with a fully coupled climate model forced by convectional heating anomalies over the MC and surrounding regions showed that El Niño events become stronger and longer-lasting after introducing anomalous convectional heating. The convection anomalies induce an extensive anomalous westerly belt over the equatorial Pacific, enhancing the El Niño by eastward-propagating Kelvin waves. Additionally, induced by the suppressed heating over the MC, the equatorially-asymmetric westerly belt reduces the meridional shear of mean easterly wind in the lower latitudes, which maintains an anomalous equatorward Sverdrup transport and in turn prolongs the El Niño event.

  • The MC also exerts a significant influence on larger-scale intraseasonal activity, such as the MJO. The MJO propagates over the MC usually in its mature phase, and tends to be weakened or blocked by the underlying islands, topography, and diurnal cycle. Generally, the eastward propagation of the MJO detours southward when approaching the MC (i.e., enhanced convection over south of Sumatra and Java, and the Timor Sea) during the boreal winter (Kim et al., 2017). The MJO-related teleconnection pattern (i.e., extratropical wave train) is the strongest when the MJO convective center is located over the MC (Adames and Wallace, 2014; Bao and Hartmann, 2014), which exerts a great impact on climate and extreme weather events worldwide.

    A variety of mechanisms have been proposed to describe the barrier effect of the MC on MJO propagation (e.g., Hsu and Lee, 2005; Oh et al., 2013; Hagos et al., 2016; Tseng et al., 2017). The island masses act to disturb air–sea interaction, reduce surface fluxes over the MC region, and hence hinder the eastward propagation of the MJO (Maloney and Sobel, 2004; Sobel et al., 2008). On the other hand, Inness and Slingo (2006) suggested that it was the orography of the islands, rather than the presence of the islands themselves, that resulted in blocking of the eastward propagation of the low-level Kelvin wave signals embedded in the MJO. Wu and Hsu (2009) found that when the MJO propagated across the MC, the large-scale circulation was disrupted by the extra lifting and sinking motions created by the topography. As a result, the convective system exhibited quasi-stationary features near the major topography. By conducting model experiments, Tan et al. (2018) reported that removing the topography over the MC led to more continuity in MJO propagation compared to the simulations with the topography. Moreover, the diurnal cycle that exists in the MC can play a marked role in weakening the MJO (Oh et al., 2013; Tung et al., 2014; Hagos et al., 2016). Oh et al. (2013) indicated that the strong diurnal cycle of convection over the MC consumed most of the available energy, and hence inhibited the development of the MJO. Hagos et al. (2016) suggested that the diurnal cycle of cloudiness might play a significant role in causing the MJO convection to stall over the large islands of the MC, mainly through its modulation on downwelling shortwave radiation.

    Zhang and Ling (2017) pointed out that at least two criteria should be satisfied for explaining the barrier effect of the MC: (1) the specific features of the MC (i.e., land–sea distribution and complex terrain) should be included, and (2) the mechanisms for both the barrier effect and its overcoming by some MJO events should be included. However, none of the existing explanations meets both criteria. The multifaceted effects of the MC on MJO need to be further investigated.

  • As a region located between the Asian monsoon and Australian monsoon, the MC experiences a wet season during the boreal winter and a dry season during the boreal summer, dominated by an annual reversal of surface winds (Chang et al., 2005b). Interactions between MC terrains and large-scale transient systems exert great impacts on the demarcation of the summer and winter monsoons. For example, the mixing of the winter and summer rainfall regimes within the equatorial belt 5°S–5°N is almost entirely the result of wind–terrain interaction [see Fig. 7 in Chang et al. (2005b); omitted].

    The climate variability and change in the MC region also exert impacts on extratropical climate. An east–west oriented dipole pattern of heating exists between the MC and central equatorial Pacific, which signifies a major variation of the Pacific Walker circulation related to the ENSO (Lau and Boyle, 1987). This dipole heating pattern, particularly the enhanced convection over Indonesia/Borneo, can cause remarkable extratropical response via modulation on the East Asian jet stream (Chang and Lau, 1982; Lau et al., 1983; Lau and Boyle, 1987). Moreover, the enhanced convection over the MC can excite Kelvin waves, which generate equatorial easterlies and hence off-equatorial anticyclonic shear vorticity over the Philippine Sea (Wang et al., 2013). This anticyclonic shear vorticity can induce boundary-layer divergence and further reinforce the anomalous western North Pacific anticyclone, a key system that bridges ENSO events in the EP and the East Asian monsoons (Wang et al., 2000, 2013; Wang B. et al., 2015). Jiang et al. (2017) suggested that the enhanced convection over the western MC led to significant anomalies of summer rainfall over Southwest China by inducing anomalous anticyclones over the western North Pacific and to the south of the Tibetan Plateau. On the other hand, Abdillah et al. (2017) pointed out that anomalous western North Pacific anticyclone was not strong enough to suppress the equatorward cold airmass flow in the midlatitudes. Instead, the influence of ENSO on the cold air outbreaks over western East Asia is mainly delivered through upper and poleward Rossby wave trains excited by convective anomalies over the MC (Abdillah et al., 2017).

    While the impact of the MC was considered as an intermediary between ENSO and extratropical monsoon, Xu and Guan (2017) proposed an independent influence of convection anomalies over the key area (10°S–10°N, 95°–145°E) on the East Asian summer monsoon. By triggering a northeastward-propagating anomalous East Asia–Pacific/Pacific–Japan (EAP/PJ)-like wave train, anomalous divergence in this key area of the MC can lead to significant anomalous rainfall from the Yangtze River valley in China to the islands of Japan. They indicated that the EAP/PJ-like wave pattern became even clearer after the ENSO and IOD effects were removed. Zhang et al. (2016c) also found a significant correlation between the interannual variation of rainfall over the MC and East Asia after removal of ENSO signals (see Fig. 6). However, the independent role of MC in extratropical monsoon remains unclear, which needs to be addressed in further research.

  • As discussed in Section 4, current models produce substantial errors in simulating and predicting climate variation over the MC, and this model deficiency may be a driver for other systematic errors. By conducting experiments with a climate version of the HadAM3, Neale and Slingo (2003) found that in the absence of the MC islands, the existing local dry bias was reduced due to eliminated cold bias over the region. This improvement in rainfall simulation over the MC led to a reduction of the systematic wet bias over the western IO and the South Pacific Convergence Zone, and a more realistic strength of the Asian summer monsoon. Jiang et al. (2013) showed improved simulation of the East Asian winter monsoon in a Coupled Model Intercomparison Project (CMIP)-type experiment compared to an Atmospheric Model Intercomparison Project (AMIP)-type experiment and the NCEP CFSv2 hindcast, attributed to the better simulation of precipitation over the western MC. Zhao et al. (2015) found that improved prediction of the SST over the western tropical Pacific near the MC enhanced the prediction skill for the early-season rainfall over southern China. Liu et al. (2015) pointed out that the boreal summer intraseasonal oscillation (BSISO) might serve as a bridge by which the forecast of MC rainfall affects the rainfall prediction over the other regions. Kim et al. (2016) argued that the wet bias over the MC and the Pacific Ocean made the western Pacific area unfavorable for MJO propagation, thus limiting its prediction skill in the ECMWF ensemble prediction system. Recently, Zhang et al. (2018c) illustrated an overestimation of the MC topographic uplifting effect on ENSO simulation. They found that eliminating the MC topographical effect produced an improved simulation of cold SST anomalies over the western tropical Pacific, which led to a more realistic representation of the combination mode dynamics, which may contribute to the phase transition of the ENSO. Overall, better representations of terrains, land–sea distribution, and air–sea interaction in the MC region in models may lead to improved sub-seasonal-to-seasonal predictions of regional rainfall variation and hence large-scale climate variability.

6.   Concluding remarks
  • Since the fundamental work by Ramage (1968), the MC has attracted increasing research interest. Significant advance in MC study has been achieved during the recent decades. The spatiotemporal complexity of the MC climate and its interaction with large-scale climate systems (i.e., ENSO and MJO) have been investigated in depth through examinations of observations, reanalysis products, and climate model outputs. Despite extensive efforts devoted, prediction of MC climate variation and its relationships to large-scale systems remain huge challenges even for the state-of-the-art models. For example, the problems exist with respect to the lower prediction skill for MC wet-season rainfall variation and its relationship with ENSO, and the earlier peak of rainfall diurnal cycle in the model simulations. Neale and Slingo (2003) emphasized the crucial role of the MC in global climate prediction, and argued that deficient rainfall over the MC could be a driver for other systematic model errors. Thus, a better understanding of MC climate variation and prediction is important for enhancing our understanding and prediction of the global climate. Particularly, the following issues should be better addressed in future.

    (1) Finer-resolution, higher-frequency, and higher-quality observational datasets are needed for exploring the complexity of multi-scale variability of MC weather and climate. For example, studies that focused on the diurnal cycle of rainfall are limited by using coarse rainfall observations (e.g., 3-hourly, 0.25° × 0.25° TRMM). A new satellite-based observation of precipitation, namely the Global Precipitation Measurement (GPM) with resolutions up to 0.1° × 0.1° every 30 minutes, will provide a better reference for model validation (Hou et al., 2014). Furthermore, the field campaign years of the MC (YMC) from 2017 to 2019 are expected to advance our understanding and prediction of MC weather and climate and related large-scale impact (http://www.jamstec.go.jp/ymc/).

    (2) Previous investigations have mainly focused on the MC wet season and dry season, while a clear theoretical perspective for the asymmetric seasonal march during the transitional seasons has been largely ignored. For example, what causes the sudden drop in rainfall and the change in atmospheric circulations during spring? What is the exact roles of the ITCZ, the landmass, the low-level wind and terrain interactions, and other processes and systems in the asymmetric seasonal march over the MC?

    (3) The influences of the MC on global climate variation (i.e., the Asian monsoon) have generally been considered as an intermediary of the ENSO. Detailed features of the independent role of the MC need to be further investigated.

    (4) The misrepresented physical processes (e.g., surface heat fluxes, clouds, and hydrological cycle) in weather and climate models seem to be the main culprits of common rainfall bias over the MC. Hence, MC rainfall simulation largely depends on the hydrology-related schemes in the models. The rainfall bias over the MC may be reduced in high-resolution (less than ~5 km) models without the cumulus parameterization, due to realistically simulated life cycle of tropical convection.

    Acknowledgments. We thank the anomalous reviewers for their constructive comments on an earlier version of the manuscript.

Reference (161)

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return