• Arblaster, J. M., , E.-P. Lim, , H. H. Hendon, , B. C. Trewin, , M. C. Wheeler, , G. Liu, , and K. Braganza, 2014: Understanding Australia’s hottest September on record [in “Explaining Extreme Events of 2013 from a Climate Perspective”]. Bull. Amer. Meteor. Soc., 95 (9), S37S41.

    • Search Google Scholar
    • Export Citation
  • Ashok, K., , S. Behera, , A. S. Rao, , H. Weng, , and T. Yamagata, 2007: El Niño Modoki and its possible teleconnection. J. Geophys. Res., 112, C11007, doi:10.1029/2006JC003798.

    • Search Google Scholar
    • Export Citation
  • Burgers, G., , M. A. Balmaseda, , F. C. Vossepoel, , G. vanOldenborgh, , and P. van Leeuwen, 2002: Balanced ocean-data assimilation near the equator. J. Phys. Oceanogr., 32, 25092519, doi:10.1175/1520-0485-32.9.2509.

    • Search Google Scholar
    • Export Citation
  • Cai, W., , A. Sullivan, , and T. Cowan, 2011a: Interactions of ENSO, the IOD, and the SAM in CMIP3 models. J. Climate, 24, 16881704, doi:10.1175/2010JCLI3744.1.

    • Search Google Scholar
    • Export Citation
  • Cai, W., , P. van Rensch, , T. Cowan, , and H. H. Hendon, 2011b: Teleconnection pathways for ENSO and the IOD and the mechanism for impacts on Australian rainfall. J. Climate, 24, 39103923, doi:10.1175/2011JCLI4129.1.

    • Search Google Scholar
    • Export Citation
  • Capotondi, A., and Coauthors, 2015: Understanding ENSO diversity. Bull. Amer. Meteor. Soc.,doi:10.1175/BAMS-D-13-00117.1, in press.

  • Colman, R., , L. Deschamps, , M. Naughton, , L. Rikus, , A. Sulaiman, , K. Puri, , G. Roff, , Z. Sun, , and G. Embery, 2005: BMRC Atmospheric Model (BAM) version 3.0: Comparison with mean climatology. Bureau of Meteorology Research Rep. 108, 32 pp.

    • Search Google Scholar
    • Export Citation
  • Dee, D. P., and Coauthors, 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553597, doi:10.1002/qj.828.

    • Search Google Scholar
    • Export Citation
  • Gates, W. L., 1992: AMIP: The Atmospheric Model Intercomparison Project. Bull. Amer. Meteor. Soc., 73, 19621970, doi:10.1175/1520-0477(1992)073<1962:ATAMIP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Gong, D., , and S. Wang, 1999: Definition of Antarctic Oscillation index. Geophys. Res. Lett., 26, 459462, doi:10.1029/1999GL900003.

  • Hartmann, D. L., , and F. Lo, 1998: Wave-driven zonal flow vacillation in the Southern Hemisphere. J. Atmos. Sci., 55, 13031315, doi:10.1175/1520-0469(1998)055<1303:WDZFVI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., 2003: Indonesian rainfall variability: Impacts of ENSO and local air–sea interaction. J. Climate, 16, 17751790, doi:10.1175/1520-0442(2003)016<1775:IRVIOE>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., , D. W. J. Thompson, , and M. C. Wheeler, 2007a: Australian rainfall and surface temperature variations associated with the Southern Hemisphere annular mode. J. Climate, 20, 24522467, doi:10.1175/JCLI4134.1.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., , M. C. Wheeler, , and C. Zhang, 2007b: Seasonal dependence of the MJO–ENSO relationship. J. Climate, 20, 531543, doi:10.1175/JCLI4003.1.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., , E.-P. Lim, , G. Wang, , O. Alves, , and D. Hudson, 2009: Prospects for predicting two flavors of El Niño. Geophys. Res. Lett., 36, L19713, doi:10.1029/2009GL040100.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., , E.-P. Lim, , and G. Liu, 2012: The role of air–sea interaction for prediction of Australian summer monsoon rainfall. J. Climate, 25, 12781290, doi:10.1175/JCLI-D-11-00125.1.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., , E.-P. Lim, , J. Arblaster, , and D. L. T. Anderson, 2014a: Causes and predictability of the record wet spring over Australia in 2010. Climate Dyn., 42, 11551174, doi:10.1007/s00382-013-1700-5.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., , E.-P. Lim, , and H. Ngyuen, 2014b: Variations of subtropical precipitation and circulation associated with the southern annular mode. J. Climate, 27, 34463460, doi:10.1175/JCLI-D-13-00550.1.

    • Search Google Scholar
    • Export Citation
  • Hudson, D., , O. Alves, , H. H. Hendon, , and G. Wang, 2011: The impact of atmospheric initialisation on seasonal prediction of tropical Pacific SST. Climate Dyn., 36, 11551171, doi:10.1007/s00382-010-0763-9.

    • Search Google Scholar
    • Export Citation
  • Hudson, D., , A. G. Marshall, , Y. Yin, , O. Alves, , and H. H. Hendon, 2013: Improving intraseasonal prediction with a new ensemble generation strategy. Mon. Wea. Rev., 141, 44294449, doi:10.1175/MWR-D-13-00059.1.

    • Search Google Scholar
    • Export Citation
  • Jones, D. A., , W. Wang, , and R. Fawcett, 2009: High-quality spatial climate data-sets for Australia. Aust. Meteor. Ocean J., 58, 233248.

    • Search Google Scholar
    • Export Citation
  • Kang, S., , L. M. Polvani, , J. C. Fyfe, , and M. Sigmond, 2011: Impact of polar ozone depletion on subtropical precipitation. Science, 332, 951954, doi:10.1126/science.1202131.

    • Search Google Scholar
    • Export Citation
  • Kao, H.-Y., , and J.-Y. Yu, 2009: Contrasting eastern-Pacific and central-Pacific types of ENSO. J. Climate, 22, 615632, doi:10.1175/2008JCLI2309.1.

    • Search Google Scholar
    • Export Citation
  • Karoly, D. J., 1989: Southern Hemisphere circulation features associated with El Niño–Southern Oscillation events. J. Climate, 2, 12391252, doi:10.1175/1520-0442(1989)002<1239:SHCFAW>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kidson, J. W., 1988: Indices of the Southern Hemisphere zonal wind. J. Climate, 1, 183194, doi:10.1175/1520-0442(1988)001<0183:IOTSHZ>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kim, H.-M., , P. J. Webster, , and J. A. Curry, 2009: Impact of shifting patterns of Pacific Ocean warming on North Atlantic tropical cyclones. Science, 325, 7780, doi:10.1126/science.1174062.

    • Search Google Scholar
    • Export Citation
  • Kosaka, Y. & , and S.-P. Xie 2013: Recent global-warming hiatus tied to equatorial Pacific surface cooling. Nature, 501, 403407, doi:10.1038/nature1253.

    • Search Google Scholar
    • Export Citation
  • Koster, R. D., and Coauthors, 2004: Regions of strong coupling between soil moisture and precipitation. Science, 305, 11381140, doi:10.1126/science.1100217.

    • Search Google Scholar
    • Export Citation
  • Kug, J.-S., , F.-F. Jin, , and S.-I. An, 2009: Two types of El Niño events: Cold tongue El Niño and warm pool El Niño. J. Climate, 22, 14991515, doi:10.1175/2008JCLI2624.1.

    • Search Google Scholar
    • Export Citation
  • Kumar, K. K., , B. Rajagopalan, , M. Hoerling, , G. Bates, , and M. Cane, 2006: Unraveling the mystery of Indian monsoon failure during El Niño. Science, 314, 115119, doi:10.1126/science.1131152.

    • Search Google Scholar
    • Export Citation
  • Larkin, N. K., , and D. E. Harrison, 2005: Global seasonal temperature and precipitation anomalies during El Niño autumn and winter. Geophys. Res. Lett., 32, L16705, doi:10.1029/2005GL022860.

    • Search Google Scholar
    • Export Citation
  • Larson, S., , S.-K. Lee, , C. Wang, , E.-S. Chung, , and D. Enfield, 2012: Impacts of non-canonical El Niño patterns on Atlantic hurricane activity. Geophys. Res. Lett., 39, L14706, doi:10.1029/2012GL052595.

    • Search Google Scholar
    • Export Citation
  • Lee, S.-K., , R. Atlas, , D. B. Enfield, , C. Wang, , and H. Liu, 2013: Is there an optimal ENSO pattern that enhances large-scale atmospheric processes conducive to major tornado outbreaks in the U.S.? J. Climate, 26, 16261642, doi:10.1175/JCLI-D-12-00128.1.

    • Search Google Scholar
    • Export Citation
  • L’Heureux, M. L., , and D. W. J. Thompson, 2006: Observed relationships between the El Niño–Southern Oscillation and the extratropical zonal-mean circulation. J. Climate, 19, 276287, doi:10.1175/JCLI3617.1.

    • Search Google Scholar
    • Export Citation
  • Lim, E.-P., , and H. H. Hendon, 2015: Understanding and predicting the strong southern annular mode and its impact on the record wet east Australian spring 2010. Climate Dyn., doi:10.1007/s00382-014-2400-5, in press.

    • Search Google Scholar
    • Export Citation
  • Lim, E.-P., , H. H. Hendon, , D. Hudson, , G. Wang, , and O. Alves, 2009: Dynamical forecast of inter–El Niño variations of tropical SST and Australian spring rainfall. Mon. Wea. Rev., 137, 37963810, doi:10.1175/2009MWR2904.1.

    • Search Google Scholar
    • Export Citation
  • Lim, E.-P., , H. H. Hendon, , and H. Rashid, 2013: Seasonal predictability of the southern annular mode due to its association with ENSO. J. Climate, 26, 80378054, doi:10.1175/JCLI-D-13-00006.1.

    • Search Google Scholar
    • Export Citation
  • Manabe, S., , and J. Holloway, 1975: The seasonal variation of the hydrological cycle as simulated by a global model of the atmosphere. J. Geophys. Res., 80, 16171649, doi:10.1029/JC080i012p01617.

    • Search Google Scholar
    • Export Citation
  • Marshall, A. G., , D. Hudson, , M. C. Wheeler, , H. H. Hendon, , and O. Alves, 2012: Simulation and prediction of the southern annular mode and its influence on Australian intra-seasonal climate in POAMA. Climate Dyn., 38, 24832502, doi:10.1007/s00382-011-1140-z.

    • Search Google Scholar
    • Export Citation
  • Marshall, G. J., 2003: Trends in the southern annular mode from observations and reanalyses. J. Climate, 16, 41344143, doi:10.1175/1520-0442(2003)016<4134:TITSAM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • McBride, J. L., , and N. Nicholls, 1983: Seasonal relationships between Australian rainfall and the Southern Oscillation. Mon. Wea. Rev., 111, 19982004, doi:10.1175/1520-0493(1983)111<1998:SRBARA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • McPhaden, M. J., , T. Lee, , and D. McClurg, 2011: El Niño and its relationship to changing background conditions in the tropical Pacific Ocean. Geophys. Res. Lett., 38, L15709, doi:10.1029/2011GL048275.

    • Search Google Scholar
    • Export Citation
  • Meyers, G., , P. MacIntosh, , L. Pigot, , and M. Pook, 2007: The years of El Niño, La Niña, and interactions with the tropical Indian Ocean. J. Climate, 20, 28722880, doi:10.1175/JCLI4152.1.

    • Search Google Scholar
    • Export Citation
  • NCL, cited 2014: NCAR Command Language, version 6.2.1. UCAR/NCAR/CISL/VETS, doi:10.5065/D6WD3XH5.

  • Neelin, J. D., , D. S. Battisti, , A. C. Hirst, , F.-F. Jin, , Y. Wakata, , T. Yamagata, , and S. E. Zebiak, 1998: ENSO theory. J. Geophys. Res., 103, 14 26114 290, doi:10.1029/97JC03424.

    • Search Google Scholar
    • Export Citation
  • Newman, P., , and E. R. Nash, 2005: The unusual Southern Hemisphere stratosphere winter of 2002. J. Atmos. Sci., 62, 614628, doi:10.1175/JAS-3323.1.

    • Search Google Scholar
    • Export Citation
  • Oke, P. R., , A. Schiller, , D. A. Griffin, , and G. B. Brassington, 2005: Ensemble data assimilation for an eddy-resolving ocean model of the Australian region. Quart. J. Roy. Meteor. Soc., 131, 33013311, doi:10.1256/qj.05.95.

    • Search Google Scholar
    • Export Citation
  • Raupach, M. R., , P. R. Briggs, , V. Haverd, , E. A. King, , M. Paget, , and C. M. Trudinger, 2009: Australian Water Availability Project (AWAP): CSIRO marine and atmospheric research component: Final report for phase 3. CAWCR Tech. Rep. 13, 67 pp.

  • Reynolds, R. W., , N. A. Rayner, , T. M. Smith, , D. C. Stokes, , and W. Wang, 2002: An improved in situ and satellite SST analysis for climate. J. Climate, 15, 16091625, doi:10.1175/1520-0442(2002)015<1609:AIISAS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Risbey, J. S., , M. J. Pook, , P. C. McIntosh, , M. C. Wheeler, , and H. H. Hendon, 2009: On the remote drivers of rainfall variability in Australia. Mon. Wea. Rev., 137, 32333253, doi:10.1175/2009MWR2861.1.

    • Search Google Scholar
    • Export Citation
  • Roff, G., , D. W. J. Thompson, , and H. Hendon, 2011: Does increasing model stratospheric resolution improve extended-range forecast skill? Geophys. Res. Lett., 38, L05809, doi:10.1029/2010GL046515.

    • Search Google Scholar
    • Export Citation
  • Saji, N. H., , B. N. Goswami, , P. N. Vinayachandran, , and T. Yamagata, 1999: A dipole mode in the tropical Indian Ocean. Nature, 401, 360363.

    • Search Google Scholar
    • Export Citation
  • Saji, N. H., , T. Ambrizzi, , and S. E. T. Ferraz, 2005: Indian Ocean dipole mode events and austral surface air temperature anomalies. Dyn. Atmos. Oceans, 39, 87100, doi:10.1016/j.dynatmoce.2004.10.015.

    • Search Google Scholar
    • Export Citation
  • Schiller, A., , J. S. Godfrey, , P. C. McIntosh, , G. Meyers, , N. R. Smith, , O. Alves, , G. Wang, , and R. Fiedler, 2002: A new version of the Australian Community Ocean Model for seasonal climate prediction. CSIRO Marine Research Rep. 240, 82 pp.

  • Seager, R., , N. Harnik, , and Y. Kushnir, 2003: Mechanisms of hemispherically symmetric climate variability. J. Climate, 16, 29602978, doi:10.1175/1520-0442(2003)016<2960:MOHSCV>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Seviour, W. J. M., , S. C. Hardiman, , L. J. Gray, , N. Butchart, , C. MacLachlan, , and A. A. Scaife, 2014: Skillful seasonal prediction of the southern annular mode and Antarctic ozone. J. Climate, 27, 74627474, doi:10.1175/JCLI-D-14-00264.1.

    • Search Google Scholar
    • Export Citation
  • Silvestri, G. E., , and C. S. Vera, 2003: Antarctic Oscillation signal on precipitation anomalies over southeastern South America. Geophys. Res. Lett., 30, 2155, doi:10.1029/2003GL018277.

    • Search Google Scholar
    • Export Citation
  • Smith, N. R., , J. E. Blomley, , and G. Meyers, 1991: A univariate statistical interpolation scheme for subsurface thermal analyses in the tropical oceans. Prog. Oceanogr., 28, 219256, doi:10.1016/0079-6611(91)90009-B.

    • Search Google Scholar
    • Export Citation
  • Stephens, C., , J. I. Antonov, , T. P. Boyer, , M. E. Conkright, , R. A. Locarnini, , T. D. O’Brien, , and H. E. Garcia, 2002: Temperature. Vol. 1, World Ocean Atlas 2001, NOAA Atlas NESDIS 49, 167 pp.

  • Stockdale, T. N., , D. L. T. Anderson, , J. O. S. Alves, , and M. A. Balmaseda, 1998: Global seasonal rainfall forecasts using a coupled ocean-atmosphere model. Nature, 392, 370373, doi:10.1038/32861.

    • Search Google Scholar
    • Export Citation
  • Taschetto, A. S., , and M. H. England, 2009: El Niño Modoki impacts on Australian rainfall. J. Climate, 22, 31673174, doi:10.1175/2008JCLI2589.1.

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., , and J. M. Wallace, 2000: Annular modes in the extratropical circulation. Part I: Month-to-month variability. J. Climate, 13, 10001016, doi:10.1175/1520-0442(2000)013<1000:AMITEC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., , M. P. Baldwin, , and S. Solomon, 2005: Stratosphere–troposphere coupling in the Southern Hemisphere. J. Atmos. Sci., 62, 708715, doi:10.1175/JAS-3321.1.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., 1979: Interannual variability of the 500 mb zonal flow in the Southern Hemisphere. Mon. Wea. Rev., 107, 15151524, doi:10.1175/1520-0493(1979)107<1515:IVOTMZ>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Uppala, S. M., and Coauthors, 2005: The ERA-40 Re-Analysis. Quart. J. Roy. Meteor. Soc., 131, 29613012, doi:10.1256/qj.04.176.

  • Valke, S., , L. Terray, , and A. Piacentini, 2000: The OASIS coupled user guide version 2.4. CERFACS Tech. Rep. TR/CMGC/00-10, 85 pp.

  • Wang, G., , and H. H. Hendon, 2007: Sensitivity of Australian rainfall to inter–El Niño variations. J. Climate, 20, 42114226, doi:10.1175/JCLI4228.1.

    • Search Google Scholar
    • Export Citation
  • Wang, W., , and M. J. McPhaden, 2000: The surface layer heat balance in the equatorial Pacific Ocean. Part II: Interannual variability. J. Phys. Oceanogr., 30, 29893008, doi:10.1175/1520-0485(2001)031<2989:TSLHBI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Watkins, A. B., 2003: Seasonal climate summary Southern Hemisphere (spring 2002): The El Niño reaches maturity and dry conditions dominate Australia. Aust. Meteor. Mag., 52, 213226.

    • Search Google Scholar
    • Export Citation
  • Weng, H., , K. Ashok, , S. K. Behera, , S. A. Rao, , and T. Yamagata, 2007: Impacts of recent El Niño Modoki on dry/wet conditions in the Pacific Rim during boreal summer. Climate Dyn., 29, 113129, doi:10.1007/s00382-007-0234-0.

    • Search Google Scholar
    • Export Citation
  • Wheeler, M. C., , H. H. Hendon, , S. Cleland, , H. Meinke, , and A. Donald, 2009: Impacts of the MJO on Australian rainfall and circulation. J. Climate, 22, 14821497, doi:10.1175/2008JCLI2595.1.

    • Search Google Scholar
    • Export Citation
  • Wilks, D., 2006: Statistical Methods in the Atmospheric Sciences. Academic Press, 592 pp.

  • Yeh, S.-W., , J.-S. Kug, , and S.-I. An, 2014: Recent progress on two types of El Niño: Observations, dynamics, and future changes. Asia-Pac. J. Atmos. Sci., 50, 6981, doi:10.1007/s13143-014-0028-3.

    • Search Google Scholar
    • Export Citation
  • Zhang, X., , and M. J. McPhaden, 2008: Eastern equatorial Pacific forcing of ENSO sea surface temperature anomalies. J. Climate, 21, 60706079, doi:10.1175/2008JCLI2422.1.

    • Search Google Scholar
    • Export Citation
  • Zhao, M., , and H. H. Hendon, 2009: Representation and prediction of the Indian Ocean dipole in the POAMA seasonal forecast model. Quart. J. Roy. Meteor. Soc., 135, 337352, doi:10.1002/qj.370.

    • Search Google Scholar
    • Export Citation
  • Zhao, M., , H. H. Hendon, , O. Alves, , Y. Yin, , and D. Anderson, 2013: Impact of salinity constraints on the simulated mean state and variability in a coupled seasonal forecast model. Mon. Wea. Rev., 141, 388402, doi:10.1175/MWR-D-11-00341.1.

    • Search Google Scholar
    • Export Citation
  • Zhao, M., , H. H. Hendon, , O. Alves, , and Y. Yin, 2014: Impact of improved assimilation of temperature and salinity for coupled model seasonal forecasts. Climate Dyn., 42, 25652583, doi:10.1007/s00382-014-2081-0.

    • Search Google Scholar
    • Export Citation
  • Zhou, T., , and R. Yu, 2004: Sea-surface temperature induced variability of the southern annular mode in an atmospheric general circulation model. Geophys. Res. Lett., 31, L24206, doi:10.1029/2004GL021473.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Observed anomalies of (a),(b) SSTs, (c),(d) MSLP, and (e),(f) Australian rainfall in (left) 1997 and (right) 2002 austral spring (SON) based on the climatology of 1982–2005. The color shading interval is 0.5°C for SST anomalies, 2 hPa for MSLP anomalies, and 0.2 mm day−1 for rainfall anomalies.

  • View in gallery

    Observed upper-layer (top 200 mm) soil moisture anomaly by the end of August in (a) 1997 and (b) 2002 expressed as a percent rank relative soil moisture (%), based on the period 1982–2005. Soil moisture data are provided by the Australian Water Availability Project (http://www.csiro.au/awap/; Raupach et al. 2009).

  • View in gallery

    MSLP (contours) and 10-m zonal wind (color shading) anomalies on 1 Sep 1997 and 1 Sep 2002 from (top) ERA-Interim, (middle) ALI, and (bottom) AMIP-type integration relative to the climatologies of 1 Sep over 1982–2005 from the ERA-Interim, ALI, and AMIP-type integration, respectively.

  • View in gallery

    (a) Climatology of the soil moisture (i.e., bucket capacity of 150 mm; Marshall et al. 2012) on 1 Sep and (b),(c) anomalous soil moisture on 1 Sep 1997 and 1 Sep 2002 from the ALI initial conditions. Anomalies are expressed as percent rank relative soil moisture (%) based on ALI initial conditions for the period 1982–2005.

  • View in gallery

    Amplitudes of climate indices in 1997 and 2002 SON. The time series of each index and the associated spatial pattern are presented in the appendix A (Fig. A1).

  • View in gallery

    (a) Reconstructed rainfall anomalies with multiple linear regression on SSTPC1, SSTPC2, DMI and SAMI for SON 1997. (b)–(e) As in (a), but each time leaving out a predictor: (b) SSTPC1, (c) SSTPC2, (d) DMI, and (e) SAMI. (a′)–(e′) As in (a)–(e), but for 2002. The color shading interval is 0.2 mm day−1.

  • View in gallery

    Australian area-mean rainfall anomalies in 1997 (light gray bars) and 2002 (dark gray bars) in the observations and reconstructions shown in Fig. 6.

  • View in gallery

    Predicted amplitudes of the DMI, SSTPC1, SSTPC2, and SAMI in P_ctrl and P_amipAL experiments.

  • View in gallery

    (a) Observed and (b)–(f) predicted Australian rainfall anomalies in SON 1997. (a′)–(f′) As in (a)–(f), but in 2002. Labels of each experiment are shown on the top-right corner of (b′)– (f′). The color shading interval is 0.2 mm day−1.

  • View in gallery

    Statistical significance on the rainfall difference between the two experiments denoted on the top-right corner of (right). The significance was tested by a two-tailed Student’s t test based on 10 ensemble members of each experiment. Significance level (α) less than 0.2 is color shaded.

  • View in gallery

    Forecasts of (left) MSLP and (right) rainfall anomalies of SON 2002 from (a),(b) P_ctrl, (c),(d) P_amipAL, and (e),(f) FAMIP_psst experiments and (g),(g) FAMIP_oisst. The color shading interval is 0.6 hPa for MSLP and 0.2 mm day−1 for rainfall anomalies.

  • View in gallery

    (left) Time series (gray bars; units are standard deviations) of SSTPC1, SSTPC2, DMI, and SAMI overlayed with respective trend lines (red) computed over the period 1982–2005. The total change (units of standard deviations) attributed to the trend over 24 yr is indicated along with its statistical significance on the top right of each graph. Significance is estimated with a two-tailed Student’s t test, assuming 23 degrees of freedom for regression. (right) Regression patterns of SSTs onto SSTPC1, SSTPC2, and DMI and of MSLP onto SAMI.

  • View in gallery

    As in Fig. 6, but the reconstruction of MSLP anomalies of SON for 1997 and 2002. The color shading interval is 0.4 hPa.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 30 30 8
PDF Downloads 21 21 5

Understanding the Contrast of Australian Springtime Rainfall of 1997 and 2002 in the Frame of Two Flavors of El Niño

View More View Less
  • 1 Centre for Australian Weather and Climate Research, Bureau of Meteorology, and CSIRO, Melbourne, Victoria, Australia
© Get Permissions
Full access

Abstract

This study investigates the causes and predictability of the different springtime rainfall responses over Australia for El Niño in 1997 and 2002. The rainfall deficit over Australia is generally assumed to be linearly related to the strength of El Niño. However, Australia received near-normal springtime rainfall during the record strong El Niño in 1997, whereas it suffered from severe drought, especially in the east, during the weak El Niño of 2002.

Statistical reconstruction of the rainfall anomalies and forecasts produced from the Australian Bureau of Meteorology’s dynamical seasonal forecast system [Predictive Ocean and Atmosphere Model for Australia (POAMA)] demonstrated that the eastward and westward shifts of the maximum SST warming of El Niño contributed to the near-normal and dry responses of Australian spring rainfall in 1997 and 2002, respectively. Hence, the contrasting rainfall responses were largely predictable. However, the dry conditions in 2002 were significantly amplified by the occurrence of the record strength negative phase of the southern annular mode (SAM), which could only be predicted with the use of realistic atmospheric initial conditions in the atmosphere–ocean coupled configuration of POAMA. Therefore, predictability of the severity of the 2002 drought over Australia was strongly constrained by the predictability of the SAM, despite the high predictability of the drier than normal condition of 2002 spring that stems from the anomalous central Pacific warming of 2002 El Niño.

Corresponding author address: E.-P. Lim, CAWCR, Bureau of Meteorology, 700 Collins St., Docklands VIC 3008, Australia. E-mail: e.lim@bom.gov.au

Abstract

This study investigates the causes and predictability of the different springtime rainfall responses over Australia for El Niño in 1997 and 2002. The rainfall deficit over Australia is generally assumed to be linearly related to the strength of El Niño. However, Australia received near-normal springtime rainfall during the record strong El Niño in 1997, whereas it suffered from severe drought, especially in the east, during the weak El Niño of 2002.

Statistical reconstruction of the rainfall anomalies and forecasts produced from the Australian Bureau of Meteorology’s dynamical seasonal forecast system [Predictive Ocean and Atmosphere Model for Australia (POAMA)] demonstrated that the eastward and westward shifts of the maximum SST warming of El Niño contributed to the near-normal and dry responses of Australian spring rainfall in 1997 and 2002, respectively. Hence, the contrasting rainfall responses were largely predictable. However, the dry conditions in 2002 were significantly amplified by the occurrence of the record strength negative phase of the southern annular mode (SAM), which could only be predicted with the use of realistic atmospheric initial conditions in the atmosphere–ocean coupled configuration of POAMA. Therefore, predictability of the severity of the 2002 drought over Australia was strongly constrained by the predictability of the SAM, despite the high predictability of the drier than normal condition of 2002 spring that stems from the anomalous central Pacific warming of 2002 El Niño.

Corresponding author address: E.-P. Lim, CAWCR, Bureau of Meteorology, 700 Collins St., Docklands VIC 3008, Australia. E-mail: e.lim@bom.gov.au

1. Introduction

Climatic impacts of El Niño depend not only on the magnitude of the anomalous surface warming in the equatorial Pacific Ocean but also on the spatial pattern or flavor of the warming. For instance, sensitivity to the flavor of El Niño has been shown for rainfall and surface temperature anomalies in Asia and North America (e.g., Larkin and Harrison 2005; Ashok et al. 2007; Weng et al. 2007; Kim et al. 2009; Larson et al. 2012; Lee et al. 2013), the Indian monsoon (Kumar et al. 2006), and rainfall in Australia (e.g., Wang and Hendon 2007; Lim et al. 2009; Taschetto and England 2009).

Much of the inter–El Niño variation of the climate impacts is related to the relative warming of the central Pacific compared to the eastern Pacific (e.g., Larkin and Harrison 2005; Kumar et al. 2006; Ashok et al. 2007; Hendon et al. 2009; Capotondi et al. 2015). El Niño events with the strongest SST anomaly in the equatorial eastern Pacific are referred to as traditional cold tongue events or eastern Pacific events, whose mechanism and basis for long-lead predictability are well understood (referred to here as cold tongue El Niño; e.g., Neelin et al. 1998). El Niño events that have maximum SST anomaly concentrated in the central Pacific are referred to equivalently as date line (Larkin and Harrison 2005), Modoki (means “pseudo”; Ashok et al. 2007), central Pacific (Kao and Yu 2009), or warm pool (Kug et al. 2009) El Niño events (referred to here as warm pool El Niño). The mechanism of warm pool El Niño events is less certain, although some of the key processes that are responsible for the development of this type of El Niños have been proposed (e.g., Wang and McPhaden 2000; Zhang and McPhaden 2008; Kug et al. 2009; Yeh et al. 2014; Capotondi et al. 2015), and their predictability is somewhat unknown largely because of model errors (e.g., Hendon et al. 2009). However, it has been reported that their occurrence appears to have been increasing (e.g., McPhaden et al. 2011), perhaps related to recent decadal changes in the background mean state (e.g., McPhaden et al. 2011; Kosaka and Xie 2013).

Australian springtime climate (September–November) is closely related to the occurrence of El Niño (e.g., McBride and Nicholls 1983) but especially sensitive to the flavors of El Niño (e.g., Wang and Hendon 2007). For instance, below average rainfall across Australia during El Niño is more strongly associated with warm pool El Niños than with cold tongue El Niños (e.g., Wang and Hendon 2007; Lim et al. 2009). In particular, Wang and Hendon (2007) showed that the massive 1997 El Niño, whose maximum SST anomaly occurred in the far eastern Pacific, produced a weak impact on Australian spring rainfall (Fig. 1, left). By contrast, they showed that the relatively weak El Niño of 2002, as judged by the magnitude of the SST anomaly in the eastern equatorial Pacific (as monitored by the Niño-3 index; e.g., Watkins 2003), resulted in a widespread devastating drought likely associated with the occurrence of a relatively strong positive SST anomaly just east of the date line (Fig. 1, right).

Fig. 1.
Fig. 1.

Observed anomalies of (a),(b) SSTs, (c),(d) MSLP, and (e),(f) Australian rainfall in (left) 1997 and (right) 2002 austral spring (SON) based on the climatology of 1982–2005. The color shading interval is 0.5°C for SST anomalies, 2 hPa for MSLP anomalies, and 0.2 mm day−1 for rainfall anomalies.

Citation: Journal of Climate 28, 7; 10.1175/JCLI-D-14-00582.1

The two different flavors of El Niño and their differing impacts on Australian climate have been reported to be predictable, at least at short lead times before model errors prevent the distinction between warm pool and cold tongue El Niños. For instance, Hendon et al. (2009) showed that, at lead times of 0–3 months, the differences in the patterns of SST anomaly during cold tongue and warm pool El Niños could be predicted with the Australian Bureau of Meteorology dynamical seasonal forecast model, Predictive Ocean and Atmosphere Model for Australia (POAMA). Lim et al. (2009) showed that the seven driest Australian springs in the period 1980–2006 were all associated with El Niños whose anomalous maximum SST warming extended far into the western Pacific and that the extreme regional rainfall deficit in each of these years was predictable with POAMA at lead times of up to one season.

Motivated by the fact that Australian spring rainfall anomalies during 1997 and 2002 were so contrasting and perhaps opposite to intuition (i.e., one might have expected severe drought during 1997 and only minor impacts during 2002), the present study explores further the causes and predictability of Australian spring rainfall for these two years. As will be shown, the short-lead-time forecasts from POAMA were very skillful in both years (i.e., POAMA correctly forecast near-normal springtime rainfall conditions in 1997 but more pronounced dry conditions in 2002). However, these apparently good forecasts raise questions about causality: Were the different patterns (or flavors) of tropical Pacific SST anomalies in 1997 and 2002 really the main causes of the rainfall difference over Australia in those two years? If they were, the nonlinear responses of Australian spring rainfall to the magnitudes of El Niño in 1997 and 2002 should have been highly predictable as far as tropical SST anomalies were realistically captured in the POAMA system (i.e., even if the initial state of atmosphere or land was not realistic).

We also ask whether the different SST patterns were the only source of predictability for the rainfall differences in the two years. For instance, Koster et al. (2004) suggested that soil moisture can significantly influence seasonal rainfall in some climate regimes. Based on the Australian Water Availability Project (AWAP) soil moisture analyses (Raupach et al. 2009; http://www.csiro.au/awap/), the upper-layer soil (to the depth of ~200 mm) of the preceding winter during 1997 was drier than normal in the east but wetter in the west of Australia, whereas during 2002 very dry soil conditions covered the entire country (Fig. 2). Did the different initial soil moisture precondition the subsequent rainfall in spring and thus act as an additional source of predictability?

Fig. 2.
Fig. 2.

Observed upper-layer (top 200 mm) soil moisture anomaly by the end of August in (a) 1997 and (b) 2002 expressed as a percent rank relative soil moisture (%), based on the period 1982–2005. Soil moisture data are provided by the Australian Water Availability Project (http://www.csiro.au/awap/; Raupach et al. 2009).

Citation: Journal of Climate 28, 7; 10.1175/JCLI-D-14-00582.1

The 2002 spring also recorded the strongest negative amplitude of the southern annular mode (SAM; e.g., Trenberth 1979; Kidson 1988; Hartmann and Lo 1998; Gong and Wang 1999; Thompson and Wallace 2000) in the past 50 yr (e.g., Hendon et al. 2014a; Fig. A1d), which followed from a massive sudden stratospheric warming over the Antarctic region in winter (e.g., Newman and Nash 2005; Thompson et al. 2005). The negative phase of the SAM (negative SAM) clearly dominated the SH extratropical sea level pressure anomaly in 2002 (Fig. 1d), with zonally uniform higher pressure over the polar region and lower pressure in the midlatitudes than the climatology. Negative SAM is associated with decreased rainfall across eastern and southern Australia (e.g., Hendon et al. 2007a, 2014a). Did this strong negative SAM contribute to the spring drought over eastern Australia? If it did, is the association between the negative SAM and 2002 rainfall anomaly over Australia correctly predicted by the POAMA forecast model? The current study is aimed to address these questions by examining historical relationships and conducting forecast experiments designed to explore the rainfall forecast sensitivity to SSTs, atmosphere, and soil moisture initial conditions.

The POAMA seasonal forecast system and design of the forecast experiments, along with the observational/verification data, are described in section 2. In section 3, we briefly review the observed conditions of large-scale ocean and atmosphere circulations and Australian rainfall in 1997 and 2002 spring [September–November (SON)] and reconstruct the rainfall anomalies with the climate indices representing dominant modes of tropical Indo-Pacific SSTs and the SH extratropical circulation so as to understand the contributions of the large-scale climate drivers to the rainfall anomalies in the two years. Then, in section 4, we analyze the POAMA forecasts for El Niño, the SAM, and Australian rainfall in 1997 and 2002 SON and compare them to the forecasts from the forecast sensitivity experiments to further elucidate the key mechanisms and sources of predictability of the rainfall anomalies in 1997 and 2002. Finally, concluding remarks will be given in section 5.

2. POAMA seasonal forecast model experiments

a. The POAMA coupled forecast system and verification data

The POAMA system consists of the Bureau of Meteorology (BoM) Atmospheric Model version 3 (BAM3; Colman et al. 2005) and the Australian Community Ocean Model version 2 (ACOM2; Schiller et al. 2002; Oke et al. 2005). The horizontal structure of BAM3 is represented by spherical harmonics with a triangular truncation at wavenumber 47 (denoted as T47, which has approximately 250-km resolution), and the vertical variation is represented by 17 sigma levels. In BAM3, the land surface is simulated by a simple bucket model for soil moisture with three levels for temperature (Manabe and Holloway 1975; Hudson et al. 2011). ACOM2 has a zonal resolution of 2° longitude and a telescoping meridional resolution of 0.5° latitude within the equatorial region (8°S–8°N), gradually changing to 1.5° latitude near the poles. ACOM2 has 25 vertical levels, with 12 levels in the top 185 m. The atmosphere and ocean models are coupled every 3 h by the Ocean Atmosphere Sea Ice Soil (OASIS) coupling software (Valke et al. 2000).

The POAMA version used in this study is version 1.5b, which was the BoM’s operational dynamical forecast system for 2007–11 for ENSO forecasts. POAMA forecasts are initialized with observed atmosphere/land and ocean conditions. The atmospheric initial conditions are provided by the atmosphere and land initialization scheme (ALI; Hudson et al. 2011). The ALI initial conditions are generated by nudging horizontal winds, temperature, and humidity in BAM3 to those of the reanalyses from ERA-40 (Uppala et al. 2005) during 1980–2001 and to the global analyses from the BoM’s numerical weather prediction (NWP) system during 2002–06 for the hindcasts. The initial conditions produced from ALI are similar to the observational analyses from ERA-40 and the BoM’s NWP system but result in less initial forecast shock than if the observational analyses were directly used as initial conditions. Land surface conditions are initialized indirectly in response to the surface fluxes generated by the nudged atmosphere.

As an example, Fig. 3 displays the initial 10-m zonal wind and mean sea level pressure (MSLP) conditions on 1 September 1997 and 1 September 2002 (displayed are the anomalies relative to the 1982–2005 daily climatology) from ALI. We also generate initial conditions by running the BAM3 model as an offline Atmospheric Model Intercomparison Project (AMIP)-style integration.1 For comparison, the observed counterparts from the Interim ECMWF Re-Analysis (ERA-Interim; Dee et al. 2011) are also displayed in Fig. 3. ALI initial atmospheric conditions are seen to largely agree with the ECMWF reanalyses (Fig. 3, top and middle).

Fig. 3.
Fig. 3.

MSLP (contours) and 10-m zonal wind (color shading) anomalies on 1 Sep 1997 and 1 Sep 2002 from (top) ERA-Interim, (middle) ALI, and (bottom) AMIP-type integration relative to the climatologies of 1 Sep over 1982–2005 from the ERA-Interim, ALI, and AMIP-type integration, respectively.

Citation: Journal of Climate 28, 7; 10.1175/JCLI-D-14-00582.1

The POAMA ocean data assimilation system (PODAS) provides an estimate of the state of the upper ocean (temperatures and currents) based on the optimum interpolation (OI) of available subsurface temperature observations (Smith et al. 1991), together with a strong relaxation of the SSTs to observed analyses. PODAS does not increment sub surface salinity but surface salinity is constrained by imposed surface flux of freshwater and with an additional relaxation to the climatology from the World Ocean Atlas 2001 (Stephens et al. 2002; Zhao et al. 2013, 2014). Ocean current increments are implemented by applying the geostrophic relation to the temperature increments following Burgers et al. (2002).

For this study, a 10-member ensemble of hindcasts for September to November was generated for the period 1982–2005. Ten ensemble members were produced by perturbing the initial atmospheric conditions by successively picking the atmospheric analysis 6 h earlier from 0000 UTC 1 September, while using the same ocean initial conditions of 1 September. We refer to this control set of hindcasts from POAMA that are initialized with realistic atmosphere, land, and ocean conditions from ALI and PODAS as POAMA control (P_ctrl).

b. Designs of sensitivity experiments

To understand the source of predictability and to elucidate the sensitivity of the seasonal forecasts to atmosphere and land initial conditions, four forecast sensitivity experiments are conducted: two coupled model experiments and two uncoupled model experiments (Table 1).

Table 1.

Configuration of POAMA forecast experiments. Forecasts were initialized on 1 Sep 1997 and 1 Sep 2002 with the conditions generated from the BoM’s forecast initialization schemes shown in the table. Detailed descriptions of ALI, AMIP-style integration, and PEODAS are provided in section 2.

Table 1.

First, to explore the impact of initial land surface conditions on the rainfall forecasts, we kept every initial condition the same as for P_ctrl except initializing the land surface variables (soil moisture and temperature) with the climatological means of 1 September of 1982–2005 that were produced from ALI (Fig. 4a). This experiment is referred to as POAMA with ALI atmosphere and climatological land initial conditions (P_aliAclimL). As discussed earlier, the upper-layer soil moisture was observed to be below normal in the east but above normal in the west by the end of winter 1997, whereas it was strongly below normal across most of the continent by the end of winter 2002 (Fig. 2). These differences are well captured in the ALI initial soil moisture fields (Figs. 4b,c). Therefore, in the P_aliAclimL experiment, 1997 spring will start out with drier than observed soil conditions in the west and wetter than observed soil conditions in the east. Similarly, the P_aliAclimL experiment for 2002 will start out with wetter than observed soil conditions across most of the country.

Fig. 4.
Fig. 4.

(a) Climatology of the soil moisture (i.e., bucket capacity of 150 mm; Marshall et al. 2012) on 1 Sep and (b),(c) anomalous soil moisture on 1 Sep 1997 and 1 Sep 2002 from the ALI initial conditions. Anomalies are expressed as percent rank relative soil moisture (%) based on ALI initial conditions for the period 1982–2005.

Citation: Journal of Climate 28, 7; 10.1175/JCLI-D-14-00582.1

To explore the sensitivity to realistic intraseasonal varying atmosphere/land initial conditions, we conducted an experiment whereby the atmosphere/land is initialized from an AMIP-style integration. In this fashion, the atmosphere/land is initialized with anomalies that are reflective of the slowly varying boundary forcing that was experienced up through the start of spring but contains no realistic information about any intraseasonal variability, especially the SAM and the Madden–Julian oscillation (MJO), which are known to directly affect Australian climate (e.g., Hendon et al. 2007a; Wheeler et al. 2009) but also to affect the evolution of El Niño in the case of MJO (e.g., Hendon et al. 2007b), other than that which is a response to the slowly evolving SST. We refer to these forecasts as POAMA with AMIP atmosphere and land initial conditions (P_amipAL).

Figures 3e,f show that the atmospheric initial conditions produced in an AMIP integration forced with the observed SSTs are different from reality (cf. Figs. 3a,b). However, they do capture the atmospheric fingerprints of the ongoing El Niño events in both years. That is, a typical Southern Oscillation dipole structure in the MSLP field with anomalous westerlies over the equatorial central Pacific is apparent in the AMIP initial conditions for 1997, whereas a weak tripole structure with high–low–high pressure anomalies from the west to the east along the equatorial Pacific with a small blob of westerly anomalies in the west of the date line is apparent for 2002. However, the extratropical variability observed on 1 September each year is absent in the AMIP initial conditions. The initial state of the upper-layer soil moisture, which was brought into the balance with the atmospheric initial conditions from the AMIP integration appears to be wetter than that from P_ctrl almost everywhere over Australia but especially over the southeast and the southwest of the continent for both 1 September 1997 and 1 September 2002 (not shown). For the two coupled experiments (P_aliAclimL and P_amipAL), the oceanic initial conditions were identical to those of P_ctrl.

We also carried out two additional AMIP-style experiments (i.e., SSTs are prescribed) to explore the response of Australian rainfall to observed and forecast SSTs of 1997 and 2002 SON. First, atmosphere-only (uncoupled) forecasts were made using the ALI atmosphere/land surface initial conditions but with the SSTs prescribed during the forecast to be the observed monthly-mean SSTs (interpolated to the model time step) from Reynolds OISSTv2 analysis (Reynolds et al. 2002). We refer to these forecasts as forecast AMIP with Reynolds OISSTs (FAMIP_oisst). Taking a step further, we created a similar ensemble of atmospheric forecasts but with the SSTs prescribed to be the predicted SSTs from the original hindcasts. We refer to these forecasts as forecast AMIP with POAMA SSTs (FAMIP_psst). These experiments are designed to assess the sensitivity of prescribing SSTs rather than allowing full coupling as well as assessing the impact of SST forecast errors on the predicted rainfall anomalies.

From each of P_ctrl, P_amipAL, FAMIP_oisst, and FAMIP_psst, a 10-member ensemble was produced for September, October, and November for the period of 1982–2005, and the ensemble-mean forecasts of the 3-month mean (SON) were computed for each year. Then, for each experiment the SON anomalies of 1997 and 2002 were computed with respect to the hindcast climatology consisting of three randomly chosen members. Thereby, systematic bias associated with model drift was removed from the forecasts (Stockdale et al. 1998). For P_aliAclimL, a 10-member ensemble of hindcasts for SON was generated only for 1997 and 2002, and their anomalies were obtained against the P_ctrl climatology. The design of experiments is summarized in Table 1.

The predictions of Australian rainfall are verified against observed behavior based on the AWAP gridded monthly analyses of rainfall (0.25° latitude × 0.25° longitude but interpolated onto the POAMA grids in this study; Jones et al. 2009). SST forecasts are verified against the Reynolds OISSTv2 analysis and MSLP forecasts are verified against the ERA-Interim MSLP data.

Finally, data analysis and visualization of the results in this study were done by the use of the NCAR Command Language (NCL; NCL 2014).

3. Observed analysis of 1997 and 2002 El Niño events and Australian rainfall

a. Review of 1997 and 2002 SON climate

The 1997 El Niño was characterized as a canonical cold tongue El Niño but with record-breaking magnitude (i.e., SST anomalies in the equatorial eastern Pacific exceeded 4°C; Fig. 1a). The maximum SST warming was confined to the far eastern Pacific, so the central Pacific SST, whose variability is important to Australian spring rainfall variations (Wang and Hendon 2007), was near average. Another interesting feature in 1997 was the occurrence of a big positive excursion of the Indian Ocean dipole mode (IOD; Saji et al.1999) with warm SST anomalies over the tropical western Indian Ocean and cold SST anomalies over the tropical eastern Indian Ocean. This 1997 IOD was the strongest positive event in the last 50-yr record (e.g., Hendon et al. 2014a; Fig. A1c), which was likely promoted by the teleconnection between the strong El Niño and the Indian Ocean, which is also known to be amplified by the local air–sea interaction over the eastern Indian Ocean during austral spring (e.g., Hendon 2003; Hendon et al. 2012). Historically, the IOD has significantly impacted on southern Australian climate in austral winter and spring seasons (e.g., Meyers et al. 2007; Cai et al. 2011b) because Australia is under the direct path of the Rossby wave train excited by the SST anomalies over the eastern and western poles of IOD (e.g., Saji et al. 2005; Cai et al. 2011b). The positive phase of IOD is generally associated with drier than normal conditions over southern Australia, and the negative phase of IOD is associated with the opposite conditions. In spring 1997, the strong El Niño whose maximum SST anomaly was confined to the eastern Pacific and the strong positive IOD were expressed in the atmosphere through the Southern Oscillation pattern whose zero contour was found eastward of the date line in the tropics and the well-developed Rossby wave train from the tropical Indian Ocean toward the west of the Antarctic Peninsula (Fig. 1c). Consequently, high pressure anomalies were observed over the Maritime Continent/northern Australia and over the Great Australian Bight/extratropical southeastern Australia, with a local minimum of the anomalous pressure increase in central Australia.

On the other hand, the 2002 El Niño was a good example of a warm pool El Niño. Maximum SST warming of ~2°C occurred in the central Pacific near the date line with less than 1°C warming in the far eastern Pacific (Fig. 1b). Positive IOD also developed during this El Niño, but the amplitude was small. Because of the maximum warming occurring over the central Pacific, the node of the Southern Oscillation was pushed far west to the eastern coastline of Indonesia along the equator, but the magnitude of the pressure anomaly was small in the tropics (Fig. 1d). A more striking feature found in the MSLP anomaly pattern was the strong negative phase of the SAM. The negative SAM is generally associated with drier conditions across eastern and southern Australia and is thought to occur because of reduced onshore moist flow from the warm Coral Sea as a result of local westerly anomalies in the east (e.g., Hendon et al. 2007a) and/or dynamically induced downward motion over the SH subtropics (e.g., Kang et al. 2011; Hendon et al. 2014b). As we will show below, the strong negative swing of the SAM in 2002 did play a primary role for the severity of the dry conditions during spring 2002, which appears to have been overlooked in previous studies.

Australia had near-normal rainfall in the midst of the record strengths of El Niño and positive IOD in 1997 spring (Fig. 1e), as discussed earlier. The southern half of the country was wetter than normal, whereas the northern end of the country, the southwest corner, and the southeast edge, where ENSO and IOD impacts are historically the largest (e.g., Risbey et al. 2009), were moderately drier than normal. In contrast, in 2002 spring extreme dry conditions occurred over most of the country but were particularly severe in the east, despite the moderate strength of El Niño in the eastern Pacific (Figs. 1b,f).

b. Reconstruction of Australian rainfall with major SST indices and the SAM index

Insight into the causes of these contrasting rainfall anomalies in these two years is provided by statistically reconstructing the rainfall anomalies. The reconstruction is based on multiple linear regression2 using as predictors four climate indices that capture the large-scale oceanic and atmospheric circulation anomalies that are well known to affect Australian rainfall (e.g., Risbey et al. 2009; Hendon et al. 2014a). These include the time series of the first and second principal components of tropical Pacific SSTs (25°S–20°N, 120°–295°E) that represent the variability of canonical eastern Pacific/cold tongue El Niño (SSTPC1) and central Pacific/warm pool El Niño (SSTPC2); the dipole mode index3 (DMI) that monitors the variability of the IOD; and the time series of the first principal component of MSLP over the SH extratropics (20°–75°S) that monitors the variability of the SAM [SAM index (SAMI)]. The time series of the four indices, their associated spatial patterns, and cross correlations between the time series are presented in appendix A (Fig. A1 and Table A1). As briefly discussed earlier, there is strong covariability between SSTPC1 and the DMI in austral spring (correlation of ~0.8 over the period 1982–2005), but we retain a separate predictor for the IOD so as to understand the contributions of independent components of SSTPC1 and the DMI from each other to the rainfall anomalies.

The standardized amplitudes of the four predictor indices during 1997 and 2002 SON are displayed in Fig. 5. SSTPC1 and the DMI show large amplitudes [two to three standard deviations (σ)] in 1997 but much smaller amplitudes in 2002. However, SSTPC2, which depicts the central Pacific warming relative to the eastern and western Pacific basins, is strongly negative (−2σ) in 1997 but moderately positive in 2002. The SAM was extraordinarily strongly negative (−2σ) in 2002 but also significantly negative in 1997 (−1σ). The strongly negative occurrence of the SAM in 2002 has been attributed largely to the extraordinary strength of a rare sudden stratospheric warming (Newman and Nash 2005; Thompson et al. 2005). The large negative amplitude of the SAM in 1997 is likely explained by the promotion of negative SAM by cold tongue El Niño (Table A1; see also Silvestri and Vera 2003; Zhou and Yu 2004; Marshall et al. 2012; Lim et al. 2013; Lim and Hendon 2015). However, the zonally symmetric component of the SH extratropical circulation (i.e., the SAM) appears to be swamped by the zonally asymmetric component resulting from Rossby waves dispersing to high latitudes from the tropical Indian Ocean associated with the strong positive IOD during 1997 spring.

Fig. 5.
Fig. 5.

Amplitudes of climate indices in 1997 and 2002 SON. The time series of each index and the associated spatial pattern are presented in the appendix A (Fig. A1).

Citation: Journal of Climate 28, 7; 10.1175/JCLI-D-14-00582.1

The reconstructed rainfall anomalies by these four indices reasonably well reproduce the main characteristics of observed rainfall anomalies across Australia, especially capturing that 1997 was not as dry as 2002 (Figs. 6a,a′). In particular, Western Australia and central eastern Australia are faithfully reconstructed to be wetter than normal, whereas the northern end, the southeast, and the southwest are reconstructed to be drier than normal in 1997. The year 2002 is reconstructed to be dry everywhere but especially in the east. Although the spatial patterns of rainfall anomalies are successfully depicted by this statistical reconstruction with the four predictors, the overall wetness of 1997 spring and the dryness of 2002 spring over the Australian continent are somewhat underestimated by this method (Fig. 7).

Fig. 6.
Fig. 6.

(a) Reconstructed rainfall anomalies with multiple linear regression on SSTPC1, SSTPC2, DMI and SAMI for SON 1997. (b)–(e) As in (a), but each time leaving out a predictor: (b) SSTPC1, (c) SSTPC2, (d) DMI, and (e) SAMI. (a′)–(e′) As in (a)–(e), but for 2002. The color shading interval is 0.2 mm day−1.

Citation: Journal of Climate 28, 7; 10.1175/JCLI-D-14-00582.1

Fig. 7.
Fig. 7.

Australian area-mean rainfall anomalies in 1997 (light gray bars) and 2002 (dark gray bars) in the observations and reconstructions shown in Fig. 6.

Citation: Journal of Climate 28, 7; 10.1175/JCLI-D-14-00582.1

Then, we repeated the multiple linear regression calculation but leaving one index out each time in order to see the sensitivity of the rainfall anomaly to the independent influence of each index. This exercise reveals that the large negative amplitude of SSTPC2 in 1997 was critical to suppress the potentially severe dry response due to the record-breaking El Niño and IOD (Figs. 6c and 7), which confirms the findings of Wang and Hendon (2007). Likewise, the positive amplitude of SSTPC2 in 2002 is found to be one of the two most important causes of the extreme dry conditions in 2002 spring (Figs. 6c′ and 7). The other important driver of the 2002 dry spring was the occurrence of negative SAM (Figs. 6e′ and 7). In contrast, the negative SAM in 1997 seemed not to have played much of a role on the anomalous rainfall (Figs. 6e and 7). Finally, because SSTPC1 and the DMI strongly covary, the independent influence of the DMI appears not to be significant during both 1997 and 2002 (i.e., its omission does not make a noticeable difference from the original reconstruction as far as SSTPC1 is included in the reconstruction; Figs. 6d,d′ and 7). Last, reconstructions of MSLP anomalies over Australia in 1997 and 2002 SON, using the multiple linear regression models with the same predictor sets, also show that the 1997 and 2002 pressure anomalies over Australia are the most sensitive to the inclusion of SSTPC2 and the SAM in the models, respectively (Fig. B1).

Reproducibility of the basic characteristics of the rainfall anomalies (i.e., near-normal conditions in 1997 spring and nationwide dry conditions in 2002 spring) based on the SST conditions, especially on the behavior of SSTPC2, accounts for reasonably good skill of predicting the contrasting rainfall anomalies of these two seasons with a dynamical forecast system, which will be shown below. However, the important contribution of the SAM in 2002 for producing dry conditions in the east may act to limit predictability for this year. These issues are discussed in the next section.

4. Dynamical forecasts of 1997 and 2002 spring

To assess the forecast skill for Australian rainfall in 1997 and 2002, we first assess how well the dynamical model predicted the relevant large-scale drivers (i.e., SSTPC1, SSTPC2, the DMI, and the SAMI). We computed the predicted SSTPC1 and SSTPC2 by projecting SST forecasts onto the observed SST EOF1 and EOF2 patterns shown in appendix A (Figs A1a,b) and computed the predicted SAMI by projecting MSLP forecasts onto the observed MSLP EOF1 pattern also shown in appendix A (Fig A1d).

The POAMA system demonstrates high skill in predicting the strong positive IOD and the strong eastern Pacific warming El Niño (i.e., large amplitudes of positive SSTPC1 and negative SSTPC2) in 1997 (Fig. 8). Similarly, the central Pacific warming El Niño (i.e., moderate positive amplitudes of SSTPC1 and SSTPC2) in 2002 is also well predicted (Fig. 8). However, the occurrence of the weak positive IOD of 2002 is missed by the POAMA forecasts. The level of skill to predict these tropical Indo-Pacific SST conditions is very similar between P_ctrl and P_amipAL, confirming the notion that atmospheric initial conditions do not play a primary role for predictability of ENSO. However, P_ctrl and P_amipAL demonstrate a substantial difference in the prediction of the magnitude of the negative SAM especially in 2002, which confirms that the occurrence of record strength negative SAM in 2002 was largely internal to the atmosphere.

Fig. 8.
Fig. 8.

Predicted amplitudes of the DMI, SSTPC1, SSTPC2, and SAMI in P_ctrl and P_amipAL experiments.

Citation: Journal of Climate 28, 7; 10.1175/JCLI-D-14-00582.1

For Australian climate, P_ctrl predicts spring 1997 to be wetter than normal and 2002 to be drier than normal, although the forecasts tend to be somewhat wetter than the observations for both years (Figs. 9a,b,a′,b′). Interestingly, the forecasts from all the other four sensitivity experiments show 1997 to be not as dry as 2002 (Figs. 9c–f,c′–f′), despite the absence of realistic atmospheric or land information at the initial state in the coupled experiments and the absence of the atmospheric feedback to the ocean in the uncoupled experiments. Therefore, these results confirm that the source of the rainfall contrast between the two seasons was the difference in the tropical SSTs and that this rainfall contrast was predictable.

Fig. 9.
Fig. 9.

(a) Observed and (b)–(f) predicted Australian rainfall anomalies in SON 1997. (a′)–(f′) As in (a)–(f), but in 2002. Labels of each experiment are shown on the top-right corner of (b′)– (f′). The color shading interval is 0.2 mm day−1.

Citation: Journal of Climate 28, 7; 10.1175/JCLI-D-14-00582.1

In detail, the rainfall forecasts initialized with climatological land surface conditions (P_aliAclimL) are reasonably similar to those with realistic land surface conditions (P_ctrl) in both 1997 and 2002 (Figs. 9c,c′). However, the magnitudes of rainfall anomalies are noticeably reduced in the central part of Australia (Northern Territory and South Australia) in both seasons in P_aliAclimL, which suggests that wetter soil conditions at the end of winter in 1997 and drier soil conditions at the end of winter in 2002 likely contributed to the enhanced and suppressed rainfall responses in the central states in spring 1997 and 2002, respectively. The sensitivity of predicted rainfall to the initial soil moisture over the central states is statistically significant4 at the 90% confidence level (c.l.) in both years (Figs. 10a,a′).

Fig. 10.
Fig. 10.

Statistical significance on the rainfall difference between the two experiments denoted on the top-right corner of (right). The significance was tested by a two-tailed Student’s t test based on 10 ensemble members of each experiment. Significance level (α) less than 0.2 is color shaded.

Citation: Journal of Climate 28, 7; 10.1175/JCLI-D-14-00582.1

Interestingly, the rainfall forecasts are found to be more sensitive to the atmospheric initial conditions than the initial soil moisture (Figs. 9d,d′). The magnitudes of predicted rainfall anomaly are much smaller for the forecasts initialized with AMIP atmospheric conditions, which are balanced to the SSTs but do not contain realistic intraseasonal variability, than for the forecasts initialized with the observed atmospheric initial conditions. In particular, the rainfall anomalies over the eastern states predicted in P_amipAL are significantly different from the original forecasts in P_ctrl at the 95% c.l. for both years (Figs. 10b,b′). A large portion of this difference stems from the first month of the forecasts for both 1997 and 2002 (not shown), which confirms that the atmospheric initial state matters for the evolution of the atmosphere up to a 1-month time scale and then the boundary forcing takes over (e.g., Hudson et al. 2013; Lim and Hendon 2015).

An important cause of the underestimation of 2002 drought in P_amipAL seems associated with the lack of skill in predicting the observed negative SAM (as shown in Fig. 8b). To understand the linkage between the negative SAM and the rainfall deficit in 2002 spring, we display global MSLP and rainfall anomaly patterns of 2002 SON from P_ctrl and P_amipAL in Figs. 11a–d. As demonstrated in Fig. 8, P_ctrl skillfully predicts the negative SAM being dominant in the SH extratropical circulation in 2002 spring, which is similar to the observed pattern (cf. Figs. 11a and 1d). P_ctrl also predicts the pattern of the associated low pressure anomaly over eastern Australia, which brings anomalous warm and dry air from the northwest inland to the southeast of the country (e.g., Arblaster et al. 2014). Furthermore, P_ctrl correctly predicts a zonally symmetric band of suppressed (enhanced) rainfall along 25°–40°S (40°–55°S), which is attributed to the increased downward (upward) motion in the subtropics (midlatitudes) induced by transient eddy momentum transport during the negative SAM (Fig. 10b; Kang et al. 2011; Hendon et al. 2014b). This negative SAM and associated rainfall of 2002 spring is not skillfully predicted in P_amipAL (Figs. 11c,d); therefore, the Australian continent is predicted to be less dry in this experiment than that in P_ctrl (Fig. 11d).

Fig. 11.
Fig. 11.

Forecasts of (left) MSLP and (right) rainfall anomalies of SON 2002 from (a),(b) P_ctrl, (c),(d) P_amipAL, and (e),(f) FAMIP_psst experiments and (g),(g) FAMIP_oisst. The color shading interval is 0.6 hPa for MSLP and 0.2 mm day−1 for rainfall anomalies.

Citation: Journal of Climate 28, 7; 10.1175/JCLI-D-14-00582.1

The 2002 rainfall forecast also shows high sensitivity to the atmosphere and ocean coupling. Although the dry condition is maintained in the east, the intensity is much weaker in the experiment where the atmosphere is forced in uncoupled mode by the POAMA predicted SSTs from P_ctrl (FAMIP_psst; Fig. 9e′). This sensitivity to coupling is statistically significant at the 95% c.l. (Fig. 10c′). The global MSLP and rainfall anomaly patterns from FAMIP_psst show that, without proper air–sea interaction, there are significantly larger rainfall anomalies in the tropical eastern Indian Ocean and in the west coast of central Africa compared to the rainfall anomalies from P_ctrl (Figs. 10c′ and 11f). Hence, a more zonally asymmetric wave pattern is found in the extratropical Indian Ocean (Fig. 11e). Consequently, the negative SAM is not predicted and the associated Australian rainfall response is weak in FAMIP_psst (Figs. 11e,f). The statistically significant rainfall difference due to the absence of the negative SAM in the FAMIP_psst experiment is also apparent in the latitude bands of 45° and 65°S as well as over Australia in Fig. 10c′.

The inability to predict the negative SAM of 2002 spring and the associated rainfall is also found in the experiment where the atmosphere is forced by the observed SSTs (FAMIP_oisst; Figs. 11g,h). In FAMIP_oisst, the local circulation over eastern Australia is not too different from that in P_ctrl (Fig. 11g), but the hemispheric-scale rainfall anomaly related to the SAM is absent; therefore, Australia is predicted to be substantially less dry in FAMIP_oisst than in P_ctrl. It should be reminded here that the atmosphere and land initial conditions used for FAMIP_psst and FAMIP_oisst were identical to those used for P_ctrl, yet Australia was significantly less dry in the two AMIP-type experiments than in P_ctrl. Therefore, these results suggest that the model’s ability to simulate the negative SAM was important to the prediction of the extraordinarily dry conditions over eastern Australia in 2002 spring. It is interesting to learn from FAMIP_psst and FAMIP_oisst that realistic representation of air–sea coupled processes is important to the skillful predictions of the SAM and its associated regional climate impacts. This connection between air–sea coupling and improved prediction of the SAM needs further study.

On the other hand, air–sea coupling does not seem to make a noticeable difference to the prediction of Australian rainfall for 1997 spring (Fig. 9e). Although southern Queensland and New South Wales (25°–30°S in the east) are predicted to be less wet in FAMIP_psst than in P_ctrl, the difference is not statistically significant (Fig. 10c). Finally, the comparison of rainfall forecasts between FAMIP_oisst and FAMIP_psst indicates that the overly wet forecasts over northern Australia in 1997 in FAMIP_psst and P_ctrl could be due to biases in the POAMA predicted SSTs (Figs. 9b,e,f). Furthermore, the rainfall difference in northern Australia between FAMIP_oisst and FAMIP_psst is statistically significant at the 95% c.l. for both 1997 and 2002 spring (Figs. 10d,d′), which implies a systematic bias in the POAMA SST forecasts that negatively impacts rainfall forecasts for northern Australia.

To summarize, the near-normal rainfall anomaly in 1997 spring resulted from the far eastward shift of the maximum SST warming of a strong cold tongue El Niño although substantially excessive rainfall in the southern part of central and eastern Australia was likely due to atmospheric processes that were skillfully captured in the POAMA model by the use of the realistic atmosphere initial conditions of 1 September 1997. In comparison, the drier than normal conditions during 2002 spring resulted from the occurrence of a warm pool El Niño event whose maximum SST warming was placed close to the date line. However, the extremity of the dryness of 2002 spring was significantly contributed to by the occurrence of record strength negative SAM. The forecast experiments that were unable to predict the large negative excursion of the SAM were unable to predict the strength of the drought in 2002. Realistic atmospheric initial conditions and good representation of air–sea interaction are found to be important for the skillful prediction of the strong negative SAM during spring 2002.

5. Summary and concluding remarks

In this study, we have addressed the causality and predictability of the differing responses of Australian springtime rainfall to the El Niño events in 1997 and 2002. The 1997 El Niño event was by many measures the strongest El Niño during the twentieth century but the springtime rainfall anomaly in Australia was weak. The 2002 El Niño was classified as a warm pool event with weak warm anomaly in the eastern Pacific but strong warm anomaly in the central Pacific and was associated with strong dry conditions across most of the Australian continent. Therefore, we posed several questions as follows:

  1. Were the anomalous SST patterns in 1997 and 2002 the main causes of the profoundly different spring rainfall anomalies over Australia and therefore the source for their predictability in both years?
  2. Did the preceding land surface condition, which was much drier in 2002 than in 1997, provide any predictability to the rainfall anomalies of the two seasons?
  3. Did the strong negative SAM of 2002 induced by a sudden stratospheric warming play a role in the Australian spring drought in 2002? If it did, is the association between the negative SAM and 2002 rainfall anomaly over Australia correctly captured by POAMA?

Results from linear reconstructions of the rainfall anomalies based on historical relationships with large-scale oceanic and atmospheric circulations suggest that the eastward shift of the maximum SST warming of 1997 El Niño was indeed the critical cause for near-normal spring rainfall in 1997. The westward shift of the maximum SST warming of 2002 El Niño was important for the enhanced dry conditions over eastern Australia, but the occurrence of strongly negative SAM was even more important for the rainfall deficit, which was not recognized in previous studies.

These findings from statistical analysis are confirmed by carefully designed dynamical forecast sensitivity experiments: Australian springtime rainfall was predicted to be less dry in 1997 compared to 2002 for all experiments that faithfully represented the differing pattern of SSTs associated with the two El Niño events. Therefore, predictability of the contrasting rainfall anomalies in 1997 and 2002 spring largely stems from the different types of El Niño in those two years.

However, the intensity of dry conditions in 2002 showed substantial sensitivity to the forecast of negative SAM. The experiments that were unable to predict the negative SAM were unable to predict the severity of the rainfall deficit, which highlights the important role played by the SAM for the dry conditions in 2002. This prominent role for the SAM also implies a lack of predictability at long lead times because the occurrence of negative SAM during 2002 appears to be largely internal to the atmosphere, presumably with its root in the strong stratospheric warming that occurred at the end of austral winter 2002. The lack of a well resolved stratosphere in the POAMA model leaves large scope for improvements for predictions of the SAM and associated Australian climate (e.g., Roff et al. 2011; Seviour et al. 2014). Representation of air–sea interaction in the model also appears to impact the prediction of the SAM. To get a better understanding of the role of air–sea interaction on the SAM prediction, more thorough assessments of predictions from a coupled versus an uncoupled forecast system will be necessary, and additional forecast sensitivity experiments to see the sensitivity of SAM forecasts to the tropical versus extratropical air–sea coupling are worth carrying out.

The antecedent upper-layer soil moisture conditions were hinted to be important for promoting rainfall responses in the central states of Australia in both 1997 and 2002. Therefore, the contribution of initial land surface conditions to seasonal predictability of Australian climate, especially to the predictability of extreme climate, and its representation in models with improved land surface components will be an interesting and useful area of research for Australian seasonal climate prediction.

Acknowledgments

This work was supported by the Victorian Climate Initiative (VicCI) and the Australian Climate Change Science Program (ACCSP). The authors are grateful to Dr. Guo Liu in the Bureau of Meteorology for the technical assistance with running the experiments. The authors are also thankful to Drs. Guomin Wang and Lynnette Bettio in the Bureau of Meteorology; Dr. Sang-Ki Lee at CIMAS, University of Miami; and the two anonymous reviewers for their constructive comments on this manuscript.

APPENDIX A

Climate Indices Used as Predictors in the Multiple Linear Regression Models

Strong correlations are found between SSTPC1 and DMI, between SSTPC1 and SAMI, and between SAMI and DMI over the period 1982–2005 (Fig. A1 and Table A1). The relationship between SSTPC1 and DMI has been reported in earlier studies with different record lengths and different data products (e.g., Lim et al. 2009; Zhao and Hendon 2009; Cai et al. 2011b; Hendon et al. 2014a). However, the other two relationships seem to be shown only in the last 30 yr (Silvestri and Vera 2003; Lim et al. 2013). For instance, Hendon et al. (2014a), who used the SAM index of the British Antarctic Survey (Marshall 2003) of 1960–2010, does not show any relationship with either SSTPC1 or DMI during the austral spring season. Several studies have proposed mechanisms of the SSTPC1 and SAM teleconnection in the SH warm seasons (e.g., Karoly 1989; Seager et al. 2003; L’Heureux and Thompson 2006; Lim et al. 2013; Lim and Hendon 2015), and the influence of ENSO on SAM is reproducible in the climate models (e.g., Cai et al. 2011a; Lim and Hendon 2015). On the other hand, DMI and SAMI show a high correlation in the past 30 yr as well, but the mechanism of the teleconnection between IOD and SAM has not been well established in the literature.

Fig. A1.
Fig. A1.

(left) Time series (gray bars; units are standard deviations) of SSTPC1, SSTPC2, DMI, and SAMI overlayed with respective trend lines (red) computed over the period 1982–2005. The total change (units of standard deviations) attributed to the trend over 24 yr is indicated along with its statistical significance on the top right of each graph. Significance is estimated with a two-tailed Student’s t test, assuming 23 degrees of freedom for regression. (right) Regression patterns of SSTs onto SSTPC1, SSTPC2, and DMI and of MSLP onto SAMI.

Citation: Journal of Climate 28, 7; 10.1175/JCLI-D-14-00582.1

Table A1.

Correlation between the SST indices and the SAMI in SON for the period 1982–2005. Statistically significant correlation at the 90% confidence level is boldfaced. The significance test was done by a two-tailed Student’s t test given 24 independent samples.

Table A1.

APPENDIX B

Reconstruction of MSLP Anomalies of 1997 and 2002 SON

To gain some insight to the mechanism behind the reconstructed rainfall with the different combinations of four predictors, we also reconstructed MSLP anomalies over the Australian region in the same manner as reconstructing rainfall anomalies shown in Fig. 6. Figure B1 demonstrates that, in 1997 spring, the pressure becomes higher over Australia and a trough between the two centers of high pressure anomalies in the north and the south of Australia becomes weaker if SSTPC2 is not included in the multiple linear regression model. These higher pressure anomalies in Fig B1c are consistent with the respective dry response over Australia in Fig. 6c. On the other hand, the pressure anomalies without the SAMI in the model are distinctively different from the reconstructed pressure anomalies with the SAMI (cf. Fig. B1e′ with Figs B1a′–d′), which is also consistent with the reconstructed 2002 spring to be less dry without the SAMI shown in Fig. 6e′.

Fig. B1.
Fig. B1.

As in Fig. 6, but the reconstruction of MSLP anomalies of SON for 1997 and 2002. The color shading interval is 0.4 hPa.

Citation: Journal of Climate 28, 7; 10.1175/JCLI-D-14-00582.1

REFERENCES

  • Arblaster, J. M., , E.-P. Lim, , H. H. Hendon, , B. C. Trewin, , M. C. Wheeler, , G. Liu, , and K. Braganza, 2014: Understanding Australia’s hottest September on record [in “Explaining Extreme Events of 2013 from a Climate Perspective”]. Bull. Amer. Meteor. Soc., 95 (9), S37S41.

    • Search Google Scholar
    • Export Citation
  • Ashok, K., , S. Behera, , A. S. Rao, , H. Weng, , and T. Yamagata, 2007: El Niño Modoki and its possible teleconnection. J. Geophys. Res., 112, C11007, doi:10.1029/2006JC003798.

    • Search Google Scholar
    • Export Citation
  • Burgers, G., , M. A. Balmaseda, , F. C. Vossepoel, , G. vanOldenborgh, , and P. van Leeuwen, 2002: Balanced ocean-data assimilation near the equator. J. Phys. Oceanogr., 32, 25092519, doi:10.1175/1520-0485-32.9.2509.

    • Search Google Scholar
    • Export Citation
  • Cai, W., , A. Sullivan, , and T. Cowan, 2011a: Interactions of ENSO, the IOD, and the SAM in CMIP3 models. J. Climate, 24, 16881704, doi:10.1175/2010JCLI3744.1.

    • Search Google Scholar
    • Export Citation
  • Cai, W., , P. van Rensch, , T. Cowan, , and H. H. Hendon, 2011b: Teleconnection pathways for ENSO and the IOD and the mechanism for impacts on Australian rainfall. J. Climate, 24, 39103923, doi:10.1175/2011JCLI4129.1.

    • Search Google Scholar
    • Export Citation
  • Capotondi, A., and Coauthors, 2015: Understanding ENSO diversity. Bull. Amer. Meteor. Soc.,doi:10.1175/BAMS-D-13-00117.1, in press.

  • Colman, R., , L. Deschamps, , M. Naughton, , L. Rikus, , A. Sulaiman, , K. Puri, , G. Roff, , Z. Sun, , and G. Embery, 2005: BMRC Atmospheric Model (BAM) version 3.0: Comparison with mean climatology. Bureau of Meteorology Research Rep. 108, 32 pp.

    • Search Google Scholar
    • Export Citation
  • Dee, D. P., and Coauthors, 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553597, doi:10.1002/qj.828.

    • Search Google Scholar
    • Export Citation
  • Gates, W. L., 1992: AMIP: The Atmospheric Model Intercomparison Project. Bull. Amer. Meteor. Soc., 73, 19621970, doi:10.1175/1520-0477(1992)073<1962:ATAMIP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Gong, D., , and S. Wang, 1999: Definition of Antarctic Oscillation index. Geophys. Res. Lett., 26, 459462, doi:10.1029/1999GL900003.

  • Hartmann, D. L., , and F. Lo, 1998: Wave-driven zonal flow vacillation in the Southern Hemisphere. J. Atmos. Sci., 55, 13031315, doi:10.1175/1520-0469(1998)055<1303:WDZFVI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., 2003: Indonesian rainfall variability: Impacts of ENSO and local air–sea interaction. J. Climate, 16, 17751790, doi:10.1175/1520-0442(2003)016<1775:IRVIOE>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., , D. W. J. Thompson, , and M. C. Wheeler, 2007a: Australian rainfall and surface temperature variations associated with the Southern Hemisphere annular mode. J. Climate, 20, 24522467, doi:10.1175/JCLI4134.1.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., , M. C. Wheeler, , and C. Zhang, 2007b: Seasonal dependence of the MJO–ENSO relationship. J. Climate, 20, 531543, doi:10.1175/JCLI4003.1.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., , E.-P. Lim, , G. Wang, , O. Alves, , and D. Hudson, 2009: Prospects for predicting two flavors of El Niño. Geophys. Res. Lett., 36, L19713, doi:10.1029/2009GL040100.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., , E.-P. Lim, , and G. Liu, 2012: The role of air–sea interaction for prediction of Australian summer monsoon rainfall. J. Climate, 25, 12781290, doi:10.1175/JCLI-D-11-00125.1.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., , E.-P. Lim, , J. Arblaster, , and D. L. T. Anderson, 2014a: Causes and predictability of the record wet spring over Australia in 2010. Climate Dyn., 42, 11551174, doi:10.1007/s00382-013-1700-5.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., , E.-P. Lim, , and H. Ngyuen, 2014b: Variations of subtropical precipitation and circulation associated with the southern annular mode. J. Climate, 27, 34463460, doi:10.1175/JCLI-D-13-00550.1.

    • Search Google Scholar
    • Export Citation
  • Hudson, D., , O. Alves, , H. H. Hendon, , and G. Wang, 2011: The impact of atmospheric initialisation on seasonal prediction of tropical Pacific SST. Climate Dyn., 36, 11551171, doi:10.1007/s00382-010-0763-9.

    • Search Google Scholar
    • Export Citation
  • Hudson, D., , A. G. Marshall, , Y. Yin, , O. Alves, , and H. H. Hendon, 2013: Improving intraseasonal prediction with a new ensemble generation strategy. Mon. Wea. Rev., 141, 44294449, doi:10.1175/MWR-D-13-00059.1.

    • Search Google Scholar
    • Export Citation
  • Jones, D. A., , W. Wang, , and R. Fawcett, 2009: High-quality spatial climate data-sets for Australia. Aust. Meteor. Ocean J., 58, 233248.

    • Search Google Scholar
    • Export Citation
  • Kang, S., , L. M. Polvani, , J. C. Fyfe, , and M. Sigmond, 2011: Impact of polar ozone depletion on subtropical precipitation. Science, 332, 951954, doi:10.1126/science.1202131.

    • Search Google Scholar
    • Export Citation
  • Kao, H.-Y., , and J.-Y. Yu, 2009: Contrasting eastern-Pacific and central-Pacific types of ENSO. J. Climate, 22, 615632, doi:10.1175/2008JCLI2309.1.

    • Search Google Scholar
    • Export Citation
  • Karoly, D. J., 1989: Southern Hemisphere circulation features associated with El Niño–Southern Oscillation events. J. Climate, 2, 12391252, doi:10.1175/1520-0442(1989)002<1239:SHCFAW>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kidson, J. W., 1988: Indices of the Southern Hemisphere zonal wind. J. Climate, 1, 183194, doi:10.1175/1520-0442(1988)001<0183:IOTSHZ>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kim, H.-M., , P. J. Webster, , and J. A. Curry, 2009: Impact of shifting patterns of Pacific Ocean warming on North Atlantic tropical cyclones. Science, 325, 7780, doi:10.1126/science.1174062.

    • Search Google Scholar
    • Export Citation
  • Kosaka, Y. & , and S.-P. Xie 2013: Recent global-warming hiatus tied to equatorial Pacific surface cooling. Nature, 501, 403407, doi:10.1038/nature1253.

    • Search Google Scholar
    • Export Citation
  • Koster, R. D., and Coauthors, 2004: Regions of strong coupling between soil moisture and precipitation. Science, 305, 11381140, doi:10.1126/science.1100217.

    • Search Google Scholar
    • Export Citation
  • Kug, J.-S., , F.-F. Jin, , and S.-I. An, 2009: Two types of El Niño events: Cold tongue El Niño and warm pool El Niño. J. Climate, 22, 14991515, doi:10.1175/2008JCLI2624.1.

    • Search Google Scholar
    • Export Citation
  • Kumar, K. K., , B. Rajagopalan, , M. Hoerling, , G. Bates, , and M. Cane, 2006: Unraveling the mystery of Indian monsoon failure during El Niño. Science, 314, 115119, doi:10.1126/science.1131152.

    • Search Google Scholar
    • Export Citation
  • Larkin, N. K., , and D. E. Harrison, 2005: Global seasonal temperature and precipitation anomalies during El Niño autumn and winter. Geophys. Res. Lett., 32, L16705, doi:10.1029/2005GL022860.

    • Search Google Scholar
    • Export Citation
  • Larson, S., , S.-K. Lee, , C. Wang, , E.-S. Chung, , and D. Enfield, 2012: Impacts of non-canonical El Niño patterns on Atlantic hurricane activity. Geophys. Res. Lett., 39, L14706, doi:10.1029/2012GL052595.

    • Search Google Scholar
    • Export Citation
  • Lee, S.-K., , R. Atlas, , D. B. Enfield, , C. Wang, , and H. Liu, 2013: Is there an optimal ENSO pattern that enhances large-scale atmospheric processes conducive to major tornado outbreaks in the U.S.? J. Climate, 26, 16261642, doi:10.1175/JCLI-D-12-00128.1.

    • Search Google Scholar
    • Export Citation
  • L’Heureux, M. L., , and D. W. J. Thompson, 2006: Observed relationships between the El Niño–Southern Oscillation and the extratropical zonal-mean circulation. J. Climate, 19, 276287, doi:10.1175/JCLI3617.1.

    • Search Google Scholar
    • Export Citation
  • Lim, E.-P., , and H. H. Hendon, 2015: Understanding and predicting the strong southern annular mode and its impact on the record wet east Australian spring 2010. Climate Dyn., doi:10.1007/s00382-014-2400-5, in press.

    • Search Google Scholar
    • Export Citation
  • Lim, E.-P., , H. H. Hendon, , D. Hudson, , G. Wang, , and O. Alves, 2009: Dynamical forecast of inter–El Niño variations of tropical SST and Australian spring rainfall. Mon. Wea. Rev., 137, 37963810, doi:10.1175/2009MWR2904.1.

    • Search Google Scholar
    • Export Citation
  • Lim, E.-P., , H. H. Hendon, , and H. Rashid, 2013: Seasonal predictability of the southern annular mode due to its association with ENSO. J. Climate, 26, 80378054, doi:10.1175/JCLI-D-13-00006.1.

    • Search Google Scholar
    • Export Citation
  • Manabe, S., , and J. Holloway, 1975: The seasonal variation of the hydrological cycle as simulated by a global model of the atmosphere. J. Geophys. Res., 80, 16171649, doi:10.1029/JC080i012p01617.

    • Search Google Scholar
    • Export Citation
  • Marshall, A. G., , D. Hudson, , M. C. Wheeler, , H. H. Hendon, , and O. Alves, 2012: Simulation and prediction of the southern annular mode and its influence on Australian intra-seasonal climate in POAMA. Climate Dyn., 38, 24832502, doi:10.1007/s00382-011-1140-z.

    • Search Google Scholar
    • Export Citation
  • Marshall, G. J., 2003: Trends in the southern annular mode from observations and reanalyses. J. Climate, 16, 41344143, doi:10.1175/1520-0442(2003)016<4134:TITSAM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • McBride, J. L., , and N. Nicholls, 1983: Seasonal relationships between Australian rainfall and the Southern Oscillation. Mon. Wea. Rev., 111, 19982004, doi:10.1175/1520-0493(1983)111<1998:SRBARA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • McPhaden, M. J., , T. Lee, , and D. McClurg, 2011: El Niño and its relationship to changing background conditions in the tropical Pacific Ocean. Geophys. Res. Lett., 38, L15709, doi:10.1029/2011GL048275.

    • Search Google Scholar
    • Export Citation
  • Meyers, G., , P. MacIntosh, , L. Pigot, , and M. Pook, 2007: The years of El Niño, La Niña, and interactions with the tropical Indian Ocean. J. Climate, 20, 28722880, doi:10.1175/JCLI4152.1.

    • Search Google Scholar
    • Export Citation
  • NCL, cited 2014: NCAR Command Language, version 6.2.1. UCAR/NCAR/CISL/VETS, doi:10.5065/D6WD3XH5.

  • Neelin, J. D., , D. S. Battisti, , A. C. Hirst, , F.-F. Jin, , Y. Wakata, , T. Yamagata, , and S. E. Zebiak, 1998: ENSO theory. J. Geophys. Res., 103, 14 26114 290, doi:10.1029/97JC03424.

    • Search Google Scholar
    • Export Citation
  • Newman, P., , and E. R. Nash, 2005: The unusual Southern Hemisphere stratosphere winter of 2002. J. Atmos. Sci., 62, 614628, doi:10.1175/JAS-3323.1.

    • Search Google Scholar
    • Export Citation
  • Oke, P. R., , A. Schiller, , D. A. Griffin, , and G. B. Brassington, 2005: Ensemble data assimilation for an eddy-resolving ocean model of the Australian region. Quart. J. Roy. Meteor. Soc., 131, 33013311, doi:10.1256/qj.05.95.

    • Search Google Scholar
    • Export Citation
  • Raupach, M. R., , P. R. Briggs, , V. Haverd, , E. A. King, , M. Paget, , and C. M. Trudinger, 2009: Australian Water Availability Project (AWAP): CSIRO marine and atmospheric research component: Final report for phase 3. CAWCR Tech. Rep. 13, 67 pp.

  • Reynolds, R. W., , N. A. Rayner, , T. M. Smith, , D. C. Stokes, , and W. Wang, 2002: An improved in situ and satellite SST analysis for climate. J. Climate, 15, 16091625, doi:10.1175/1520-0442(2002)015<1609:AIISAS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Risbey, J. S., , M. J. Pook, , P. C. McIntosh, , M. C. Wheeler, , and H. H. Hendon, 2009: On the remote drivers of rainfall variability in Australia. Mon. Wea. Rev., 137, 32333253, doi:10.1175/2009MWR2861.1.

    • Search Google Scholar
    • Export Citation
  • Roff, G., , D. W. J. Thompson, , and H. Hendon, 2011: Does increasing model stratospheric resolution improve extended-range forecast skill? Geophys. Res. Lett., 38, L05809, doi:10.1029/2010GL046515.

    • Search Google Scholar
    • Export Citation
  • Saji, N. H., , B. N. Goswami, , P. N. Vinayachandran, , and T. Yamagata, 1999: A dipole mode in the tropical Indian Ocean. Nature, 401, 360363.

    • Search Google Scholar
    • Export Citation
  • Saji, N. H., , T. Ambrizzi, , and S. E. T. Ferraz, 2005: Indian Ocean dipole mode events and austral surface air temperature anomalies. Dyn. Atmos. Oceans, 39, 87100, doi:10.1016/j.dynatmoce.2004.10.015.

    • Search Google Scholar
    • Export Citation
  • Schiller, A., , J. S. Godfrey, , P. C. McIntosh, , G. Meyers, , N. R. Smith, , O. Alves, , G. Wang, , and R. Fiedler, 2002: A new version of the Australian Community Ocean Model for seasonal climate prediction. CSIRO Marine Research Rep. 240, 82 pp.

  • Seager, R., , N. Harnik, , and Y. Kushnir, 2003: Mechanisms of hemispherically symmetric climate variability. J. Climate, 16, 29602978, doi:10.1175/1520-0442(2003)016<2960:MOHSCV>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Seviour, W. J. M., , S. C. Hardiman, , L. J. Gray, , N. Butchart, , C. MacLachlan, , and A. A. Scaife, 2014: Skillful seasonal prediction of the southern annular mode and Antarctic ozone. J. Climate, 27, 74627474, doi:10.1175/JCLI-D-14-00264.1.

    • Search Google Scholar
    • Export Citation
  • Silvestri, G. E., , and C. S. Vera, 2003: Antarctic Oscillation signal on precipitation anomalies over southeastern South America. Geophys. Res. Lett., 30, 2155, doi:10.1029/2003GL018277.

    • Search Google Scholar
    • Export Citation
  • Smith, N. R., , J. E. Blomley, , and G. Meyers, 1991: A univariate statistical interpolation scheme for subsurface thermal analyses in the tropical oceans. Prog. Oceanogr., 28, 219256, doi:10.1016/0079-6611(91)90009-B.

    • Search Google Scholar
    • Export Citation
  • Stephens, C., , J. I. Antonov, , T. P. Boyer, , M. E. Conkright, , R. A. Locarnini, , T. D. O’Brien, , and H. E. Garcia, 2002: Temperature. Vol. 1, World Ocean Atlas 2001, NOAA Atlas NESDIS 49, 167 pp.

  • Stockdale, T. N., , D. L. T. Anderson, , J. O. S. Alves, , and M. A. Balmaseda, 1998: Global seasonal rainfall forecasts using a coupled ocean-atmosphere model. Nature, 392, 370373, doi:10.1038/32861.

    • Search Google Scholar
    • Export Citation
  • Taschetto, A. S., , and M. H. England, 2009: El Niño Modoki impacts on Australian rainfall. J. Climate, 22, 31673174, doi:10.1175/2008JCLI2589.1.

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., , and J. M. Wallace, 2000: Annular modes in the extratropical circulation. Part I: Month-to-month variability. J. Climate, 13, 10001016, doi:10.1175/1520-0442(2000)013<1000:AMITEC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., , M. P. Baldwin, , and S. Solomon, 2005: Stratosphere–troposphere coupling in the Southern Hemisphere. J. Atmos. Sci., 62, 708715, doi:10.1175/JAS-3321.1.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., 1979: Interannual variability of the 500 mb zonal flow in the Southern Hemisphere. Mon. Wea. Rev., 107, 15151524, doi:10.1175/1520-0493(1979)107<1515:IVOTMZ>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Uppala, S. M., and Coauthors, 2005: The ERA-40 Re-Analysis. Quart. J. Roy. Meteor. Soc., 131, 29613012, doi:10.1256/qj.04.176.

  • Valke, S., , L. Terray, , and A. Piacentini, 2000: The OASIS coupled user guide version 2.4. CERFACS Tech. Rep. TR/CMGC/00-10, 85 pp.

  • Wang, G., , and H. H. Hendon, 2007: Sensitivity of Australian rainfall to inter–El Niño variations. J. Climate, 20, 42114226, doi:10.1175/JCLI4228.1.

    • Search Google Scholar
    • Export Citation
  • Wang, W., , and M. J. McPhaden, 2000: The surface layer heat balance in the equatorial Pacific Ocean. Part II: Interannual variability. J. Phys. Oceanogr., 30, 29893008, doi:10.1175/1520-0485(2001)031<2989:TSLHBI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Watkins, A. B., 2003: Seasonal climate summary Southern Hemisphere (spring 2002): The El Niño reaches maturity and dry conditions dominate Australia. Aust. Meteor. Mag., 52, 213226.

    • Search Google Scholar
    • Export Citation
  • Weng, H., , K. Ashok, , S. K. Behera, , S. A. Rao, , and T. Yamagata, 2007: Impacts of recent El Niño Modoki on dry/wet conditions in the Pacific Rim during boreal summer. Climate Dyn., 29, 113129, doi:10.1007/s00382-007-0234-0.

    • Search Google Scholar
    • Export Citation
  • Wheeler, M. C., , H. H. Hendon, , S. Cleland, , H. Meinke, , and A. Donald, 2009: Impacts of the MJO on Australian rainfall and circulation. J. Climate, 22, 14821497, doi:10.1175/2008JCLI2595.1.

    • Search Google Scholar
    • Export Citation
  • Wilks, D., 2006: Statistical Methods in the Atmospheric Sciences. Academic Press, 592 pp.

  • Yeh, S.-W., , J.-S. Kug, , and S.-I. An, 2014: Recent progress on two types of El Niño: Observations, dynamics, and future changes. Asia-Pac. J. Atmos. Sci., 50, 6981, doi:10.1007/s13143-014-0028-3.

    • Search Google Scholar
    • Export Citation
  • Zhang, X., , and M. J. McPhaden, 2008: Eastern equatorial Pacific forcing of ENSO sea surface temperature anomalies. J. Climate, 21, 60706079, doi:10.1175/2008JCLI2422.1.

    • Search Google Scholar
    • Export Citation
  • Zhao, M., , and H. H. Hendon, 2009: Representation and prediction of the Indian Ocean dipole in the POAMA seasonal forecast model. Quart. J. Roy. Meteor. Soc., 135, 337352, doi:10.1002/qj.370.

    • Search Google Scholar
    • Export Citation
  • Zhao, M., , H. H. Hendon, , O. Alves, , Y. Yin, , and D. Anderson, 2013: Impact of salinity constraints on the simulated mean state and variability in a coupled seasonal forecast model. Mon. Wea. Rev., 141, 388402, doi:10.1175/MWR-D-11-00341.1.

    • Search Google Scholar
    • Export Citation
  • Zhao, M., , H. H. Hendon, , O. Alves, , and Y. Yin, 2014: Impact of improved assimilation of temperature and salinity for coupled model seasonal forecasts. Climate Dyn., 42, 25652583, doi:10.1007/s00382-014-2081-0.

    • Search Google Scholar
    • Export Citation
  • Zhou, T., , and R. Yu, 2004: Sea-surface temperature induced variability of the southern annular mode in an atmospheric general circulation model. Geophys. Res. Lett., 31, L24206, doi:10.1029/2004GL021473.

    • Search Google Scholar
    • Export Citation
1

In the Atmospheric Model Intercomparison Project, atmospheric models were forced with prescribed SSTs (Gates 1992). A set of 10-member ensemble atmospheric initial conditions was generated from the AMIP-style integration forced by observed SSTs (Reynolds et al. 2002), starting from 1 June with atmospheric perturbations.

2

where y is a predictand, x denotes predictors, b denotes multiple linear regression coefficients, and M is the number of predictors [Wilks 2006, Eq. (6.24)].

3

DMI = SST(10°S–10°N, 50°–70°E) − SST(10°S–0°, 90°–110°E) (Saji et al. 1999).

4

Statistical significance on the difference of two ensemble means was tested by a two-tailed Student’s t test based on two sets of 10-ensemble-member forecasts for 1997 and 2002.

Save