Influence of the Madden-Julian Oscillation phase on rainfall and intra-seasonal dry spells: Halls Creek case study

Climatology (historical weather observations) together with the Madden-Julian Oscillation (MJO) index forecast and seasonal outlooks can be useful in northern rangeland management decision making.

Climatology can provide long term ‘best-bet’ dates for wet season onset, green date, production point and end of growing season: these can be used in combination with seasonal outlooks to inform decisions about wet season stock movement, fire management, weed control, joining windows, adjusting stocking rates, planting for restoration or crops and the like.

The MJO is a global-scale feature of the tropical atmosphere associated with weekly to monthly periods of enhanced and suppressed rainfall over northern Australia. It influences the timing, development and strength of the major global monsoon patterns, and tropical cyclones; and is associated with variations in wind, cloudiness, and rainfall (Bureau of Meteorology (BoM) website).

Recent observations and an experimental forecast of the MJO are available from the BoM website. The MJO observations and forecast can be applied to any location and compared with local rainfall events. The MJO phase diagram illustrates the progression of the MJO through different phases, which generally coincide with locations along the equator around the globe. RMM1 and RMM2 are mathematical methods (Real-time Multivariate MJO) that combine cloud amount and winds at upper and lower levels of the atmosphere to provide a measure of the strength and location of the MJO (the index). When the index is within the centre circle of the diagram the MJO is considered weak, meaning it is difficult to discern (NACP climate workshop, Katherine 2019).

The graph indicates that when the calculations of RMM1 and RMM2 indicate the MJO is weak, it is difficult to discern where the influence of the weather system is being felt.
Image 1. Reading the phase diagram aka ‘the pigtail’


Historical MJO data and the BoM Halls Creek composite rainfall record were compared by DPIRD to localise existing general information of relevance to the pastoral industry.  Summary statistics of rainfall data from 1974-1975 to 2017-2018 at Halls Creek show that phases 5 & 6 are most likely to coincide with rainfall. The MJO also influences ‘dry’ periods during the northern wet season in phases 8, 1 and 2 (Wheeler et al. 2009). Phases 1 & 2 are least likely to coincide with rain at Halls Creek.

The graph indicates that there is a 2 mm anomaly around the mean when the MJO is in the sixth phase.
Image 2. Wet season daily rainfall anomaly by MJO phase at Halls Creek, WA


The same record set described above was used to investigate the occurrence of ‘rainfall effective for plant growth’, applying the rule-of-thumb threshold commonly used in the Kimberley (50 mm or more over 3 days with follow-up rainfall two weeks later). Rainfall meeting this threshold occurred at least once in 27 of the years in the records 1974-1975 to 2017-2018 and in most MJO phases, but never in phase 2 (note: gap in MJO data 1977-1978 excluded from both datasets).

The 23 years where ‘effective rainfall’ at this threshold was not recorded coincided with years of near or below median rainfall. Informal observation of seasonal grass growth is at odds with the low incidence of ‘rainfall effective for growth’ in the record (ie grass grows even when the threshold is not met); suggesting that the rule of thumb is due for a revision in this locality.

Image 3. Annual rainfall anomaly at Halls Creek (July-June) 1974-1975 to 2017-2018


 Other major drivers of rainfall at Halls Creek include El Niňo and La Niňa conditions, the Indian Ocean Dipole and regional soil moisture (Sharmila and Hendon 2019).

Further analyses already underway to improve our understanding of the MJO and other climate drivers include:

  • Regional drivers of ‘unseasonal’ and variable rainfall- Sharmila Sur (University of Southern Queensland)
  • New and improved forecast tools for the MJO- Matt Wheeler (BoM)

Additional analysis of the MJO and Halls Creek datasets could be used to investigate:

  • If the length of time that the MJO remains in phases 4, 5 and 6 is correlated with the amount of rainfall received (i.e. increased length of time ‘stalled’ may lead to increased rainfall) at Halls Creek
  • A revision of rule-of-thumb thresholds for ‘rainfall effective for growth’ at Halls Creek

To find out more about the MJO or the Northern Australia Climate Program in your area, contact your local climate mates, Anne Marie Huey (East Kimberley/VRD) on 08 9191 7069 or Jardine MacDonald (West Kimberley/Pilbara) on 08 9192 5507


Ryan, K. 2019: Madden-Julian Oscillation influence of phase on rainfall and intraseasonal dry spells: Halls Creek case study. Australian Rangeland Society 2019 conf. proceedings (abstract).

Sharmila, S. and H.H. Hendon, 2019: Causes and predictability of multi-year variability of Northern Australia wet-season rainfall. AMOS-ICTMO2019 conf. proceedings (abstract).

Wheeler, M.C., H.H. Hendon, S. Cleland, H. Meinke, and A. Donald, 2009: Impacts of the Madden–Julian Oscillation on Australian Rainfall and Circulation. J. Climate, 22,1482-1498,

For more information contact Kath Ryan, Development Officer, Kununurra, on +61 (0)8 9166 4000 or