Predicting Melbourne’s Wet Periods

Charlie Nelson
February 2011

Melbourne’s wettest periods occur in the six months of spring plus summer (referred to hereafter as the springsummer season).  As shown in Chart 1, there have been nine springsummer seasons where rainfall has exceeded 570mm – and all but two of these seasons have exceeded 600mm.  Note that in the charts, the year shown refers to the springsummer starting in September in year X and ending in February in year X+1.  The last springsummer shown on the chart started on 1 September 2010 and ended on 28 February 2011.

Chart 1

By contrast, the autumn plus winter period (autumnwinter season) has never reached 500mm

My previous research, in 2009, showed that Melbourne’s rainfall is correlated with the 18.6 year lunar node cycle (see Appendix 1 for a description of the cycle).  Melbourne tends to have wet periods of 9.3 years alternating with dry periods of 9.3 years.

In Chart 2, I have overlaid the spring plus summer rainfall data with the lunar node cycle.  In this chart, the positive part of the cycle represents periods when the moon is between the major and minor standstill while the negative part represents periods when the moon is between the minor and major standstill.  My earlier analysis showed that wet periods were mostly between major and minor standstills and dry periods were mostly between minor and major standstills.

Chart 2


All of the nine wettest springsummer seasons have occurred in the positive part of the cycle, between major and minor standstills!  Not one springsummer season occurring in the negative part of the cycle, between minor and major standstills, has exceeded 514mm.

Most of the wet springsummer seasons have occurred in the middle years between the major and minor standstill, the exception being 1903 which occurred in a late year.

It is also noteworthy that there is only one positive part of the lunar node cycle which has not had a wet springsummer season – the period 1876 to 1885.  All others have had one only except for the period 1988 to 1997 which had two wet seasons.

Thus, wet springsummer seasons do not occur randomly but mostly occur within narrow windows near to the middle of the periods between major and minor standstills of the lunar node cycle.  The probability that this outcome is a fluke result is very low – about one chance in 500.

Correlation is one thing and causation another.  I will now be exploring the possible causal chain between the lunar node cycle and wet and dry periods in Melbourne in particular and south east Australia more generally.  For the purposes of prediction, we also need to establish the tripping mechanism which determines the particular year which is very wet.  La Nina events and Indian Ocean Dipole events are not consistently associated with the wettest years, unlike the lunar node cycle.

The Bureau of Meteorology (BOM) and the CSIRO are not considering the role of the moon’s cycles as a driver of rainfall and as a result their predictions are not as accurate as they ought to be.

In Special Climate Statement 22, issued on 1 October 2010, BOM discussed Australia’s wettest September on record in 2010.  With regard to Melbourne, the statement said:

Although 2010 has been the wettest year in Melbourne since 2005, January-September rainfall has still been 35mm below normal, and it is likely that 2010 will be a fourteenth consecutive year of below average rainfall for the city.

Melbourne went on to record its wettest year since 1996 with a total of 777mm, 20% above the long-term average of 648!  At the time the statement was written, it was clear that the Indian Ocean Dipole favoured high rainfall in Melbourne and surrounds (for the first time in several years) and so did the state of the lunar node cycle.

In May 2010, the South Eastern Australian Climate Initiative (SEACI) released a report from Phase 1 of their research into climate variability and change in south-eastern Australia.  At a cost of $7.5 million, the sixty-odd scientists from CSIRO, BOM, and other organizations found that large-scale factors that influence the climate of south-eastern Australia cannot explain the decline in autumn rainfall.  These factors include the El Nino – Southern Oscillation (ENSO), the Indian Ocean Dipole (IOD) and the Southern Annular Mode (SAM).  The study found that:

  • Changes in the Hadley Cell are leading to increasing surface pressure across the region, represented by the sub-tropical ridge (STR);
  • The increase in the intensity of the STR is well correlated with decreasing rainfall in all seasons except summer;
  • There is evidence that the changes in the Hadley Cell and STR are linked to global warming, rather than simply being a natural fluctuation in the global circulation.

The report states that: “Yet another possibility is that the current [rainfall] decline is a natural cycle in the climate of south-eastern Australia”.  “..., it could be argued that a return to wetter conditions is likely in the near future, and that any association of the current dry period with global warming is simply fortuitous.  This is considered to be unlikely by SEACI researchers”.

But there has been a return to wetter conditions and it was happening before the final draft of the SEACI report was finished!  And it was confirmed within months of the report being released - the Murray Darling Basin had its wettest year on record in 2010!.  My research has found that in comparison with three natural drivers of rainfall in south-eastern Australia, the STR is of minor importance.  And the SEACI report does not even consider the influence of the lunar and solar cycles!  Autumn 2011 will provide an important test of the global warming versus natural cycle theories.

My research indicates that it is likely that there will be no further wet springsummers (with 600mm of rainfall) until the late 2020’s or the early 2030’s.  If there is to be one before then it will be in 2011.  But rainfall is likely to be close to average except for the period 2015 to 2022 which is likely to be drier than average.

Appendix 1: the lunar node cycle

The Sun’s declination changes from +23.5° to -23.5° between the solstices due to the Earth’s rotational axis being tilted at about 23.5° from the axis of orbital motion around the sun.  The Moon also changes in declination by the same average amount over a period of four weeks, the period of the Moon’s orbit around the Earth.  But unlike the Sun, the maximum and minimum declination of the moon varies because the Moon’s orbit around the Earth is inclined at 5° to the plane of the Earth’s orbit around the Sun.  The orientation of the 5° inclination relative to the tilt of the earth’s axis rotates around the earth , due to the influence of the sun’s gravity, over a period of 18.6 years.  Thus, the maximum declination of the moon varies between 18.5° and 28.5° over an 18.6 year cycle.

The two points at which the Moon’s path crosses the ecliptic are known as the nodes.  These nodes slowly move around the ecliptic, taking 18.6 years to complete one cycle.

At a minor standstill, which last occurred in 1997, the declination of the moon varies from -18.5° to +18.5° over its month.  At a major standstill, which last occurred in 2006, the declination of the Moon varies from -28.5° to +28.5° over its month.

This means that the Moon swings both further south and further north of the equator at a major standstill and its swings are closer to the equator at a minor standstill.

These variations are easily observed and have been known about since ancient times.

I identified the relationship between Melbourne’s rainfall and the lunar node cycle empirically (not having a background in climate research) but then found there is quite a deal of literature on this topic, most notably by Robert Currie, a professor from the New York State University at Stonybrook.

I happened across Currie’s work when reading “Weather’s greatest mysteries solved!” by Dr Randy Cerveny, President’s Professor in Geographical Sciences specializing in weather and climate at Arizona State University.  In a chapter titled “The mystery of the great American dust bowl” he discusses Currie’s work on cycles and notes that Currie believed that the tidal changes caused by the 18.6 year lunar node (or lunar declination) cycle created variations in regional climates around the world.

In one paper Currie identified an 18.6 year lunar cycle existing in 1,015 out of 1,219 weather records (R. G. Currie and D. P. O’Brien, “Deterministic Signals in USA Precipitation,” International Journal of Climatology 10 (1990): 795-818).


Appendix 2: my previous documented research into Melbourne’s rainfall

In April 2009, I made a submission to the House of Representatives Industry, Science and Innovation Committee enquiry into long-term meteorological forecasting in Australia.  In my submission, I documented my finding that rainfall in Melbourne and the Murray Darling Basin is quite strongly influenced by the 18.6 year lunar node cycle.  The existence of cycles in rainfall, with a predictable driver of those cycles, should be very important for long-term planning.

By August 2009, I had also quantified the impact of the Indian Ocean Dipole, a variation of temperature gradients between the east and west of the tropical Indian Ocean, which influences moisture transport across the Australian continent and so has an impact on rainfall in southeastern Australia.  My August 2009 paper is available online at .

But these two factors were still unable to explain the sudden worsening of the drought in 2006.  By September 2009, I had identified a rare behaviour of the sun which is very likely to have caused this further drop in rainfall.  I was able to advise Melbourne Water and the Murray Darling Basin Authority that there were prospects for a return to average rainfall.  In 2010, Melbourne’s rainfall exceeded the long-term average by 137mm.  It was the wettest year since 1996.  Rainfall in January and February 2011 has also been well above average.  My findings were documented in a July 2010 paper “Melbourne’s long drought explained and prospects for the future” available at