Seasonal Climate Outlook Message for November 2004 to February 2005

The bottom line

Border-Line El Niño Persists 01/11/04

A number of articles have appeared in the general press lately discussing whether or not an El Niño has developed this year.

At present, sea surface temperatures (SST) in the Pacific are not in a classic El Niño pattern such as occurred from mid 2002 to mid 2003. However, from a risk management view point, the fact that SST in the central Pacific from the international dateline running east are warmer than normal (+0.5 to 1.5oC) is cause for concern.

Research has shown that it is the central part of the Pacific Ocean that has a major impact of our seasonal outlook. When SST in this region are warmer than normal, there is an increased risk that our rainfall and water supply across eastern Australia will be below average. We are describing this pattern as a "border-line El Niño".

As an example, in 1992 and 1993 a borderline El Niño sea surface temperature pattern could be found in the Pacific. Both of these years produced below average rainfall for large parts of Queensland but not to the extent of the 2002/2003 El Niño event.

To find out more on conditions in the Pacific try the Bureau of Meteorology " El Niño wrap up " at or the US Climate Prediction Centre at For the latest sea surface temperature maps have a look at or at

With the recent patchy storm rain leaving many still looking for some relief rain, the next passage of the MJO is somewhat anticipated. It last occurred in early October unfortunately triggering little rain. If its timing remains current it will next be expected in the first 2 weeks of November.

The MJO is simply a band of low air pressure originating off the east coast of central Africa travelling eastward across the Indian Ocean and northern Australia roughly every 30 to 60 days. Research has shown the MJO to be a useful indicator of the timing of potential rainfall events across much of Queensland.

Given the growing interest in the MJO, and with funding from the DPI&F, GRDC and CRDC we have developed a site that will allow anyone interested to track its passage.

The full story

SOI Remains In Near Zero Phase

Based on the shift in value of the SOI from the end of September to the end of October, the SOI is in a "Consistently Near Zero" phase. This is the third month in a row the SOI has remained in a "Consistently Near Zero" phase.

Therefore based on the SOI and available rainfall records there is a 50 to in a few locations 70% chance of getting above median rainfall through to the end of January for most of the eastern quarter of Queensland. Across the rest of the state, the chance of getting above median rainfall through to the end of January is lower between 30 to 50%.

For example Winton has a 40% chance of getting above its long-term November to January median rainfall of 132 mm, Dalby has a 50% chance of getting above its long-term November to January median rainfall of 238 mm, Aramac has a 56% chance of getting above its long-term November to January median rainfall of 157 mm and both Townsville and Mackay have a 60% chance of getting above their long-term November to January median rainfall of 400 and 500 mm respectively.

For there to be an overall improvement in the seasonal outlook for Queensland, it would help if the SOI rose to a "Consistently Positive" pattern for a couple of months at least.

Similar to rainfall probabilities in Queensland, there is a 30 to in a few locations 60% chance of getting above the long-term November to January median rainfall across the rest of Australia. The higher rainfall probabilities (50- 60%) can be found mainly in Western Australia. The latest rainfall probability maps for Queensland, Australia and the world are at or

As with any probability based forecast system it is important to consider the opposite aspect. For example Cunnamulla has around a 65% chance of getting above 80 mm over November to January. This also means that there is a 35% chance of NOT getting the 80 mm over November to January.

Another way of looking at this is in around 6 to 7 years out of 10 (or around two thirds) with the current SOI pattern, Cunnamulla has received at least 80 mm over November to January. Therefore in around 3 to 4 years out of 10 (or one third), Cunnamulla has gotten less than 80 mm over November to January

When interpreting rainfall probabilities for your area it may make it easier to think of them in these terms:

1. Probabilities above 80% highlight a high chance 2. Probabilities above 60% highlight a reasonable chance 3. Probabilities below 40% highlight a relatively low chance 4. Probabilities below 20% have a low chance

In shouldn't be a surprise, the final wheat yield outlook for the 2004 season in Queensland shows a below average forecast yield at the end of September of 1.12 t/ha. This is 19% below the long-term median of 1.39 t/ha. Yields have varied greatly throughout the state. For example in central and southeast Queensland, yields have been below average with some regions falling in the worst 10% of all years. However, for some parts of the southwest Queensland crop yield expectations are closer to normal.

This wheat yield outlook is based on a shire scale. It does not take into account crop area planted and is purely a yield forecast. Nor does not take into account individual property circumstances or the effects and damage from poor crop nutrition, pests, diseases, frosts and distribution of planting rain within a shire. For more information on the APSRU/DPI&F regional wheat crop outlook contact Andries Potgieter on (07) 46881417 or try where a full copy the wheat crop outlook can be found.

For detail on farm level crop management strategies for CQ and SQ access

Many people like to follow the relationship between the SOI and rainfall patterns in more detail. To do that, have a look at what happened in your area over November to January in the following years; 2003, 2001, 1995, 1990, 1985, 1984, 1980, 1979, 1978, 1968, 1967, 1966, 1961, 1960, 1959, 1958, 1954 and 1949 and compare the rainfall recorded with your ' normal ' rainfall for November to January.

For example, during November to January at Emerald in those years, above average rainfall was recorded 6 times, near average rainfall was recorded 10 times and below average rainfall was recorded twice. At Roma, above average rainfall was recorded 4 times, near average rainfall was recorded 7 times and below average rainfall was recorded 7 times.

Information on what rainfall patterns where like for November to January in those years can be found at or in Australian Rainman.

To estimate rainfall patterns in Australia over the last two to three hundred years, sources of information such as the annual florescence in cores of coral in the Great Barrier Reef and Pacific are measured. These cores suggest that the driest 30-year period occurred around 1770 to 1799 with the wettest 30-year period occurring from 1950 to 1979. This period of statistically above average rainfall has had many implications as it coincided with the rapid development and expansion of broad acre agricultural throughout Australia post World War Two.

Since then a serious downward trend in rainfall has occurred across eastern Australia especially in the economically important cropping and horticultural belts.

Many producers will be interested to see if the recent dry years are just part of short-term climate variability or, as suspected, part of longer-term ongoing climate change and a return to a more realistic rainfall pattern. Rainfall trend maps are at

Official rainfall records are available for most locations in Australia from around the mid to late 1800 ' s. Even over this short period there have been significant variations in rainfall (and temperature) over large parts of Australia.

For an example, Gayndah has 134 years of rainfall records and a long-term average annual rainfall of 766 mm. However using a 10 year moving average it becomes apparent that there have been prolonged periods of above and below the long-term average annual rainfall of 766 mm. The highest 10 year moving average rainfall was 953 mm in 1956. Since this peak, the 10 year moving average rainfall at Gayndah has fallen to 672 mm at the end of 2003. This is a shift of around 280 mm (or approximately 11 inches). This downward trend since the mid 1950 ' s is evident across most of eastern Australia and especially in Queensland.

This downward trend in rainfall patterns across most of the state is part of the reason we recommend getting updated accurate information on rainfall levels for your location. This will not only help interpret seasonal outlooks but provide a more realistic picture of what rainfall may be expected. This is important when developing management strategies and especially when considering large capital investments such as purchasing property.

Key Points To Consider When Using Climate Information

When incorporating climate forecasts into management decisions, it could be worthwhile to consider some of the following general rules of thumb developed from feedback from climate users.

*The first point is to be sure of your source of information and what it is actually suggesting. Do not take a quick grab of information from any source including radio, TV or the internet and assume what you heard/saw/read applies to your location.

*Management decisions should never be based entirely on one factor (such as a weather or climate forecast). As always, all factors that could impact of the outcome of a decision (such as soil moisture/type, crop, pasture type/availability, commodity prices, machinery, work force, transport, finance, costs, seasonal outlook etc) should be considered.

*Try to access local information or at least be aware of the long-term median for your location so you can correctly interpret the forecast.

*It has also been shown to be useful to do a cost benefit analysis of any decision with a climate risk factor eg What will I gain if I get the desired outcome from this decision? What will I lose if I don ' t get the desired outcome from this decision? What other cost neutral options do I have if any?

*Do not to take a forecast for a specific period (eg September to November) and expand it out (eg late summer). Update the information regularly.

An interesting site from the 'Climate Variability In Agriculture' (CVAP) research and development program is well worth looking at. It highlights some case studies on how producers and businesses have used (to varying levels of success) climate and weather information in their decision making processes.

Last updated: 7 November 2004