A DROUGHT warning system capable of more accurately predicting rainfall up to six months in advance is being developed for the south-eastern and south-western grain belts of Australia.
If the research proves successful, Western Australian growers could find themselves in a better position to make long-term business decisions based on the expected seasonal rainfall for their region.
The weather prediction models could be refined to the point where growers could use them to help decide to take an extended holiday rather than face an uphill battle against drought, or plant a high-yielding variety grain instead of another in good rain seasons.
Agriculture Minister Kim Chance said improving the reliability of weather forecasting systems was the priority for WA's agricultural industry.
"Traditionally weather forecasting on the west coast of Australia has been notoriously difficult to achieve with any degree of certainty," Mr Chance said.
"The current research that is going towards improving these systems is the single most important thing we are doing to assist farmers at the moment.
"If we were able to say with a degree of certainty that it would be better for a grower to take a well-earned holiday than to plant a crop from one season to the next, or alternatively to take advantage of expected good conditions, it would save the country millions of dollars and dramatically improve farmers' profits, far more than anything else that has been done before."
Agriculture Department climatology research officer David Stephens is leading the research project as part of the Managing Climate Variability Program (MCVP), with support from the Grains Research and Development Corporation (GRDC) and the State Government.
Dr Stephens said the decision to undertake the research was made after none of the existing models had successfully predicted the biggest El Nino in history, in 1997.
"Mal Lamond of Lamond Weather Services and I noticed that a sequence of events in the southern mid-latitudes proposed by Harry van Loon had preceded this El Nino, so we decided to publish a paper on it," Dr Stephens said.
"The GRDC and farmers were interested in an El Nino prediction index (EPI) that came out of this and asked us to submit a three-year research project which they funded."
Dr Stephens said most forecasting systems only delivered three-month seasonal outlooks and did not show much accuracy until early June or July.
By this time growers had already made most of their major cropping decisions.
Forecasters find it difficult to make accurate predictions about rainfall in the critical autumn period because the world's weather patterns do not become firmly established until June or July.
Forecasters tend to rely on atmospheric and ocean indicators from equatorial areas and on the Southern Oscillation Index (SOI), which compares the air pressure difference between Darwin and Tahiti.
Dr Stephens developed the EPI, which averages air pressure anomalies over Alice Springs and Mildura between July and September, and found that when this became negative the chances of a dry year and an El Nino increased.
Dr Stephens also created a MeanSOI, which averages the traditional SOI with pressure oscillations along the equator, the southern mid-latitudes and Hawaii.
He found that air pressure changes in mid-latitudes precede the development of El Nino events and determine the strength of Pacific warming.
Joint work with Lamond Weather Services found that the strength of Pacific warming was a key indicator of where the impacts of major droughts are more likely to occur.
Dr Stephens was able to forecast both eastern Pacific sea surface temperatures and May to October rainfall for much of the country from early in the year, using a new model called an ENSO Sequence System (ESS).
Dr Stephens' team discovered that a large decline in the MeanSOI between October and May is a clear indicator of whether an El Nino will develop and is a strong indicator of major droughts in the grain belt.
In early 2005, other forecasters were making predictions until early June that an El Nino was developing and dry conditions could continue in Australia.
But Dr Stephens noticed that the MeanSOI had started to rise, so he issued a statement saying that an El Nino and drought conditions were unlikely - a prediction that has proved accurate.
He said the improved forecasts had the ability to revolutionise farming across Australia and other parts of the world affected by El Nino and La Nina events.
"This system offers the potential for farmers in regions affected by ENSO to better prepare for dry or wet conditions with more lead time than existing forecasting systems," Dr Stephens said.
"The system picks analogue years, similar years from the past to the present year.
"When these analogue years agree we would then suggest that farmers should pay close attention to the outlooks."
Dr Stephens said it was not possible for the research to be 100pc accurate because weather was too complex and spatially varying to predict perfectly, but he was confident of its accuracy.
"When the ESS was tested at predicting May-October rainfall it was found to have low/moderate skill at the beginning of the year, such as three to four months lead time," Dr Stephens said.
"We are now testing whether local indicators such as sea surface temperatures to the north-west of Australia and local barometric pressure can improve on the global scale ESS."
Dr Stephens said the information on predictions would be made available to growers to include in business planning, in the Agriculture Department's ENSO technical summary and on the climate section of its website.
"In the growing season outlook on the Agriculture Department's website we issue a median rainfall ranking map for the analogue years and a discussion on local factors that affect the global scale analogue selection," he said.
"For example, this year the global scale ESS has indicated below-average rainfall for the growing season since February, however, cold sea surface temperatures to the north of Australia have dampened moisture inflow into the Australian region and we have suggested that conditions should be drier than what ESS has suggested."
Dr Stephens said growers would need to factor in local conditions before they could be confident about making decisions on cropping inputs on the basis of climate risk forecasts.
"We suggest farmers look at the outlooks in the context of soil moisture, time of break in the season and commodity prices," Dr Stephens said.
"We also suggest growers look at the skill of the forecasts, which we map on the website.
"When analogues are in agreement and skill is indicated by the forecasts then we would have more confidence in the outlooks.
"When no skill is indicated, then we suggest farmers pay more attention to existing conditions such as stored soil moisture.
"It is hoped that a proposed second stage of the Indian Ocean Climate Initiative will enable this research to continue into the future."