Initial results from tests of a climate forecasting model show it could be sufficiently accurate to be used as a profitable management tool by farmers in Western Australia’s southern Wheatbelt.
The tests were funded by the Managing Climate Variability (MCV) program, which is a collaborative initiative administered by the Grains Research and Development Corporation (GRDC).
The research revealed the ‘dynamical’ Predictive Ocean Atmosphere Model for Australia (POAMA) model was more than 70 per cent accurate in predicting whether growing season rainfall was above or below the median at two sites in this region.
When applied to the years 1980 to 2007, forecasts made by the system on May 1 were 71 per cent accurate in predicting above or below median seasonal rainfall to the end of October at Nyabing, and 78 per cent accurate at Katanning.
According to researchers from the CSIRO, the Department of Agriculture and Food (DAFWA) and the Bureau of Meteorology, farmers in the southern Wheatbelt could, in wetter years, gain about $50 per hectare by basing their nitrogen fertiliser decisions on the model’s forecasts.
Changes to stocking rates and crop-to-pasture ratios could provide additional profits in wetter years of about $16/ha, and further potential benefits of $20,000 to $200,000 per annum could be achieved by dry-seeding early based on a ‘perfect 10-day rain forecast’.
“Application of the ‘dynamical’ POAMA model in WA’s southern Wheatbelt yielded between $100,000 to $200,000 in extra income per average farm and year, excluding the risk of losses from frost, and has potential to improve farm decision making in these areas,” DAFWA researcher Ian Foster said.
But he said the system was significantly less accurate in WA’s northern grainbelt, and had no value in this area.
When applied to the years 1980 to 2007, the model was only 57 per cent accurate in predicting growing season rainfall at Mullewa, and 68 per cent accurate at Merredin.
Dr Foster also warned a lot of work was still required to improve the model.
“To be of benefit to farmers, a climate forecasting model has be correct seven to eight times out of 10 to reliably generate profits,” he said.
“This is not a new miracle system that will take all the risk out of farming, but in areas where it is reasonably accurate, it will certainly help growers.”
Dr Foster said the real potential value of the forecasting system was in allowing farmers to capitalise on potential gains in wetter seasons.
“Particularly in CSIRO’s work with nitrogen decisions, the stronger gains come from taking advantage of really good years, and years when there is more stored subsoil moisture at the start of the growing season,” he said.
Dr Foster said tests of the POAMA model had also revealed it to be remarkably accurate in predicting seasonal conditions experienced in many different regions of Australia in 2010.
“However, further testing under the project is required to confirm its reliability and accuracy in different parts of the country,” he said.
Dr Foster said the ‘dynamical’ POAMA model differed from the statistical forecasting system currently used by the Bureau of Meteorology in Australia.
“The ‘dynamical’ method, which uses the physics of atmospheric and oceanic interactions, is believed to give a more complete picture of seasonal conditions than traditional statistical methods and is seen as the way of the future for seasonal forecasts,” he said.
“Traditional statistical forecasting methods are running into trouble because of climate change, which is weakening the reliability of guidance from historical records.”
