Blooms of blue-green algae in Australia's largest river, the Murray could be predicted up to four weeks before they occur thanks to a new computer model being developed at Adelaide University with the help of industry.
The model-an Australian first-has already shown early signs of success.
Adelaide University student Mr Gavin Bowden is working on the project as part of his PhD with the Department of Civil & Environmental Engineering. His research is funded by an Australian Research Council SPIRT grant, in partnership with United Utilities Australia (formerly known as North West Water).
United Utilities has a 25-year contract with SA Water to operate 10 water treatment facilities along the lower Murray, providing water to more than 150,000 South Australians.
Outbreaks of toxic blue-green algae pose a major threat to the quality of water.
"Until now, traditional attempts to predict outbreaks of blue-green algae have been unsuccessful," Mr Bowden said.
He has achieved some success using a special computer modelling process which mimics the human brain. The "artificial neural networks" utilised by the computer model enable it to "learn" the key factors which contribute to an algal outbreak, enabling the prediction of major blooms before they arise.
Mr Bowden has been supplying his computer model with data provided by SA Water and the Murray-Darling Basin Commission. The data was compiled from water samples collected every week over the last 20 years at Morgan in the Riverland.
"A range of environmental factors in the water are taken into account by the model," Mr Bowden said.
"These include nutrients, such as nitrogen and phosphorus, the flow and temperature of the water, turbidity and colour.
"There are other factors which influence the development of blue-green algae, but the beauty of using neural networks is that you can get reasonable predictions by including just the major factors."
Early results are extremely promising. When the computer's forecasts are measured against actual events, the predictions come close to the mark. Although not 100% accurate, they clearly show the peaks and troughs in algal bloom development.
"This is far better than other previous modelling techniques, and is very promising. The model picks up the overall pattern of algal activity-not only that, it's picking it up a full four weeks in advance," Mr Bowden said.
Right now, Mr Bowden is developing the model on his computer based at the University. He believes a fine-tuned, fully workable model that can be used by United Utilities is only about six months away.
"The ultimate goal is to have this model sitting either at the head office in the city or at the treatment plants themselves, where they can be used to forecast algal blooms.
"The importance of predicting the blooms is that, due to the associated algal toxins, they present a major water quality problem; and their treatment can be costly. If the model forecasts a bloom four weeks in advance, industry can get the appropriate treatment processes ready and tackle the problem straight away, saving time, money, and maintaining the quality of the water."
Another important aspect to the computer model is the ability to better understand the factors involved in toxic algal outbreaks. Mr Bowden said the computer model could be "interrogated" to reveal the principal factors causing algal blooms.
He's now looking to adapt the computer model for other uses, such as the all-important problem in the Murray: salinity.
"The cost of salinity to domestic and industrial water users in Adelaide, as estimated back in 1984, was US$22 million per year, so the problem is probably much worse now," Mr Bowden said.
"I'm looking at improving the forecasts used for salinity, which could be used in conjunction with other models.
"If you knew a few weeks in advance that a large volume of highly saline water was moving down the Murray, water could be pumped to the Adelaide reservoirs before or after that to avoid pumping the saline water."
But the applications of the computer model don't stop there. Mr Bowden believes the model could be used throughout the world to predict problems of water quality.
"All you need is the right data for a particular water supply and you can apply the same modelling processes.
"Wherever water quality is of concern, it can be used," he said.
Mr Bowden will deliver a paper on his findings at the Hydro 2000 conference in Perth in November this year, and also at the 2nd International Conference for Applications of Machine Learning to Ecological Modelling in Adelaide in November/December.
His research is supervised by Professor Graeme Dandy and by Dr Holger Maier, whose own PhD work formed the basis of this study. Mr Bowden also acknowledged the great support received from Mr Neil Palmer at United Utilities and Mr Mike Burch at the Australian Water Quality Centre.
The above post is reprinted from materials provided by Adelaide University. Note: Content may be edited for style and length.
Cite This Page: