All of Australia will from next year have access to a new modelling system that will improve the scope, accuracy and length of weather forecasts, which could help prepare for future repeats of Queensland's devastating floods.
Prime Minister Julia Gillard said the government would "look into" flood forecasting models, but conceded improvements may be limited.
This is because weather prediction is not an exact science, as it's constrained by principles such as Chaos Theory, which limits the length of predictions.
But new projects by the Bureau of Meteorology will boost the accuracy of predictions and the scope of weather data on offer.
This includes tweaks by United States mathematicians to 89-year-old prediction algorithms created during the First World War and still in use by meteorologists today.
It will be built into a data-modelling system that will be implemented in Australia's southern states this year and Queensland during 2012. As part of a $45 million five-year Next-Generation Project, the modelling system will allow meteorologists to spend less time typing in numbers into an antiquated database and more on analysing changing weather patterns.
Project leader Terry Hart, a 38-year veteran bureau forecaster and researcher now in the weather and oceans services department, said weather prediction is still based largely on human interpretation, but the new modelling system will make predictions more accurate and available.
"We want the forecasters to be able to focus and identify high-impact weather systems, like the floods or the Black Saturday bushfires."
"Predicting weather is a bit like quantum mechanics. It can't be 100 per cent accurate because we don't have measurements at every point all over the world, and those points of uncertainty grow in time.
"You may predict the weather in the dry season with accuracy of a couple of weeks, but perhaps only a few days say in Queensland where a low pressure system may develop very quickly."
Hart said the limit of weather prediction in accordance with Chaos Theory is limited to two weeks, and said forecasts can be made accurately up to seven days.
To this end, the bureau will use the data-modelling system to introduce probability accuracy ratings into its predictions, which will be included in television weather forecasts.
The Graphical Forecast Editor, part of the modelling system, will make life easier for meteorologists by pre-populating forecasts with predictions based on raw data. Previously, meteorologists had to manually enter data, meaning they were limited in the number of forecasts they could do due to time constraints.
"Forecasters would look at the raw data, work it out in their head what the prediction is and would have to write out the information for every [suburb]," Hart said.
"This model changes them from being writers to painters."
Meteorologists will enter data into global models to produce weather patterns for the entire country. Grid areas have also been redefined to be far more accurate, meaning forecasts can be localised to an area as small as 3 kilometres.
Hart said the biggest beneficiaries will be regional Australians who must rely on state forecasts to determine local weather. This method was far less accurate than those enjoyed by metropolitan residents because the bureau did not have the resources to produce the localised predictions.
Users can visit the bureau website to view a digital map and click on any unspecified area to see precise and specific data.
"It means residents will have weather particular to their region, which might vary considerably between communities at the foot of a mountain to adjacent areas in an open [plane]," Hart said.
The modelling system, built in the US but adapted for Australia, runs over the bureau's $30 million Sun supercomputer, and can support software additions customised for any of the bureau's divisions.
NSW will receive the Graphical Forecast Editor in September, along with South Australia and the ACT. It will hit Queensland, Western Australia and the Northern Territory from next year, but it is already up and running in Victoria and Tasmania.