Australia has developed nationally consistent bushfire modelling and prediction technology aimed at boosting support for frontline fire crews.
The new Spark Operational bushfire simulation technology has been developed by the CSIRO and the National Council for Fire and Emergency Services (AFAC).
When fully operational, the bushfire predictions will be used by fire and emergency services agencies across Australia.
Currently, emergency planners and frontlines crews can be hindered by varying approaches to bushfire modelling in different states and over different types of landscapes.
CSIRO Chief Executive Dr Larry Marshall said last year’s devastating bushfires had “changed the game forever”.
“So, we have changed our game too, by unleashing new science and technology to protect our firefighters and Australian communities,” Dr Marshall said.
“This advanced system will help firefighters outthink fire, to anticipate its actions and to get ahead of it, so they can beat it.”
Dr Mahesh Prakash, the senior principal research scientist at the organisation’s data research arm, CSIRO Data61, said the technology would give fire crews and emergency planners a better idea of where a fire would travel over the course of up to 10 hours.
“It’s a predictive model, which allows a range of people to forward predict what bushfire spread will look like, both spatially and temporally,” he said.
“So it allows planners, decision-makers and emergency managers to get an indication of what a given bushfire might look like, in the next five to six to 10 hours or so.”
He said that would allow crews to be better prepared and to know which fires they should focus most attention on.
Dr Prakash said there were clear benefits to having bushfire modelling that was nationally consistent.
“Right now, what happens is when you make a prediction in one state, and you’re crossing state boundaries, it doesn’t quite work in a consistent way.
“For example, Victoria might have a given prediction and New South Wales might have something slightly different.
“So we really do not want that going forward, and that’s the key reason why we have been developing something that cuts across state boundaries.
“There’s been some issues with consistency across state boundaries in terms of providing warnings, especially in those border towns, so having a nationally consistent model would be hugely helpful there,” Dr Prakash said.
He said different climatic conditions and types of vegetation, key parameters that drive bushfire behaviour, also varied widely across Australia, but the new technology would produce nationally consistent predictions.
The new technology builds on the Phoenix RapidFire bushfire simulator, developed by the University of Melbourne and others that has been used for a number of years to understand and predict bushfire behaviour.
It will be integrated with the latest meteorological information, as well as land slope, vegetation and other landscape information to help emergency planning.
The fully operational version of Spark Operational will also help fire agencies to generate detailed data driven emergency warnings to communities.
The technology’s development has received funding from the Minderoo Foundation and is expected to become fully operational in the next two to three years.
