The 2009 Victorian Bushfires Royal Commission identified a need to improve the ability to assess the potential impacts of bushfires on rural and urban/rural interface communities. This need related to both an immediate threat and possible extreme scenarios involving future threat.
The project developed a proof of concept simulation system that demonstrated that it is possible to provide critical fire-planning information to emergency services, government and the public in an integrated system.
The Fire Impact and Risk Evaluation Decision Support Tool (FireDST) – proof of concept shows that an integrated systems can assist fire managers to make quick decisions in extremely complex situations. FireDST could help fire managers and incident management teams decide where to direct firefighting resources by simulating the potential impacts a bushfire may have on community assets, infrastructure and people.
Geoscience Australia's Bob Cechet led the FireDST project, working with Geoscience Australia researcher Ian French. FireDST involved a high-calibre collaboration on separate but related projects (see below) with the University of Melbourne's Dr Kevin Tolhurst, Dr Jeff Kepert from the Bureau of Meteorology, and Dr Mick Meyer from CSIRO Marine and Atmospheric Research, who led their respective CRC projects.
The FireDST proof of concept is a probabilistic system. It uses probability theory to predict the likelihood of neighbourhood and house loss, the potential health impacts of bushfire smoke and identifies the areas that are most likely to be affected in a bushfire. The system achieves this by processing known information about ignition location, weather (humidity, temperature, wind speed/direction) and vegetation. FireDST can be run under predicted weather conditions, and be modified for different scenarios, such as changes to wind strength and direction. It uses a parallel computing platform to produce a large number of simulations almost as quickly as a single simulation, thereby creating a suite of possible outcomes.
The FireDST project provided fire agencies with a preview of decision support approaches that might possible as computational capacity increases in five to ten years’ time.
As a prototype, FireDST has already revolutionised fire simulation modelling. It promises to be an invaluable asset to fire managers working under pressure to make quick decisions that affect lives and infrastructure.
In his PhD research, the University of South Australia’s Alex Sims quantified the danger of driving in a bushfire. He has examined the factors that contributed to that risk when evacuating during a fire and compared the results with other options, such as abandoning a car to find shelter or staying and defending property.