Warning to bushfire fighters

Bushfire fighters in the eastern states could be risking their lives if they rely on the current forest fire spread prediction tables.

New Project Vesta figures show that the forest fire spread prediction systems developed by bushfire researcher Alan McArthur in the mid 1960's can under-predict the potential rate of spread by as much as 3 times.

CSIRO has warned rural fire brigades around Australia of the new findings.

Almost a hundred experimental fires were lit during Project Vesta, a six-year, multi-million dollar study investigating the behaviour of summer fires in dry eucalypt forest.

The research project involved CSIRO and the Western Australian Department of Conservation and Land Management (CALM), supported by Australian fire and land management agencies, local government and corporate sponsors.

"These warnings arise from preliminary analysis of the field data collected over the past two summers. The warnings are posted as a guide only and may be subject to change as the evidence gradually emerges from still to be completed data analysis," says CSIRO's Mr. Phil Cheney, leader of Project Vesta.

A preliminary examination of the behaviour of experimental fires conducted during Project Vesta has raised a number of important points that include:

McArthur forest fire spread table under-predicts

The fire spread table on the back of the McArthur Mk 5 Forest Fire Danger Meter under-predicts the potential rate of spread over most fire danger indices.

Shrub fuel is important in fire spread

Forest fires in fuels with a developed shrub layer taller than one metre can spread up to three times faster than predicted by McArthur's forest fire spread table. Fires in litter fuels with a low shrub layer can spread two times faster.

Watch out when the wind speed reaches 15 km/h

There appears to be a threshold wind speed around 12-15 km/h in the open that makes a huge difference in the behaviour of forest fires. Fires in heavy fuels may spread deceptively slowly, well below their potential rate of spread, when the wind speed is below the threshold. A slight increase in wind speed can result in a big jump in fire behaviour.

Forest winds are spatially very variable

Fire behaviour observed at one location is not necessarily the same elsewhere in the forest. Detailed wind measurements showed that gusts under the canopy do not travel more than 40 m. Five-minute mean wind speeds at one location can be ± 20% of the measured value at another location. This can make a big difference in fire behaviour, particularly around the threshold wind speed.

Line fires don't wait

A fire starting from a line greater than 100 m long will burn at its potential rate of spread immediately. It may take 2-4 minutes for the flames to develop their full dimensions but the fire is already travelling at full speed before this happens. Conversely, a fire lit from a point ignition and whose head fire remains narrow may spread all day and still not reach its potential rate of spread.

Firefighters are advised to:

Despite these research results, there is no reason to change the McArthur Forest Fire Danger Rating System, says Mr Cheney.

"The weather has not changed and the present manning levels based on experience should be still applicable'" he says. "Likewise, the degree of suppression difficulty at each fire danger class is the same as it always was. However, forest fires have a considerably higher potential rate of spread than that predicted by the table on the back of the meter."

The fire spread table in McArthur's Forest Fire Danger Meter was derived from relatively small fires of intensities generally less than 2000 kW/m and observational reports of spread of wildfires, says Mr Cheney. McArthur's table may still predict reasonably well during the first hour of development of a fire if the head fire remains narrow - but be aware that there could be a sudden jump in fire behaviour if the head fire becomes wider. Under prescribed burning conditions satisfactory results will be achieved using the accepted prescribed burning guide.

"We have a great deal of analysis to carry out on the Project Vestadata but the above warnings stand out," says Mr Cheney. "There appears to be a complex interaction between the fuel structure (litter, shrubs, and bark), wind speed, and fire spread and it will be at least two years before we can quantify all the important variables."

Reprinted here with permission from:
Mick Crowe

Communication Manager
CSIRO Forestry and Forest Products