[Voice over: Phil Brien, Fire Behaviour Specialist]
Speaker: Phil Brien
So we're looking today at some of the considerations that have gone into the development of the Taradale planned burning case study.
For those who may not be aware, there was a fire which started near Taradale on January 9th, which was the day that we saw some very large destructive bushfires occur across Victoria.
In particular, of consequence, we think about the Ravenswood-Fogarty Gap fire, we think about the Longwood fire, and we think about
the Wawa fire, making this one of the most destructive single-days bushfire runs in Victoria in the last several decades.
What was really interesting about this fire at Taradale is it was suppressed at less than one hectare in size, when it actually had the potential to do significant damage to communities around Drummond, Malmsbury, Lauriston, and even Kyneton.
When we're looking at these case studies and how they occur, the first thing we need to look at is the weather, which is either forecast to occur on the day or that has occurred on the day if we're looking back in time.
On a forecast day, we can look at things such as the Forest Fire Danger Index and the Grass Fire Danger Index, which are key indices used to help us think about fire spread and fire intensity. We can also look at the Fire Behavior Index, which has replaced that given the implementation of AFDRS in 2022.
When we look at both of these indices together, we get a very quick high-level view of what may happen in the landscape.
We can break this down further, so we can look at things such as drought factor, Keetch-Byram Drought Index, the curing in the grasses, grass fuel conditions, grass fuel loads, temperature, relative humidity.
One of the most important elements for us to think about is the fuel that are there in the landscape.
So we can look at things such as the Phoenix fuel type, Phoenix being one of our main bushfire simulation tools.
So if we zoom right in these areas around Castlemaine, in around Taradale, we can see very quickly that our dominant fuel type here is forest with shrub fuel type 3006, which is one of the highest fuel loads and highest maximum fuel hazard fuels we've got in all of Victoria.
Phoenix is our main simulator we use in Victoria.
So when we're doing a bushfire simulation, there are all sorts of data inputs we have to consider. This is the fuel that's there, the topography, how close the roads are to the areas we're looking at, whether or not there are any disruptions in the landscape which might halt or impact the spread of the fire, the location of the different assets that we care about, the wind modifiers, the rates of fuel accumulation within the forest, and what suppression resources we have available.
It's also important for us to consider how fuels accumulate over time.
So what's loading up on the screen in front of me is all of the different fuel types which we use in our modelling in Victoria, quite literally hundreds of them.
And if we scroll all the way down to 3006 here, which is the dominant forest type around the Taradale fire, we can have a look at how different components of fuel hazard, and also the load which is calculated from that, accumulate after exposure to fire.
So what we can see is our elevated fuel loads here generally max out at about six tons to the hectare between 15 and 20 years after
fire, and that gives us an overall fuel hazard of four or very high.
When we combine this with our surface fuels and our bark fuels, generally we start to think that an area like this that has been burnt starts to carry fire in as little as eight years and returns to maximum potential fire behaviour in as little as 15 years.
When we also think about the undulating topography, the continuity of the forest here, this is really a convective hotspot for our region, where we have potential for the largest and most intense bushfires.
When we start thinking about that convective behavior or convective potential, it's also important that we understand what's happening in the atmosphere.
We get access to all of the Bureau's weather station data, and if we zoom right out and we pick a nearby area, let's say Bendigo, we can view what they've modeled is happening at different levels of the atmosphere throughout the day.
By understanding what's happening at these various levels of the atmosphere, in some cases up to 13 or 15 kilometers above the surface, we can actually understand how that's going to change the fire behavior we see on the ground.
And on days like Black Saturday, Ash Wednesday, what's come to be known as the Black Summer bushfires, these complex atmospheric interactions have really been dominant drivers of fire behavior.
So assuming all of this information is accurate, and what we're doing here is looking at what may have happened had there been no planned burning in the landscape, we can hit run.
Now, what Phoenix is doing here is it's looking at a modified version of a MacArthur rate of spread, which is an old forest model developed in the 1970s that's still highly accurate today.
So when we look at how that would have spread through the landscape, we're seeing fire spread, which is these yellow lines, over time
under a northwesterly before that westerly and southwesterly change hits.
What's most interesting here is because we're looking at MacArthur, realistically, this fire is probably spreading a lot further to the
southeast than these yellow lines suggest, and each of these red dots, which indicate potential spot fires, some of these are
probably taking and forming new fires.
So if we're looking at this as fire behavior analysts talking about where the actual risks on the day are, we're saying our potential impact area is probably something a little bit more like this.
And that's really important for things like our community warnings, for prioritizing how and where we're going to be trying to put this fire out, and recognizing that looking back, when we're talking about what's been achieved by keeping this fire small, we're only talking about that small area in the yellow.
But risks to Carlsruhe, Kyneton, Green Hills, they actually existed, and this could have been a lot worse than even the simulator is suggesting.
If we wanted to test that further, we can use one of these spreadsheets where we have access to probably 20 different fire behavior models that are used around Australia.
So when looking at this fire and what it's done and when developing these case studies, it's also important that we don't just rely upon data inputs layers, because these aren't really real until you've gone out and viewed them yourself.
So my team and I have gone out into these areas quite recently, around Taradale Lower Kangaroo Track, and we've walked through the areas that have been previously burnt.
And if I find our fire history, what we're looking at here is our 2022 burn, the burn which effectively stopped the spread of that fire.
And if we go back just a little bit further, you're also seeing the 2008 burn where the fire started.
Now, my team and I have gone out there with some research scientists from the CFA and from DEECA. We've actually physically assessed the fuels and had a look at how the different forest species have changed accumulation rates over time, assessed how things like long-term drought have slowed that down, and viewed how the fire likely behaved on the day.
And what's really interesting is when we've gone out to this 2008 area, we were expecting to see fuel loads up around 30 tons to the
hectare in some areas.
But the fuel loads we're actually observing were down around the high teens and low 20s through a lot of the area around where the fire
consumed.
You didn't have to go far before you found those 30, but we're talking about fire intensity being two-thirds of what it could have been just in that area.
What we've witnessed is the fires actually consumed surface litter at a very slow rate, and we think this is because the continuity or connectivity in that surface layer is quite low.
It hasn't allowed the fire to build rapidly, it hasn't allowed it to spread quickly, and it hasn't allowed it to start consuming the next level of
strata of fuel, so moving from the surface into the near surface, into the elevated.
What's happened is when the fires come down to this track here, it's got to the edge and it hasn't actually had enough intensity to get across that road in most cases.
When we've gone down and we've observed what's actually happened here, the fire has started to consume a little bit of elevated fuel and a little bit of bark fuel in a single tree on that road, which has allowed it to spot across the road into the 2022 burn.
Had the 2022 burn not been here and we were looking at 20 to 30 tons to the hectare of fuel and extreme fuel hazards, it would have been game over.
That fire would have consumed that fuel rapidly.
In areas nearby, we saw extreme surface fuel hazards, we saw extreme bark hazards with stringy bark. It would have taken maybe half an hour to an hour for that fire to have established, which even under these midmorning conditions would have been really hard to suppress.
Because we're in this 2008 burn where we've gone out and observed super low surface fuels, no continuity, bark fuels that will not spot, it's allowed our crews to get in there and actively suppress this fire very rapidly.
And what you're seeing on your screen is there was almost no spread into that 2022 burn. What it's done is it's delayed the buildup of that fire, it's really mitigated the potential fire behavior, and it's allowed our local crews from Castlemaine and surrounds to get in there and rapidly stitch this thing up.
If we think about what happened at Ravenswood only a couple of kilometers to the northwest, that gives you a great indication of what could have happened otherwise.
And when we think about how many of our firefighters, our resources, our aircraft were all targeted at Ravenswood earlier in the day,
it's a little bit scary to think about what may have happened if we would have had to split them across two fires.
Page last updated: 20/08/26