The McArthur Forest Fire Danger Index was developed in the 1960s by CSIRO scientist AG McArthur. It is a score from 0 to 100 to measure the degree of danger of fire in Australian forests.
Additionally, a Grassland Fire Danger Index has been developed
The Fire Danger Ratings (FDR) are derived from the Fire Danger Index (FDI) scores. They use the plain language terms of low-moderate, high, very high, severe, extreme and catastrophic to give the general public an indication of the fire danger and the consequences and risk to life should a fire start.
The diagram below shows a Fire Danger Ratings signboard, an ever present image on the Australia landscape.
Fire Danger Ratings sign
The table below shows the relationship between the Fire Danger Ratings, the Forest FDI and the Grassland FDI.
Note that Victoria uses the term “code red” instead of catastrophic.
Fire Danger Index and Ratings relationship
Forest Fire Danger Index (FFDI)
To calculate the FFDI we can use the tool called the Forest Fire Danger Meter Mk5, also known at the McArthur Meter (see picture below) or computer programs or smartphone apps that perform the same calculations.
Forest Fire Danger Meter Mk5
In order to use the meter to calculate the FFDI, the user must enter four inputs. They are:
- KBDI, days since rain, mm of rain since 9am
- Air temperature
- Relative humidity
- Average wind speed
For an advanced firefighter, the parameters of “days since rain”, “Keech-Byram Drought Index” and “mm of rain since 9am” can be ignored as they are used to derive the drought factor which is include in Bureau of Meteorology reports.
Instructions on how to use the meter are found on the face of the meter. More detailed information is on the reverse side of the meter.
Importantly, the reverse side also has the table for establishing the fire behaviour parameters of rate of spread, flame height and spotting distance.
Grassland Fire Danger Index
The Grassland Fire Danger Index can be found by using the Grassland Fire Danger Meter.
The Grassland Fire Danger Meter works in a similar fashion to the Forest Fire Danger Meter with two important differences.
Firstly, rate of spread has been removed; it should be calculated separately using the CSIRO Fire Spread Prediction System/Meter. Secondly, the index value now exceeds 100 with the index being open ended. While McArthur’s original system was always open ended, the original meter only showed a maximum index value of 100 as it was thought that would be ‘worst possible’ fire weather conditions likely to be expected in Australia. However, this value has been exceeded on several occasions since 1966.
Under the current FDR system a GFDI greater than 150 is deemed to be catastrophic.
In order to use the meter to calculate the GFDI, the user must enter four inputs. They are:
- Degree of fuel curing
- Air temperature
- Relative humidity
- Wind speed
This gives an index of the degree of difficulty of suppressing fire in a standard, average pasture carrying 4 tonnes per hectare.
Mark 4 Grass Fire Danger Meter
Danger Meter Limitations
The Fire Danger Meter is used to give firefighters an approximate guide as to fire behaviour that can be expected.
One of the biggest limitations is how the meter behaves at its extremities. Below an FDI of 12, the meters will tend to over predict, particularly in calm conditions. At FDIs over 100, there is a tendency for under prediction. This is further exacerbated when the atmosphere is unstable (cHaines >10) as the actual ROS can be up to three times higher than predicted.
Human error can also cause inaccurate predictions. This could be from incorrect use of the meter or underestimation of inputs.
It should always be remembered that the results given by any predictive tool are simply an estimation and by no means exact. They are also very susceptible to human error and as such care must be taken when applying observations and interpreting results.
Rate of Spread
Of the four parameters listed above, the rate of spread (ROS) is obtained first as it feeds into almost every other fire calculation we are interested in.
When we refer to rate of spread we are generally referring to the forward rate at the head. However, rate of spread differs on the head, the flanks and the back of the fire. Thus we will examine the:
- Forward rate of spread
- Backing rate of spread
- Flank rate of spread
Forward Rate of Spread
The rate of spread of a fire is measured in kilometres per hour (km/h) and refers to the speed at which the fire travels. In most cases, the forward rate of spread is of key importance as it will often be where the highest intensity fire is and is aligned with the fastest direction of travel. Rate of spread on the flanks and heel may not be the same.
Due to its intensity, the head fire zone is where the most destructive fire occurs as the fire is being driven into unburnt fuel. On a head fire the flame depth is defined as the area behind the fire front that is covered in continuous flame and is often quite deep and depending on the rate of spread can end up being many tens of metres deep.
Backing Rate of Spread
The backing fire on the other hand is generally of lower intensity and moves at a much lower rate of spread when compared to a head fire as the fire will be trying to burn into the wind.
Flank Rate of Spread
A flank fire will generally be of a lower intensity and rate of spread than a head fire. In the flank zone, the fire is essentially burning into itself with the wind. The fire’s rate of spread and intensity perpendicular to the wind (to the left and right) will be similar to the same fire on a calm day (no wind).
Always remember that variations in wind direction, no matter how slight, can rapidly transform a flank fire into a head fire.
Rate of Spread and Topography Upslope
Fire behaviour and slope share an exponential relationship regardless of fuel type. This relationship is formed through the interaction between flame and fuel as the slope of the ground brings the two closer together. Similar to wind, slope will assist pre-heating the fuel ahead of the fire by bringing the flames closer to the fuel, allowing more efficient convective heating of the fuel to occur and increasing the probability of direct contact between flame and un-burnt fuel.
For every additional 10 degrees of upslope change, a fire’s rate of spread will double.
ROS multiplication factor
Above 30° the exponential relationship begins to break down due to complex interactions with convective pre-heating and flame contact.
Effect of upslope
Forest Rate of Spread
Forest ROS is determined by way of a lookup table located on the reverse side of the McArthur Meter and is a relationship between FFDI and fuel load – see table below.
The diagram below is another way of showing the relationship between ROS and FFDI.
Relationship between forest ROS and FFDI
Grass Rate of Spread
Grassland rate of spread is significantly affected by dead fuel moisture, fuel condition, fuel curing and wind speed.
To calculate this rate of spread we use the grassland fire spread meter, specifically produced by the CSIRO to predict the ROS of a grassfire under given conditions. To use the meter correctly the user will need to input:
- Relative humidity
- Grass curing
- Wind speed
- Pasture condition (natural, grazed or eaten out)
The meter will output fuel moisture content (FMC) and rate of spread (ROS). Under an optimum set of conditions, the maximum practical rate of spread for a grass fires is about 20 km/h. Reports of “the fire was travelling faster than my vehicle” have generally been regarded as due to error of parallax than reality.
A fire lit from a spot may spread slower in the initial stages of a grass fire than from a fire line. In the photo below we can see three fires lit at the same time: one fire lit at a spot, one along a 50 metre line and one along 100 metre line. This picture shows how the width of the fire front influences the rate of spread. If the head remains narrow, the fire will spread more slowly than a fire with a broad head.
Grassland Fire Spread Meter
Current research indicates that a grass fire beginning from a 200 metre fire front (as opposed to a spot fire) will immediately reach its maximum intensity. When a flank fire becomes the head fire due to wind change, the size of the fire increases rapidly. If the flank is several kilometres long, it immediately becomes a very large head fire. When mopping up it is therefore important to black out the flank most likely to become the head fire if there is a forecast wind change.
Grass fire spread
Heath Rate of Spread
Rate of spread of a fire in a heath type fuel is dictated by the fuel continuity, density, height and wind speed. In general the density and arrangement of fuel is continuous so wind speed is the driving factor of rate of spread.
Heath ROS relationship to wind speed on level ground
Fire ROS in heath is often described by firefighters as forest fire intensity at a grass fire speed.
Heath fire run driven by 80 to 90 km/h winds
Flame height is the characteristic most often used to describe fires. It is the height of the flame measured vertically above the ground. As flame height is usually variable, an average value should be used. Take sample readings every 30 seconds for 5 minutes and average the results. Practice assessing 1, 2 or 5 metres vertically.
Untrained observers may overestimate the true flame height. In grass fires flame length may be greater than flame height, (depending on wind influence). Flame height is the vertical height above the ground.
Flame length and height
As the above diagram indicates, estimation of flame height can be quite complex, particularly in windy or hilly conditions as the flames themselves will be laid down closer to the fuel bed.
Forest Flame Height
Flame height is a relationship between FFDI and fuel load and is typically bounded by a crown fire. Once a crown fire has been established, measurement of flame height becomes pointless.
The diagram below demonstrates the escalating behaviour between FFDI and flame height and how even at relatively low FFDIs a high total fuel load can result in a crown fire. Also note that anything above 15 metres is assumed to be a crown fire (Forest Fire Danger Meter Mk5).
FFDI vs. fuel load
A crown fire occurs when the intensity of a surface fire is high enough to ignite and burn the tree canopy above. The surface fire is also igniting rising gasses produced by the breaking down of fuel as it burns. Generally a crown fire cannot sustain itself without an intense surface fire underneath, although under severe conditions a crown fire may continue to spread ahead of a surface fire for a short distance until it will ultimately collapse.
Crown fires are the most difficult and dangerous fires to suppress.
Chances of a crown fire developing are higher with:
- Unusually low FMC
- Unstable atmospheric conditions
- Abrupt changes in wind speed and direction
- Mass spotting and formation of junction zones
- Increased surface fire intensity (running up a hill)
- Low tree canopy
- Presence of ladder fuels and high elevated fuel loads
In the lead up to the development of a crown fire, torching of some tree canopy will begin to occur with the surface fire occasionally pulsating or surging into the canopy. At this point even the slightest deterioration in fire weather or terrain is likely to result in the development of a fully formed crown fire.
Once a running crown fire has formed, a fire’s rate of spread will increase significantly and can be up to 2 to 3 times faster than prediction meters indicate.
Fully formed crown fire
Grass Flame Height
With grassland flame height, consideration of the pasture condition needs to be made. As such there are three different flame heights to be considered with the worst obviously being natural ungrazed pasture.
Grassland ROS vs. flame height
Eucalyptus Forest Spotting
Whether it’s short range spotting across a fire trail or long range spotting kilometres ahead of a fire front, when it comes to bush fire suppression, spotting complicates everything. When fires break containment it’s usually attributable to spotting.
Spotting is the process whereby embers are drawn up into a fire’s convection column and blown ahead of the main head fire, igniting new fires in the process.
When fuel types were discussed in Chapter 2, special mention was made of the various types of barks and the hazard they present to bush fire suppression. The probability of spotting and the type of spotting is significantly dependant on those bark types present and their distribution on the fireground.
Bark however isn’t the only thing that causes spotting. The term firebrand is a generic term that can be used to describe any burning fuel that, in this case, can be carried by the wind to start new fires. Firebrand material can include:
- Fern fronds
In vegetation dominated by fibrous barked trees, frequent short range spotting is a continual hazard as spot fires can often land and take hold just out of reach of crews or in a location that would be dangerous for crews to access. These types of spot fires can make an otherwise successful containment strategy, completely useless.
Active spotting - long and medium range - Deans Gap 2013
If numerous spot fires begin around you in unburnt fuel, the main fire front is not far away and you are in danger. Short range spotting may be only a few metres or up to 3 kilometres.
Candle barks present a different type of spotting hazard as they can drift many kilometres aloft due to their aerodynamic properties before they land and initiate spot fires. Spot fires 30 kilometres ahead of an active fireground have been recorded.
One of the key things to remember on a fireground is that when falling embers are successful in starting spot fires, the FMC is beginning to drop to a level that will be conducive to erratic fire behaviour. This is a visual indicator and a vital component of your situational awareness on any fireground.
Eucalyptus forest spotting distance vs. FFDI
The diagram above is a graphical representation of the tables found on the reverse side of the Mk 5 McArthur meter. Notice that fuel load also plays a significant role in the potential spotting distance. Also notice that under an FFDI of only 25 (High to Very High), spot fires can travel up to 2 kilometres.
Pine Forest Spotting
Although spotting will occur in pine forest, it will typically be confined to within 200 metres of the main fire front. Due to the nature of the pine fuel though, spotting may be extreme and under some rare circumstances travel up to 2 kilometres. When compared to adjacent eucalyptus forests though, the pine forest spotting is generally 1/5 to 1/10 the distance.
Spotting in grassland is a much lesser issue than in forest, however it still presents a headache to crews.
Strong wind can often blow smouldering debris such as manure, seed heads or detached clumps of grass ahead of the fire or across control lines. These spotting products often travel no more than 100 metres and don’t materially affect the overall rate of spread of a fire. They do, however, complicate control efforts as they continually challenge established or identified control lines.
If, however, grassland is dotted with the odd fibrous tree, under the right conditions, fire will climb the tree and spotting products can be thrown in a similar way to how it would be in a forest, initiating spot fires ahead of the existing fire fronts.
Intensity and Suppression Thresholds
Fire intensity is an estimate of the heat energy being released by a forest or grass fire per metre of fire line at the head fire. The calculation of this parameter is dependent on two things. Rate of Spread and Fuel load.
Where: I is the Intensity measured in kW/m
w is fuel load measured in tonnes per hectare
r is the rate of spread measured in km/h
This calculation will yield an intensity measured in kW/m. How can we interpret this number though?
If a fire is burning in 20 t/ha of fuel and the rate of spread is 1 km/h the intensity would be:
If an average two bar bathroom heater yields 1 kW of heat energy then the above calculation tells us that this particular fire will have a heat output of 10000 bathroom heaters per meter.
A typical two bar 1 kW bathroom heater
If we were to have an intensity of 10000 kW/m, then the FDI is likely to be between 25 and 50, most likely closer to 50 with a strong likelihood of a fully developed crown fire – see table 15.
If we have 10 metres of fire line in front of us then it would feel like we were standing in front of 100,000 (10 x 10000) bathroom heaters.
The intensity figures previously calculated allow for an appreciation of fire intensity with respect to suppression thresholds. The following table shows approximate fire suppression thresholds in forests, based on 20 t/ha fuel load.
Approximate fire suppression thresholds (forest based on 20t/ha
The following table (Cheney and Sullivan, 1997) indicates levels of suppression difficulty in grassland. Although it pre-dates the introduction of ‘Severe’ and ‘Catastrophic’ to fire danger ratings, it indicates the increasing suppression difficulty as the index moves beyond ‘high’.
Difficulty of suppression in grassland
The suppression difficulty threshold for direct attack in forest is usually about 2000 or possibly 3000 kW/m but up to 10000 kW/m for grass fires. This is because grass fire has a lower heat/flame residence time and flame depth, less spotting and lower flame height than forest fire.
Lighting a backburn ahead of a large grass fire is difficult because of the potential of the fire to change direction under the influence of wind. As grass fires generally move more rapidly than forest fires there is also the problem of time required to complete the backburn.
Graders and similar machines can quickly create mineral earth breaks which are usually effective due to the reduced likelihood of spotting in grass fuels.
Practical Applications of Fire Behaviour
In this section we will look at using fire in:
- Hazard reduction burning
- Backburns to combat wildfire
We will consider:
- Spot vs. line ignition
- Risk vs. reward
- Backburns to combat grass fire
Hazard Reduction Burning vs. Back Burning
Hazard reduction burning is the application of fire under specified environmental conditions to a pre-determined area at a time, intensity and rate of spread required to achieve objectives. These objectives may be:
- Life and property protection – the reduction/modification of fuel loads around assets
- Fire protection generally – the modification of fuel hazards over broader areas to provide barriers to the spread of wildfire and to create areas that will assist in making fire suppression safer and more effective
Prescribed burn plans for hazard reductions are designed to address a variety of issues. These may include lighting patterns, safety, traffic management, smoke management and conditions to meet environmental objectives.
A backburn is a fire intentionally lit along an established control line to consume fuels in the path of an existing wildfire.
For both hazard reduction burning and back burning the task of the firefighters is to achieve the required objectives of the burn. To do so, all of the knowledge of this manual comes into play to keep the fire under control and at the desired intensity.
Spot vs. Line
While there are a number of patterns that can be used when starting a hazard reduction or back burn, the basic choice is between spot and line ignition.
Spot fire ignition is commonly used when a low intensity fire is required. The spacing of the spots can be reduced to increase the intensity.
If conditions are mild, a line of fire can be used to produce a more intense fire. This will also cause a greater reduction in fuel load.
It stands to reason that if a fire is lit up in a single line in excess of 100 metres, the fire will rapidly reach is full intensity, ROS, flame height and spotting potential. On a fire trail which is only a few metres wide, this presents an immediate issue as intensity may be far too great for crews. If the fuel at the edge of the containment line is excessive and dry, the intensity may be too great or too dangerous for crews to operate safely. Beyond 2000 kW/m you may need to be at least ten metres away from the fire.
To mitigate against this, fires can be lit in various patterns such as spots or vertical lines so that ignition points burn away from containment lines a significant distance before they meet up and allow the full ROS potential to be achieved.
Closely spaced spots
Loosely spaced spots
There are pros and cons to each technique and they need to be assessed on a case by case basis, particularly when time is of the essence.
There are, of course, other ways to control the intensity of a fire, including: choosing the time of day to start it and using topography such as slope to control it.
Both fires shown Figure 88 and Figure 89 were lit on the same day, in the same fuel and under a similar FDI.
Hazard reduction downslope ignition
Hazard reduction upslope ignition
Risk vs. Reward
Lighting backburns in difficult fireground conditions becomes a trade-off between risk and reward. The diagram below shows that on a typical bad fire day, the FFDI varied from 18 in the early morning to 63 (Severe) at the peak daily FFDI. The corresponding intensities are between 3500kW/m and 15000 kW/m.
Calculated level ground intensity. Mt Victoria 17-10-2013
Using the understanding that a fire outputting 2000 kW/m has a safe working distance of approximately 10 metres, it can be seen that on this particular day, it wasn’t until after 1700 hrs that the fire line intensity was within a level that supported any kind of reasonable backburning operation.
That’s not to say high risk backburning can’t be considered, just that its risk needs to be well understood and balanced against the potential outcomes under the fire danger conditions being experienced.
Complex backburn environment
Understanding the changes in weather (observed or forecast), reading the land and gaining situational awareness from visual cues (smoke colour, convection columns, spotting behaviour, etc) are all essential to making the right decision.
Backburns to Combat Grass Fire
Lighting backburns to contain grassfires may be complicated by the rapid rate of spread of the fire and the immediate response to wind changes. A ten kilometre flank may have developed in one or two hours so time frames are very different from slower moving forest fires.