🔥 Fire Predictor Pro · 3 hour fire spread

Starting…
⚠️ Indicative sketch only – not an official fire spread prediction. Shapes are simplified estimates from modelled weather and assume flat ground and uniform fuel. Verify with Aurora, a qualified FBAN, official forecasts and what you see on the fireground. Read the disclaimer at the bottom of this page before use.

1 · Fire location (ignition point)

Search an address or place, tap the map to drop the pin, drag the pin, or type coordinates.

Weather at the ignition point (Hour 1 – modelled unless you've typed over it)

Temperature
–
°C
Humidity
–
% RH
Wind (10 m)
–
km/h
Wind from
–
 
Delta-T
–
 

2 · Fuel type

90%
Adjust grass condition, continuity & fuel load
Considerations (optional – off by default)
⚠️ The grassland and grassy woodland models do not allow for recent rainfall. Forest allows for it only through the Drought Factor, and mallee and heath only through rain in the last 48 h. Fuel wet from recent rain may burn slower than shown until it dries out.

3 · Weather for each hour (modelled forecast – edit any value)

Temp °C
RH %
Wind km/h (10 m)
Wind from °
⚠️ Modelled weather – not observed. Open-Meteo forecast for the pin location. Check against a Belt Weather Kit, the portable station or official observations and type in corrections. Typed wind is treated as 10 m open wind.

Predictions (Rough prediction from forecast weather – may vary or be incorrect)

Fire intensity guide & suppression tactics (kW/m)

< 500
Direct attack at the head with hand tools and crews
500 – 2,000
Direct attack with tankers and machinery
2,000 – 4,000
Direct attack at the head difficult – parallel/indirect attack, aircraft support
4,000 – 10,000
Direct attack unlikely to succeed – indirect attack, backburning, aircraft on flanks
10,000 – 30,000
Head fire uncontrollable – defensive tactics, asset protection, flanks only
> 30,000
Extreme – suppression ineffective, focus on life safety and warnings
General rule-of-thumb guide only. Always follow DFES doctrine and the direction of the IC or Ops.
How the shape is drawn

Hour 1: the head travels ROS × 1 h from the pin, in the direction the wind is blowing to. The tear drop is then drawn back from that head point: round at the head, widest near the front and tapering to a point behind the ignition.

Hours 2 and 3: every point on the previous hour's edge spreads as its own small tear drop for that hour's ROS and wind (Huygens' principle, the same idea used by fire spread simulators such as Phoenix and Spark). The new head starts from whichever part of the old fire faces the new wind: the head if the wind holds, or a flank after a wind change. The flank then becomes a wide new head front.

Default – wind-based: see below. The fixed ratio is kept as an option for comparing with hand drawings.
Fixed ratio, Length : Width : Back: with 4 : 2 : 1, the fire is twice as long as it is wide, and ¼ of the length is behind the ignition point (the back burn), so ¾ is the head run.
Fixed ratio, Head : Flank : Back spread: the head, each flank and the back spread at 4 : 2 : 1 of each other.
Wind-based: grass uses length : breadth = 1.1 × U0.464 (U ≥ 5 km/h, Taylor 1997 as used in CSIRO Spark); other fuels use 1 + 0.0012 × U2.154 (Alexander 1985, also used in the NSW RFS short-fire-run method). The back spread comes from the ellipse geometry, so it gets very small in strong wind.

Positions are calculated in metres on the ground and converted to latitude/longitude with great-circle maths, so the overlay matches the map scale at any zoom. Assumes flat ground, continuous uniform fuel and a fire that is already established. A new point ignition, especially in grass, takes time to reach full speed. Slope, spotting, breaks and suppression are not included.

Disclaimer. Fire Predictor Pro is provided for general information only. It is an independent, unofficial tool and is not produced, endorsed or validated by AFAC, DFES, Landgate, the Bureau of Meteorology or any fire agency. The predicted fire shapes are simplified geometric estimates, not a fire spread simulation or an official prediction, and must not be presented as one.

Outputs may be inaccurate, incomplete or out of date because of errors in the weather forecast or inputs, default fuel values, model limitations, software errors, or conditions on the fireground that differ from those assumed. The tool does not account for slope, terrain, spotting, fuel breaks, changes in fuel, fire acceleration from a new ignition, or suppression. Actual fire spread can be much faster, slower or in a different direction than shown.

Do not rely on this tool as the sole basis for any operational, safety, warning, evacuation, planning or other decision. Always use official sources (including Aurora, AFDRS fire danger ratings, BoM forecasts and warnings, and qualified Fire Behaviour Analysts), together with your training, local knowledge, the direction of the IC and direct observation of the fire. Fire behaviour can change rapidly and unpredictably.

To the maximum extent permitted by law, the author accepts no responsibility or liability for any loss, damage, injury or death arising directly or indirectly from the use of, or reliance on, this tool or its outputs. By using this tool you accept these terms.

Models: as used in Fire Behaviour Pro and the Landgate Aurora Fire Behaviour Calculator – CSIRO Grassland (Cheney et al. 1998; Cruz et al. 2015) and Grassy Woodland; Dry Eucalypt Forest/Vesta (Cheney et al. 2012); Mallee Heath (Cruz et al. 2013); Heathland/Temperate Shrubland (Anderson et al. 2015). Fire shape: length-to-breadth from Taylor (1997, as in CSIRO Spark) for grass and Alexander (1985) for other fuels. Optional slope rule: McArthur, R = R × exp(0.069θ) (Noble et al. 1980), limited to ±20°. Weather: Open-Meteo modelled data or user entry.

Created by Ben Davies