Predicting fire thunderstorms

Figure 2: Pyrocumulonimbus from the Sir Ivan fire (12 February 2017). The smoke plume is located between the dashed lines. The Briggs smoke-plume equations describe how the smoke-plume structure changes for various combinations of firepower, fp, and wind speed U, and also how the temperature in the plume decreases with height. Using the temperature and humidity data collected from a weather balloon or forecast model data, the height at the top of the stable barrier, Z, is determined, and the minimum warm humid “boost” θ+, necessary for fire thunderstorms to form, is calculated. This allows the plume temperature at Z to be calculated from θ+, and the corresponding firepower is found by inverting the plume temperature equation. This gives the minimum firepower required for fire thunderstorms to form in this atmospheric environment, a.k.a. the Pyrocumulonimbus Firepower Threshold (PFT). Photo: Sam Crothers NSW RFS.

Fire thunderstorms are thunderstorms that develop in bushfire plumes. Their formation has much in common with conventional thunderstorms: both require warm humid air to be lifted into an unstable layer above. A sufficiently large fire can provide this lift and boosts the temperature and humidity of the lifted air. While meteorologists have developed great skill in forecasting conventional thunderstorms, fire thunderstorm prediction remains a substantial challenge; anticipating this temperature and humidity boost and lift for a given fire is very difficult.

A research team from the Bureau of Meteorology & the Bushfire and Natural Hazards Cooperative Research Centre have discovered how to diagnose the minimum fire heat required for fire thunderstorm development in a given environment. Forecast maps of this minimum fire heat help forecasters identify how the threat of fire thunderstorm formation varies in space and time.

Fire thunderstorms can produce rapid and unpredictable changes to fire intensity and spread caused by intense downburst winds, enhanced ember showers, lightning ignitions and even tornadoes; all of which are extremely hazardous to fire crews. Accurate fire thunderstorm forecasts will significantly improve firefighter safety. This article explains how thunderstorms and fire thunderstorms develop and how the research team from the Bureau of Meteorology & Bushfire and Natural Hazards Cooperative Research Centre have developed an approach that improves fire thunderstorm forecasting. The work parallels research performed by Rick McRae and Jason Sharples on the same topic reported in the April 2018 edition of this magazine (Issue 65).

Thunderstorm development

Most thunderstorms develop when warm humid air near the ground is forced upwards and rises rapidly in an unstable atmosphere. There is almost always a stable barrier between the warm humid air and the unstable layer above (see Figure 1). The upward forcing, or “lift”, needs to force the warm humid air through the stable barrier layer.

Importantly, the upper layer is only unstable if the lower layer is sufficiently warm and humid, and the instability is only realised when clouds form in the lifted air above the stable barrier layer. This occurs when the heat produced by the condensation of water in the cloud makes it warmer and hence more buoyant than the surrounding clear air. This causes the cloud to billow upwards, producing the characteristic puffy, white, cauliflower appearance of such clouds.

The relative strength of each layer determines how favourable the atmosphere is for thunderstorm development. For example, a very warm and humid lower layer with a weak stable barrier and very unstable conditions above will highly favour severe thunderstorm development, because it won’t take much lift to get the warm humid air past the barrier – and the high instability means there’s abundant energy from condensation that can be released. On the other hand, if the lower layer is too cool and dry, the upper layer might not be unstable, and the stable barrier could be quite large, in which case no amount of lift will generate a thunderstorm.

Fire thunderstorm development

Fire thunderstorm development is similar to conventional thunderstorm development in that warm humid air in the smoke plume is lifted until it penetrates the stable barrier, and cloud formation releases the instability in the upper layer (see Figure 1). There are two main differences: first that the fire provides a boost to the lower layer temperature and humidity making conditions more favourable for thunderstorm development and second, the buoyancy of the fire plume provides the lift.

However, if fire thunderstorms are to form, then that “boost” has to survive all the way up to the top of the stable barrier. This is a problem for most fire plumes because turbulence mixes air in from outside the plume diluting the plume gases more and more as they rise. This is why only large, intense fires can produce fire thunderstorms.

A strong boost also decreases the stability, so that an environment that is too stable for thunderstorms can still support fire thunderstorms.

In short, a sufficiently large and intense fire may boost the warm humid lower layer air, providing buoyant lift to penetrate the stable barrier layer, while increasing the instability of the upper layer.

A different approach to fire thunderstorm forecasting

Weather balloons, which measure how the temperature and humidity change with height, provide meteorologists with a precise measure of the warm humid lower-layer, the associated instability (if present) above and the strength and depth of the stable barrier layer. This information is used to make thunderstorm forecasts. In contrast, it is difficult to anticipate the temperature and humidity boost a fire might provide (especially for a fire not yet ignited), which makes fire thunderstorm prediction a major challenge.

A research team in the Bureau of Meteorology tried a reverse approach to the fire thunderstorm prediction problem. Rather than try and anticipate the temperature and humidity boost a specific fire might provide, they found a way to identify the minimum height the plume must rise, Z, and the minimum boost, θ⁺, needed at that height for fire thunderstorms to form (Tory et al. 2018). The values of Z and θ⁺ calculated from weather balloon or forecast model data provide a qualitative assessment of the favourability of the atmosphere for fire thunderstorm development: Large Z means the plume must rise to greater heights, and large θ⁺ means the plume needs to be even warmer when it gets to Z. Both require more heat from the fire.

Figure 3: Spatial distribution of PFT (left panel) and PFT-flag (right panel) as the wind change arrived at the Sir Ivan fire, (5 PM local time, 12 February 2017), near the time the photo in Fig. 2 was taken. The left panel shows a relatively broad band of reduced PFT (higher fire thunderstorm threat), with a value of about 300 GW at the fire site (indicated by the arrow). The PFT-flag shows a narrower band of favourability, because while the plume favourability changed little further to the south and west, the fire danger conditions were weaker, suggesting fires would not be hot enough to produce fire thunderstorms there.

Identifying a firepower threshold

The same research team has since quantified the fire thunderstorm favourability assessment by using the Briggs plume-rise equations to determine the firepower (f_p, the rate heat energy enters the plume) required to deliver the minimum boost, θ⁺, to the minimum boost height, Z, for a given wind speed, U. It is essential to consider the wind speed because the stronger the wind the more the plume bends over, which means the fire needs to produce even more heat for the plume to rise all the way to Z. Figure 2 illustrates how these quantities relate to the smoke plume that initiated the Sir Ivan fire thunderstorm.

Although the Briggs equations were derived to describe the behaviour of smoke emitted from chimney stacks, they have been found to describe well the structure and properties of wildfire plumes. This includes the overall plume shape and how they bend over in the wind. One equation describes how the temperature in the plume decreases with height, for a given firepower and windspeed. By inverting this equation, the firepower needed to produce a plume that reaches the height Z with a θ⁺ boost in a background wind of U, can be determined. This firepower is the theoretical minimum necessary for fire thunderstorms to form, i.e., any less-powerful fire will, in theory, not be able to generate a fire thunderstorm. Thus, it represents a fire thunderstorm firepower threshold, which gets abbreviated to PFT, where the ‘P’ is short for pyrocumulonimbus: the technical term for fire thunderstorm.

Many approximations and simplifications were made to reduce a highly complex process down to a single PFT equation, dependent on only three main input variables:

PFT=Constant×Z²×θ⁺×U

This means the PFT in practice could potentially have some errors. However, PFT forecast maps provide very useful insight into how the threat of fire thunderstorm formation varies with the passage of meteorological phenomena such as sea-breezes, cold fronts and thunderstorm outflows. They alert forecasters to the times and places where fire thunderstorms are most likely. The left panel in Figure 3 shows the PFT associated with a wind change corresponding to the Sir Ivan fire thunderstorm formation (Figure 2).

Using the threshold in forecasting

Fire weather forecasters and fire behaviour analysts have known for decades that weather conditions that favour plume development (warm and humid with light winds) do not favour large and intense fires (hot and dry with strong winds). The opposite is also true. For fire thunderstorms to form the mix of these conditions needs to be just right. The PFT-flag is designed to identify this mix. It is the ratio of the PFT to a modified fire-weather index. Effectively, the PFT-flag filters out conditions where a large fire is unlikely. The smaller the value the more favourable the mix of plume-friendly to fire-friendly weather conditions, and the greater the threat of fire thunderstorm development. The top right panel shows the PFT-flag for the Sir Ivan fire thunderstorm.

Australian fire-weather forecasters in New South Wales and Queensland have been trialling PFT and PFT-flag forecasts during the spring fires of 2019. They have found them to be very useful for predicting periods of high fire thunderstorm threat, several days in advance, and for identifying meteorological features that trigger deep moist pyro-convection. An expansion of the trial is planned for the southern Australian fire season. Feedback from fire-weather forecasters and fire behaviour analysts will help refine the diagnostic and contribute to improved fire thunderstorm forecasts in future seasons.

For more information, go to www.bnhcrc.com.au

References

Tory, K. J., 2018: Models of buoyant plume rise. Bushfire and Natural Hazards Cooperative Research Centre Research Report No. 451. https://www.bnhcrc.com.au/publications/biblio/bnh-5267
Tory, K. J., W. Thurston and J. D. Kepert, 2018: Thermodynamics of pyrocumulus: A conceptual study. Mon. Wea. Rev., 146, 2579–2598. DOI: 10.1175/MWR-D-17-0377.1 https://journals.ametsoc.org/doi/pdf/10.1175/MWR-D-17-0377.1

About the Author

Kevin Tory

+ posts

Kevin Tory is a Senior Research Scientist with the Australian Bureau of Meteorology & Bushfire and Natural Hazards Cooperative Research Centre. His research interests include tropical cyclones, thunderstorms, tornadoes, and recently fire weather. His current research on pyrocumulonimbus fuses his passion for fire weather and thunderstorms.