This year’s fire season in south-east Queensland and north-east New South Wales has already resulted in loss of life and property, propelling the issue of increasingly early fire danger into the national debate.
At a media conference on September 8, Inspector for Queensland Fire and Emergency Services Andrew Sturgess described the weather conditions before the fires as “an historic event”.
“We’ve never seen fire danger indices, fire danger ratings, at this time of the year, as we’re seeing now. Never seen this before in recorded history.”
Is that correct? RMIT ABC Fact Check investigates.
Sturgess is close to the mark.
In making the claim, he referred only to Queensland, so Fact Check examined only records for Queensland in assessing the claim.
Fact Check obtained daily records from the Bureau of Meteorology about bushfire danger since January 1950 for four south-east Queensland fire districts: Port Curtis; Moreton; Darling Downs; and Maranoa Warrego.
The records are a district average of the Forest Fire Danger Index (FFDI) for the day, which is comprised of data including temperature, wind speed, relative humidity and drought.
They showed the index reached unusually high levels for the time of year in south-east Queensland in the first eight days of September 2019.
On September 6, Maranoa Warrego, Darling Downs and Port Curtis recorded higher average FFDI values than at any time in those districts in any June, July, August or early September back to 1950.
In Maranoa Warrego and Darling Downs, the FFDIs recorded on September 6 were the two highest recorded between June 1 and September 8 of any year in any district.
But in the district of Moreton, stretching along the coast from the NSW border to the north of Brisbane, higher values than were recorded in early September were recorded in August 1991 and August 2004.
How do we measure fire danger?
In the 1950s and 1960s, a fire scientist named Alan McArthur devised a way to forecast the potential danger posed by bushfires in forested areas of Australia.
In the McArthur Forest Fire Danger Meter, first seen in operation in 1967, McArthur combined measurements of air temperature, relative humidity, wind speed and fuel conditions with a formula to account for the effect of drought.
The drought formula considers how long it has been since it last rained and the amount of the last rainfall to assess how dry or moist the soil is, and therefore how combustible the fuel is likely to be. The fuel is the bark, leaves, and small branches accumulating in the landscape.
McArthur used the measurements to create a Forest Fire Danger Index from one to 100, in which one represented the lowest risk and 100 the highest.
The meter is used by the Bureau of Meteorology and rural fire authorities in forecasting the likelihood and potential severity of bushfires.
What are fire danger ratings?
Sturgess spoke of “fire danger indices, fire danger ratings”.
The danger ratings are published on signs by the side of the road as well as the news media.
The bureau develops the danger ratings in consultation with state fire authorities. Other risk factors may be considered when determining a fire danger rating, but the Forest Fire Danger Index is a major factor.
When the index was introduced, the range of 1 to 100 was translated into five danger ratings: Low; Moderate; High; Very High; and Extreme.
After the Black Saturday bushfires in Victoria in February 2009, the system was revised nationally to include index values above 100 and new ratings categories.
The Queensland government adopted two new categories in September 2009: Catastrophic representing index values above 100; and Severe between Very High and Extreme. Low and Moderate were combined, leaving six categories.
|Forest Fire Danger Index||Fire Danger Rating category|
|100+||Catastrophic (Code Red*)|
*Victoria uses the term “Code Red” instead of “Catastrophic”.
- Low-moderate: a fire that can be easily controlled and poses little or no risk to life or property
- High: a fire that can be controlled, where loss of life is unlikely and damage to property will be limited
- Very high: a fire that can be difficult to control, with flames that may burn into the tree-tops. During a fire of this type, some homes and businesses may be damaged or destroyed
- Severe: a fire that may be uncontrollable and move quickly, with flames that may be higher than roof tops. A severe fire may cause injuries and some homes or businesses may be destroyed
- Extreme: a fire that may be uncontrollable, unpredictable and fast moving. The flames will be higher than roof tops. During an extreme fire, people may be injured and homes and businesses may be destroyed
- Catastrophic: a fire that may be uncontrollable, unpredictable and fast moving. The flames will be higher than roof tops. Many people may be injured and many homes and businesses may be destroyed.
For fires expected to be in the “extreme” and “catastrophic” categories, even well-constructed homes that are well-prepared for fires are no longer safe, and residents are warned that leaving is the only option for survival.
Sturgess made his claim at a media conference about the bushfires that had broken out in south-east Queensland and northern New South Wales.
He referred only to Queensland, so Fact Check has not examined records for New South Wales in assessing his claim.
He mentioned the Queensland towns of Applethorpe, around 200km south west of Brisbane, and Sarabah, 90km south of Brisbane.
He closed his remarks by saying: “So this is an omen if you will, a warning of the fire season we are likely to see ahead in the south-eastern parts of the state, the driest parts of the state, where most of our population is.”
Given this context, Fact Check has assessed the claim against historical records for south-east Queensland.
Which time period?
Sturgess made his claim on September 8 and said “at this time of the year”.
After making the claim at the media conference he said: “Fire weather has never been as severe this early in spring.”
Given this context, Fact Check has assessed the claim against historical records for the months of June, July, August and the first eight days of September.
In other words, Fact Check considers “this time of the year” and “this early in spring” to refer to any time after temperatures begin rising from their winter lows.
What data is available?
Fact Check approached the Bureau of Meteorology, who suggested a dataset of averages for four south-east Queensland fire danger districts — Darling Downs, Maranoa Warrego, Moreton and Port Curtis — that are based on the bureau’s rainfall districts, published on its website.
The dataset was developed by bureau researcher Andrew Dowdy. It begins in 1950 and divides Australia into a grid of approximately 5km by 5km.
An index value is recorded for each grid point for each day, using the maximum temperature for the day, humidity at 3pm, the afternoon wind speed and a drought factor.
The bureau told Fact Check the gridded dataset was “intended for broad scale analysis at a regional level” and provided Fact Check with daily averages of the grid points in each district. It confirmed that the four districts being analysed were suitable for this regional analysis.
Andy Pitman, director of the ARC Centre of Excellence for Climate Extremes at the University of New South Wales, said he would use data from individual weather stations in the target area.
He said there could be issues with averaging over a large area: “If you average over an area you can imagine a situation where half a district beats all records, and half is low giving you an average that is lower than truly reflecting the values in one region.”
But he said there was merit in BoM’s approach, noting that the bureau may have a reason for suggesting it “that makes it more appropriate to the data they know is reliable”.
“I do not think there is a right/wrong here,” he told Fact Check in an email.
Hamish Clarke, a research fellow at the School of Atmospheric and Life Sciences at the University of Wollongong, said that using data from individual stations, as well as averaging gridded values, were both valid methodologies.
“An average value will always obscure higher and lower values … You just need to be clear what you’re reporting.
“Note, the same logic applies to station records: we assume that the weather station represents the average of the area surrounding it.”
Based on this advice, Fact Check has used daily district averages from the bureau’s gridded dataset to assess the claim.
What does the data show?
The bureau provided Fact Check with values for the Forest Fire Danger Index every day from January 1, 1950, to September 30, 2019 for four districts in south-east Queensland: Darling Downs; Maranoa Warrego; Moreton; and Port Curtis.
Each value is an average of all the grid points in the district for that day.
Fact Check has plotted the data relevant to Sturgess’s claim on the graph below.
All values from the same date of each year are plotted in the same vertical column, illustrating their historical severity.
The highest value for 2019 (shown in red on the graph) between June 1 and September 8 occurred on September 6 for all four districts.
Darling Downs recorded 71.97, Maranoa Warrego was 81.76, Moreton was 54.57 and Port Curtis was 51.54.
For Darling Downs, Maranoa Warrego and Port Curtis, these were higher than any value for the months of June, July, August or the first eight days of September in any year since 1950.
For Moreton, there were higher values on August 14, 2004 (65.28) and August 24, 1991 (55.88).
Looking at the four districts as a single region, the September 6 records for Maranoa Warrego of 81.76 and Darling Downs of 71.97 were higher than any other day in this period.
The next highest (67.9) occurred in Maranoa Warrego on September 1, 1961.
What happened in September 2019?
As outlined by the bureau in a special climate statement on September 24, the event began on September 2, with a fire burning in Sarabah in the Gold Coast hinterland.
On September 5 and 6, south-east Queensland was hit by an almost record-breaking early season heatwave. The bureau said the soaring temperatures were “generally second only to the exceptional winter heat” of August 23 and 24, 2009.
Rainfall in the lead-up to September 2019 had been below average on a range of timescales from months to years,leading to a prolonged and severe drought.
The bureau said at almost all sites in south-east Queensland, the drought factor (the dryness of fuels) was at or near the highest value used for the Forest Fire Danger Index.
From September 6, this combination of high temperatures, low humidity, strong winds and dry conditions led to elevated fire danger.
The Sarabah fire, having burned for several days, spread to Binna Burra on September 6. Fires also took hold around Stanthorpe and Applethorpe, with homes destroyed overnight.
By September 7, more than 50 fires were burning across Queensland, with the rainforest tourism property Binna Burra Lodge engulfed on the morning of September 8.
Principal researcher: Matt Martino, online editor
Additional research: Lucinda Beaman, Sydney editor
- Bureau of Meteorology, Forest Fire Danger Index, daily averages, January 1, 1950 to September 8, 2019
- Bureau of Meteorology, Special Climate Statement 71, Severe fire weather conditions in south-east Queensland and north-east New South Wales in September 2019, 24 September 2019
- CSIRO, McArthur Mk5 Forest Fire Danger Meter
- CSIRO, Fire danger meters
- Geoscience Australia, scientific topics, community safety, bushfire
- Chiara M. Holgate, Albert I. J. M. van Dijk, Geoffrey J. Cary, Marta Yebra, International Journal of Wildland Fire, Using alternative soil moisture estimates in the McArthur Forest Fire Danger Index, 18 August 2017
- Andrew J. Dowdy, Graham A. Mills, Klara Finkele and William de Groot, Australian fire weather as represented by the McArthur Forest Fire Danger Index and the Canadian Forest Fire Weather Index, The Centre for Australian Weather and Climate Research, June 2009
- Christopher Lucas, Centre for Australian Weather and Climate Research, Australian Meteorological and Oceanographic Journal 60, On developing a historical weather data-set for Australia, March 2010
- Bureau of Meteorology, Long-term temperature record, A short history of Australian observational records
- Bureau of Meteorology Rainfall Districts
- Grant Williamson, Lynda Prior, Matt Jolly, Mark Cochrane, Brett Murphy and David Bowman, Measurement of inter- and intra-annual variability of landscape fire activity at a continental scale: the Australian case, IOP Publishing, 23 February, 2016
- Bureau of Meteorology, Fire Weather Knowledge Centre, Fire weather services
- AEMC, National Bushfire Warnings Taskforce, Australia’s revised arrangements for bushfire advice and alerts, 2009/2010 Fire Season, 7 October 2009
- Bushfire and Natural Hazard Cooperative Research Centre, Research gives insight into community bushfire response, 1 June 2018
- Andrew Dowdy, Climatological Variability of Fire Weather in Australia, Bureau of Meteorology, February 2018