Once again, bushfires are laying waste to Australian homes and communities, this time in Tasmania, with reports of 65 or more properties destroyed in Dunalley alone. Fire emergency services in other states are also on alert as the country bakes in sweltering heat. Fortunately, at the time of writing, there has been no reported loss of life.
Bushfires are the downside of the desire of many Australians to live close to bushlands. Concerns that global warming will aggravate this downside risk are repeatedly amplified through the media and blogosphere:
No weather event can conclusively be linked to climate change, but it is worth noting that the frequency of extreme events like these from now on will have to be accepted as the “new normal”.
The first phrase is certainly true, while the second presumes a degree of clairvoyance not accessible to most scientists.
It still is too early to know how the remainder of the bushfire season will play out, but this doesn’t stop us from putting the history of bushfires in Australia into some perspective.
In what follows, we try and make sense of the real risk as revealed by the cold reality of data unfettered by prejudices and biases. For bushfires, the factors that can be analysed with some objectivity are exposure and vulnerability, essentially the disposition of homes vis-à-vis the bush, and fire history.
Exposure and vulnerability
The single variable that explains most of the vulnerability of a home to bushfire is its distance from the bush. Research conducted by Professor John McAneney and Dr Keping Chen in 2004 and 2010 mapped the location of houses destroyed in some major historical fires and their distance to the nearest bushland interface. They estimated for the Royal Commission into the 2009 “Black Saturday” Victorian fires that historically about 85% or more of the properties lost since the 1967 Hobart bushfires were located within 100 metres of the bush. They found no record of buildings further than 700 metres being lost to bushfires, even though we know embers can travel much further than this.
The likelihood of a property being destroyed in a fire if located within 50 metres of bushland under extreme fire weather conditions is around 50-60%. For homes within 10 metres of bushland in Marysville and Kinglake during the 2009 Victoria Black Saturday fires, it was around 80 to 90%. Some 60% of all homes destroyed in these two towns fell within this distance category (less than 10 metres).
Using nearest distance to bushland boundaries as a measure of exposure, we estimate that around 34,000 addresses in Tasmania lie within 100 metres of the bush, approximately 11% of all addresses in the State. Figure 1 shows the distribution of numbers of addresses by distance to bushland in the entire State, whereas Figure 2 shows our best understanding of the disposition of homes in Dunalley.
Figure 1: Numbers of addresses categorised by distance to bushland boundaries in Tasmania.
Figure 2: Dunalley, Tasmania – Bushland delineation (green lines) and location of building address points.
The exposure metric we defined here is not the only factor to determine the bushfire risk: we also need to consider the likelihood of having the right weather conditions, of ignition to occur, and so on. These factors are not easily quantifiable as some depend on human behaviour, the serendipity of putting down fires in their early stages, and so on.
History does not repeat itself but, as Mark Twain said, it rhymes, and so it is useful to examine the historical record to identify the patterns of bushfire losses in Australia.
The first reference to bushfires in Australia by European explorers was registered in Abel Tasman’s diary on the December 2 1642 during his exploration of Tasmania’s east coast. Fires are natural to the Australian landscape. Its native biota has evolved in and is adapted to fire conditions. Viewed from this perspective, human settlement is the aberrant factor to Australia, not fires.
To build a long-term picture of the historical losses for Australian bushfires, staff at Risk Frontiers have searched reports in early newspapers and other documents including police and coroners’ reports left to us by our forefathers. Since 1994, we have developed a database of past natural hazard events that caused either loss of life or property, or both.
This database is called PerilAUS and is reasonably complete for bushfire losses. Table 1 shows the breakdown of 14,000 home losses by state since 1926, the first major loss event of the 20th century.
The state of Victoria has borne the greatest share of the destruction, in spite of not having had the largest number of events. Tasmania lies in third place in terms of number of properties lost, mainly because of the 1967 Hobart fires, when more than 1550 houses were destroyed.
The numbers in Table 1 reflect raw numbers as recorded at the time. Since then, the population has obviously increased, as has numbers of exposed properties. These changes were analysed by Crompton et al with the results before and after normalisation shown in Figure 3. The normalised figures indicate the likely losses if all fires had impacted upon 2011 societal conditions — an apples-by-apples comparison. The normalised figures show no long-term upward or downward trend. (Interestingly, this is also the case for some 30 similar normalisation studies of economic and insured losses from various extreme weather events and in different countries.)
The main finding of Crompton et al is that the trend of increasing numbers of buildings destroyed (before normalisation) in bushfires in Australia can be explained by an increased population that chooses to live near bushland.
Figure 3: The top graph shows the aggregate number of homes destroyed each year, while the bottom graph has been adjusted to account for the increase in the number of buildings since the original event.
The stars in Figure 4 show the locations of historical fires in Tasmania that caused property losses. These have tended to occur more often around the Hobart region, where most people live and more vulnerable properties are located.
In conclusion, we show that from an historical perspective there’s nothing new about bushfires in Tasmania, nor the potential for large losses. It was only a matter of time until the conjunction of bushfires and home destruction occurred.
We do not yet have a complete picture of the events that have unfolded in Tasmania, but will be keeping a close watch.