Climate Change And Bushfires: An Assessment
From a historical perspective Australia’s recent bushfires are not uncommon and are certainly not unprecedented.
Climate change has yet again been blamed for another natural disaster, this time the recent bushfires in NSW. But much more important is the role of poor land-use planning decisions that are increasing our nation’s vulnerability to fire, and other natural perils. We examine these issues in the light of Australia’s history of building losses to bushfire over the last century.
Bushfire losses in Australia
Prime Minister Tony Abbott was correct to assert that bushfires are a fact of life in Australia. Conflagrations may occur whenever favourable combinations of fuel, weather and ignition sources exist. When communities are in the way, large losses are always possible.
Back in 1947, James Foley, in his treatise on bushfire risk in Australia, gave descriptions of historic fires in NSW. These were laced with comments such as “the worst in the memory of the oldest residents”; “most disastrous bush fire known”; “most serious fires for years”; “one of the most disastrous fires known”, and so on.
Each fire it seems is always worse than the last.
One way of measuring the severity of a fire is to look at the loss of property. While raw losses from bushfires, as for most other natural disasters, show an increasing trend towards higher costs, the graph below shows what this history looks like after losses have been “normalised” to take into account that we now have more homes in harm’s way.
There’s no trend in the graph. Bushfire losses can therefore be explained by the increasing exposure of dwellings to fire-prone bushlands. No other influences need be invoked. So even if climate change had played some small role in modulating recent bushfires, and we cannot rule this out, any such effects on risk to property are clearly swamped by the changes in exposure.
The result is unsurprising given that it has been a consistent conclusion from many other studies in different countries and across many different hazards, in fact some 30-odd different peer-reviewed studies to date. And the IPCC (2012) underscored this conclusion.
What about New South Wales?
So much for the national picture. What about NSW: are losses in this state this early in the season unusual?
We answer this question again using Risk Frontiers’ PerilAUS database, which lists natural peril events causing material property damage or loss of life. PerilAUS suggests that since 1926 early season fires (August-September-October) have occurred in about 25% of years. They do not always herald losses later in the season.
And narrowing our focus even further to the Blue Mountains, research shows that destructive bushfires have affected all townships in the Blue Mountains, from Blackheath in the west, to Emu Heights in the east. The most common months for damaging bushfires have been November and December. But October fires are hardly out of the ordinary.
One fire occurred on October 7, 1926, 10 days earlier than the date of the most recent destruction (October 17, 2013). It is likely that this fire (or fires) had been burning even earlier but this was the date on which most of the damage occurred.
So from a historical perspective our recent fires are not at all uncommon and are certainly not unprecedented.
Increasing the risk
So if it’s not earlier fires or more frequent fires, what is causing changes in bushfire losses? The answer lies in exposure.