Extreme Weather in Australia

The sharp-eyed will have noticed a comment from Roger Pielke Jnr directing me and any other interested readers to an article by two Australians, Ryan Crompton and John McAneney, who had actually done what I had asked someone to do: estimate the losses caused by natural disasters in Australia.

The article he linked to is behind a paywall, but there is an earlier one in the Australian Journal of Emergency Management in 2008 which you can read. Crompton and McAneney normalised costs in much the same way as Roger Pielke Jnr did for the US. Storms are the great villain here, as they are in the US, but we don’t have tornadoes. Hail seems to be our great agent of destruction, rivalling thunderstorms.

While I wrote of the damage caused by floods, droughts and fires, the latter two disasters don’t make it into the big Australian insured losses list. I guess that one reason is that while fires are a great subject for television, they seem mostly to burn up scrub and bush. Droughts have serious impact on rural and export incomes, but don’t appear as major insurance events, unlike storms.

We don’t have tornadoes, but we did have Cyclone Tracy in Darwin, and the authors make the point that since it occurred Australian building regulations have been tightened to reduce the  likelihood of damage should another such storm occur. We do learn, even if it seems to take time. The earthquake in Newcastle also caused authorities there to look hard at building regulations, though the authors say that no one much takes seismic activity seriously in Australia. I can remember a tremor in Canberra in the late 1940s, and my son felt the Newcastle quake when he and I were bushwalking north of that city — I somehow missed it.

The Australian Book of Disasters offers twelve that are weather-related, like Cyclone Tracy (1974), Cyclone Mahina of 1899 that devastated Bathurst Bay in north Queensland and killed more than 400 people, and the great Melbourne storm of 1934 that killed 35 people. Broome in WA probably has the record for storms, with cyclones in 1875 (59 dead), 1887 (140), 1894 (40+), April 1908 (50+), December 1908 (50+), 1910 (0), 1912 (150+) and 1935 (unknown numbers). Who so many deaths? Broome was the home of the pearling fleet, and the town is exposed.

Pearl buttons are no longer important, and the much smaller pearling fleet now has radio and radar, but the Broome cyclones have not been as severe since the war. The big three in the past decade or so have been the Sydney hailstorm of 1999, the Victorian bushfires of 2009 (173 dead), and the Brisbane floods of 2011 (more than 40 dead). Each of these was costly in insurance terms, too, and Crompton and McAneney have updated their 2008 study to include these later disasters, so now there have now been four peak periods the mid 1960s.

It is hard to see any sign of a link between these disasters and ‘climate change’, and of course the record in the 19th and earlier 20th centuries doesn’t really assist the supposition that there is one. As I wrote in my earlier post, the IPCC itself says that there is no significant connection.

Crompton and McAneney say, in their 2008 paper, that ‘societal factors — dwelling numbers and values — are the predominant reasons for the  increasing cost of insured losses due to natural disasters in Australia’, and go on to give the take-home message:

‘Australia can, if it chooses, control where and how people live and build.It is now relatively easy to identify homes vulnerable to threats such as tropical cyclone, hailstorm, bushfire, riverine flood, coastal flooding, etc. at least to an accuracy sufficient to underpin prudent policy decisions.’  [If we did this we would get] ‘immediate improvements in community resilience to both current and future climates. The choice is ours.’

I agree entirely. I am usually careful in making sure that I have read what I ought to read, but had it not been for Roger Pielke Jnr’s comment (for which I am most grateful) I would not have known about this most important paper. To the best of my knowledge, it was not mentioned in the mainstream media, or if it was, then briefly indeed.

Join the discussion 5 Comments

  • John McAneney says:

    The Crompton and McAneney (2008) study has been recently
    updated with the methodology and results now available on the Insurance Council
    of Australia’s website:


    We have
    also published (Crompton et al. 2010) normalised losses for historical building
    damage and fatalities due to Australian bushfires extending back to 1926 when
    the first major losses occurred in the 20th century. This paper, along with some 30 others in different jurisdictions
    and for different perils, fails to find any trend after normalising for the
    losses for changes in population, wealth and inflation.

    This being
    the case we then examined
    how long it might be until a climate change signal first becomes detectable
    in loss data. We (Crompton, Pielke and McAneney 2011) did this for US
    hurricane losses where we have the benefit of both high-resolution modelling of
    projected hurricane activity, as well as over a century of economic losses from
    landfalling hurricanes. The so-called emergence time scale turns out to be a
    very long time: 260 years for an 18-model ensemble and a range of 120- 550
    years when individual global climate change models are used to set the boundary
    conditions. The implications are pretty clear: disaster losses due to natural
    perils will continue to increase as population and wealth grow in vulnerable
    areas, but climate change will play little part in aggravating the problem for
    a long time to come.

    None of
    this means that climate change isn’t a problem but if we genuinely want to reduce
    this nation’s exposure to natural disasters, then adapting to the current
    climate would be place to a good start. And for location-specific perils such
    as bushfires and floods, improved and risk-informed land use planning practices
    are essential. By way of example, we found that 25% of the homes destroyed in the
    2009 Victorian bushfires in Marysville and Kinglake were located within 1 m of
    the bushland; 60% were within 90 m. Overall the probability of home destruction
    in those first 10 m was over 90% (Chen and McAneney, 2010). Given the
    conditions on the day, the outcome is unsurprising.

    Happy to
    provide copies of papers if you email me at riskfrontiers@mq.edu.au.


    K. and J. McAneney, 2010. Chen, K. and J. McAneney, 2010. Bushfire Penetration into Urban Areas in Australia: A Spatial Analysis.
    Invited submission for the Royal Commission in to the 2009 Victorian bushfires.
    Risk Frontiers.

    Crompton, R.P. and K.J. McAneney. 2008. Normalised
    Australian insured losses from meteorological hazards: 1967 -2006. Environ. Science & Policy 11:

    R. P., K. J. McAneney, K. Chen, R. A. Pielke Jr., and K. Haynes, 2010. Influence
    of Location, Population and Climate on Building Damage and Fatalities due to
    Australian Bushfire: 1925-2009. Weather, Climate and Society, Vol. 2: pp. 300-310, doi:

    Crompton, R. P., R. A.
    Pielke Jr.
    and K. J. McAneney,
    2011. Emergence time
    scales for detection of anthropogenic climate change in US tropical cyclone
    loss data. Environmental Research
    Letters, 6, 014003, doi:

    Crompton, R.P. 2011. Normalising
    the Insurance Council of Australia Natural Disaster Event List: 1967–2011. Risk Frontiers.

  • Greenparrot says:

    Hmm. My household insurance jumped 50% this year. I also have a modest holiday house and when I looked at reinstating a policy that had lapsed for 6 months. This premium was near double. So,either your findings are wrong or the insurance companies are profiteering.

    • donaitkin says:

      It’s hard to know from your comment just what has happened in your case. Insurance companies operate rather like bookies: they need to estimate likely outcomes accurately and ensure that at the end of the day they are not out of pocket.And they do need to make a refit as well as pay claims. It makes sense for them to cry doom about climate, first because it encourages people to take out insurance, and second because they already have governments and organised science saying so anyway.

  • […] written about this subject before, and mentioned approvingly both Roger Pielke Jnr of the USA and John McAneney of Australia, who […]

Leave a Reply