25 January 2011

Updated -- Normalized Disaster Losses in Australia

[UPDATE: The data in the graph above was updated on 31 January 2011 and presented here.]

Ryan Crompton at Macquarie University has updated the Crompton and McAneney (2008) analysis of disaster losses in Australia to account for the estimated losses from the 2011 Brisbane floods, as well as each of the years following their analysis to the present.  You can see the results in the figure above.  The figure shows meteorological disasters (bushfire, flood, hailstorm, thunderstorm, tropical cyclone) for the period 1966 to present using an apples-to-apples comparison of losses across hazards.

Here is a note that Ryan sent me which explains the figure:
The figure shows the Crompton and McAneney (2008) Figure 2(b) adjusted to include the seasons 2006/07 - 2009/10 and the loss from the recent Queensland floods (the only loss included in the current 2010/11 season). The losses in the four complete seasons added to the time series have been normalised to 2005/06 values as per Crompton and McAneney (2008) so that losses have been reduced from their original values for consistency with the other data. The Insurance Council of Australia (ICA) has released updated figures (PDF) on the insurable losses in Queensland as a result of flooding, with the current estimate at $1.2 billion. With the final loss from this event not yet known and likely to increase, the original loss of $1.2 billion has been included in the above figure rather than the equivalent 2005/06 normalised value (which would be a bit lower).
You can find Crompton and McAneneny (2008) here, and if you do not have access to that journal you can read a very nice summary of that work here in PDF.

Writing with Kevin Roche, John McAneneny discussed lessons of the recent floods in a op-ed:
The city of Brisbane, like many other towns in Queensland, is built on a floodplain.

Compared with the current disaster, there have been even bigger floods in the past: in 1841 and 1893 when flood waters topped 8.35 meters, some 3.9 m above the latest peak.

After the 1974 floods, the senior engineer with the Hydrometeorology Branch of the Bureau of Meteorology, G. Heatherwick, warned that heavier rainfalls were possible over the Bremer river catchment with higher flooding in Ipswich, independent of the flood mitigation effects of the Wivenhoe Dam.

This is not to deny that climate change is a real concern: few continue to believe that only positive outcomes will arise from the continued heating of the planet.

The latest research, however, just published in the international journal Environmental Research Letters by Ryan Crompton, Roger Pielke Jr. and John McAneney suggests that it may be centuries until we can be confident that climate change is influencing disaster losses.

If we truly wish to reduce the scale of future disasters in Australia, we need risk-informed land planning policies with risks appropriately priced by an active insurance market.

In simple terms, for flood and bushfire, this means an end to unmanaged development of flood plains or within bushlands.

In the Black Saturday fires, studies by Risk Frontiers showed that 25% of the home destruction in the most affected towns of Marysville and Kinglake took place physically within bushlands; 60% occurred within 10 m of bushland boundaries.

At these distances, chances of home survival are low and not surprisingly few resisted the flames.

At the beginning of last century, lives and property were lost as people lived and worked in the bush; today the problem remains as people chose to live in or near the bush for lifestyle reasons. The analogy to floodplains is obvious.

Extreme weather is not new to Australia with some 95% of all property losses in natural disasters since 1900 attributable to flood, hail, bushfire or tropical cyclones.

However, excluding the case of improved construction standards for residential homes in cyclone-prone areas of the country, successive Australian governments, at all levels, have failed to address shortcomings of land policy solutions to mitigate against the potential impacts of natural disasters.

In the emotive days after the Black Saturday fires of 2009, the then Australian Prime Minster, Kevin Rudd, stated that we would rebuild impacted communities "brick by brick."

There was no immediate consideration about reducing risks. Let's hope the official reaction to these floods is more measured.

9 comments:

markbahner said...

"However, excluding the case of improved construction standards for residential homes in cyclone-prone areas of the country, successive Australian governments, at all levels, have failed to address shortcomings of land policy solutions to mitigate against the potential impacts of natural disasters."

Geez. I think I need to be made King of Australia. And not some fake king. A real King king.

Here's what Brisbane should do:

1) Go to Google Earth and identify every scrap of earth downstream of the Wivenhoe dam.

2) Figure out how much of that earth is *below* the riverbank level, and how much is so far above the riverbank level that it could never flood, even if all the water was emptied from the dam.

3) Determine the value of every single acre that's below the riverbank level.

4) Designate the pieces of land that are least valuable as "flood land" and designate the most valuable pieces of land as land to be protected.

5) When the next potential flood comes, breach the banks of the river, or put in a temporary obstruction in the river, to flood the least valuable land.

6) Compensate the owners of the flooded land well beyond the market value of the damage to their property. For example, assume they have farmland with crops worth $500,000. Then give them $5,000,000 for their crops, if their crops are totally lost.

This method avoids flooding of valuable pieces of land and more-than-fully compensates those whose land does get flooded.

This method also completely avoids any need for Brisbane to alter its present development.

David Palmer said...

Roger,

I linked to this post in a discussion with another person and he responded based on a report from the Insurance Council of Australia that the figures on which that updated graph is based do not include mining claims or other large claims, which are likely to blow the $1.2 billion out of the water.

Would you care to comment.

Roger Pielke, Jr. said...

-2-David Palmer

I'll pass on your question to Ryan Crompton, but my suspicion is that the estimated costs of the 2011 flood from ICA are apples to apples with the other data in the series. But if you want to double or triple the costs, it doesn't change the overall conclusion. I'll report back what I hear. Thanks.

Ryan Crompton said...

-2- David Palmer

Thanks for your comment. The ICA report that you refer to is where I got the current estimated loss from and is linked to in Roger’s blog post (see the PDF link in my note to Roger). I plan to update the graph as the loss is updated and I deliberately included the loss as an original value rather than a normalised value because the final value is not yet known. I agree with Roger that the overall conclusion will hold even if the final loss increases considerably, especially when you consider that the final value still needs to be normalised (reduced) to 2005/06 values.

Nick said...

It's a bit of a stretch to cite your recent co-authored study,based solely around one extreme phenomenon, North Atlantic hurricanes, as support for the more general claim that "it may be centuries until we can be confident that climate change is influencing disaster loss."

Ryan Crompton said...

-5- Nick

Not sure if this is to me or Roger but happy to pick up on it. The paper you refer to (Crompton et al. 2011) shows this to be the case for global weather-related natural disaster losses to the extent that they remain correlated to US hurricane losses.

Roger Pielke, Jr. said...

-5-Nick

To follow up Ryan's response, you may want to have a look at this post:

http://rogerpielkejr.blogspot.com/2011/01/signals-of-anthropogenic-climate-change.html

Especially see this comment:

http://rogerpielkejr.blogspot.com/2011/01/signals-of-anthropogenic-climate-change.html?showComment=1294270826673#c1325592703329990992

Ryan Crompton said...

The ICA estimated insured loss for the QLD floods has been updated and has increased from $1.2 to $1.51 billion:

http://www.insurancecouncil.com.au/Portals/24/Media%20Centre/2011%20Media%20Releases/QLD%20figures%20Jan30.pdf

David Palmer said...

Hi Roger and Ryan,

Thank you for your prompt response which also with the explanations given is most helpful.

David

Post a Comment