24 January 2012

Follow Up: 2011 Brisbane Floods

Just over a year ago, Brisbane, Australia experienced its worst flooding since 1974, resulting in billions of dollars in damage. Immediately after the event the focus of attention turned to the management of the Wivenhoe Dam, which was built after the 1974 floods to prevent a repeat.

This past week, as the Queensland flood commission is wrapping up it investigation, The Australian has uncovered evidence that the Wivenhoe managers operated the dam according to an incorrect procedure. Here is a summary from the Sydney Morning Herald:
On Monday and Tuesday The Australian newspaper alleged engineers operating the Wivenhoe Dam used the wrong water-release strategy, breaching its operation manual, in the lead-up to the January 2011 Brisbane flood.

It reported SEQWater engineers, who operate the dam, failed to move to a higher water release strategy early enough, contributing to the Brisbane and Ipswich floods.

The paper used emails between SEQWater and the WaterGrid to back up their claims.

It went on to accuse the commission of overlooking the documents and accepting at face value evidence from engineers who said the manual had been followed correctly.

The commission was in possession of the emails but did not make them publicly available.
The release of the information contained in the emails has prompted a re-opening of the Floods Commission inquiry and a delay in the Queensland state election. Anna Bligh, the Premier of Queensland (pictured at the top of this post with Prime Minister Julia Gillard), is facing an electoral defeat based on polling, prompting calls that the election is being delayed for politically strategic reasons.

Writing very recently in the open-access journal Water, Robin van den Honert and John McAneney, of Macquerie University provide a comprehensive review and assessment of the 2011 floods and their impacts, and which is likely to serve as the definitive study of the event for some time to come. Here is the paper's abstract:
The 2011 Brisbane Floods: Causes, Impacts, Implications

On 13th January 2011 major flooding occurred throughout most of the Brisbane River catchment, most severely in Toowoomba and the Lockyer Creek catchment (where 23 people drowned), the Bremer River catchment and in Brisbane, the state capital of Queensland. Some 56,200 claims have been received by insurers with payouts totalling $2.55 billion. This paper backgrounds weather and climatic factors implicated in the flooding and the historical flood experience of Brisbane. We examine the time history of water releases from the Wivenhoe dam, which have been accused of aggravating damage downstream. The dam was built in response to even worse flooding in 1974 and now serves as Brisbane’s main water supply. In our analysis, the dam operators made sub-optimal decisions by neglecting forecasts of further rainfall and assuming a ‘no rainfall’ scenario. Questions have also been raised about the availability of insurance cover for riverine flood, and the Queensland government’s decision not to insure its infrastructure. These and other questions have led to Federal and State government inquiries. We argue that insurance is a form of risk transfer for the residual risk following risk management efforts and cannot in itself be a solution for poor land-use planning. With this in mind, we discuss the need for risk-related insurance premiums to encourage flood risk mitigating behaviours by all actors, and for transparency in the availability of flood maps. Examples of good flood risk management to arise from this flood are described.
Based on the new reporting from The Australian on the possible errors in flood management and the comprehensive analysis in van den Honert and McAneney (2011), it is clear that bad decision making played a major role in the disaster. The bad decisions were the result of mismanagement, a deeply flawed management architecture, or what seems to be increasingly likely -- both.

3 comments:

  1. The government made a mistake? That can't be. It must be something else like climate change. Government bureaucrats are smarter than everyone else so they can't make mistakes. All kidding aside, what about our own floods on the upper Missouri river last year. From what I understand, it was pretty clear from snow cover and snow depth measurements that there were going to be problems but there seemed to be a conscious decision to not draw down the dams in anticipation of the spring floods. It almost seems like its policy out of Washington to let the rivers flood. So there was no error, just a different set of priorities.

    ReplyDelete
  2. The elephant in the room: climate change. In 2007, the state government released a document warning of future drought due to global warming. Thus, dam operators were pushed to keep as much water on hand as possible. With the reservoir at 100%, when the rains came, there wasn't sufficient reserve available, and flooding ensued.

    The government's policy of scaring the bejesus out of the people to get them on board with climate change regulations contributed to, if not caused this disaster. If there had been no pressure to keep dam levels high, there would have been room to absorb the rains.

    ReplyDelete
  3. While the motivation may differ, the same error of judgment was the exacerbating cause of Japan's nuclear reactor problem recently. In the above story, the cause was an exaggerated concern for AGW/AGCC/"climate disruption". Whereas in Japan, the cause was an exaggerated fiscal concern coupled with overly optimistic procedures.

    Then there was hurricane Katrina and, again, the failure of local government, which was exacerbated by the intransigence of residents, when responding to an imminent threat.

    Actually, the cause in each of the cases was effectively equivalent. In each of the cases a wrong prescription, based on a flawed diagnosis, promoted a distorted perception of reality, leading to an incorrect risk assessment and followthrough.

    I wonder how many levels of oversight would have been sufficient to reduce the risk.

    ReplyDelete