Worst case scenarios 4: Droughts and floods
Another in my series of worst case scenarios. A bit more locally oriented this time.
There have been important changes in the way we think about droughts, especially in relation to public policy. In the past, droughts were viewed as natural disasters, requiring a public policy response, which was typically triggered when an area was declared drought-affected. This happened so often that some parts of Queensland were drought-declared one year in three. But a climate event that happens as regularly as this, can’t be regarded as an unpredictable disasters. It is simply part of the natural cycle, which farmers, water supply enterprises and others must take into account in their normal management. This is the central feature of the National Drought Strategy adopted in the 1990s, which faced its first real test in the recent large-scale drought. Although the drought was one of the worst in Australian history, most observers agree that it was handled more successfully than previous droughts that were, in purely climatic terms, less severe.
But recognition of droughts as part of the natural cycle does not provide a complete answer. No matter how good your management, it is not economically feasible to prepare against the worst-case droughts. A common criterion used in water supply is to prepare plans based on a 1 in 100-year drought, that is, one that would be expected to occur once in every hundred years on average. Implicitly, though, this means that if the drought is worse than the 1 in 100 years, normal planning will fail, and some emergency response will be needed.
A further twist to the worst-case scenario is added by the process of climate change. Estimates of the 1 in 100 year drought level are based on historical observations of the climate. But climate change will make some areas drier and others wetter. In areas where average rainfall declines, severe droughts will become more frequent.
In economic terms, droughts are normally a bigger problem for Australia than floods. But the worst-case scenario is considerably worse for flood than for drought. During a 1 in 10,000-year flood event, the Wivenhoe Dam could fail catastrophically, putting between 50 000 and 250 000 lives at risk. This possibility might seem so remote as to be negligible, but for a dam with a life of 50 years, the chance of a 1 in 10000 year event occurring at some point becomes 1 in 200, enough to be a threat worth considering
Fortunately, SEQWater is currently engaged in an upgrade which should protect against this remote possibility by raising the height of the spillway and dam wall to a level that should protect against a 1 in 100 000 year event.
It might be asked how it’s possible to make any sort of rational decision about whether to spend money on risks like this. The easy answer is that anything that reduces risk is worth doing, but since it is never possible to eliminate risk completely, it’s not possible to dodge the necessity for hard choices. The best way is to work out the average number of lives saved and compared the amount that would be needed to save the same number of lives in some other way, for example, through improvements in medical care. A typical estimate is that a treatment that is just on the margin of acceptability will cost around 100 000 for each additional year of life gained, which translates to about $5 million for each life saved from a disaster such as a flood.