01 February 2011

Cyclone Yasi in Context

As the incredibly powerful Cyclone Yasi bears down on Queensland, I thought that it might be worthwhile to put the storm into longer term context.  The Crompton and McAneney (2008) paper that I have cited several times of late on normalized Australian insured losses from extreme weather events includes this noteworthy comment:
The average annual weather-related normalised damage over the 40-year period is AUD$820 million with a standar deviation of AUD$960 million. The recent past has been relatively benign in terms of loss activity, with annual damage over the most recent 5 years averaging AUD$420 million, close to half the average annual loss over the entire period of the Disaster List.
In the absence of systematic evaluations of data, it would be all too easy to consider recent experience (whether that be the past week or the past decades) to somehow be "normal" and thus the basis for expectations of the near future. Such expectations, when out of sync with a longer term perspective, can easily lead to misplaced judgments of risk and contribute to poor decision making.

Such a systematic evaluation of the long-term tropical cyclone landfall record in eastern Australia was published last summer in Climate Dynamics by Jeffrey Callaghan and Scott Power (2010).  Callaghan and Power find a long-term trend of much fewer landfalls of intense cyclones (i.e., Category 3, 4, and 5) in the region.  They write:
The linear trend in the number of severe TCs making land-fall over eastern Australia declined from about 0.45 TCs/year in the early 1870s to about 0.17 TCs/year in recent times—a 62% decline.
The figure at the top of this post comes from their paper and comes with the following caption:
Fig. 1 The number of severe tropical cyclone (TC) land-falls in each TC season from 1872/1873 to 2009/2010 inclusive. The corresponding linear trend of -0.0021 TCs/year is also shown. This represents a decline of approximately 60% over the full period.
They find evidence for a relationship between intense cyclone landfall activity and the ENSO cycle, reflecting the natural variability of the system. They speculate about a connection between the significant decrease in landfalls and global warming, but find little convincing evidence of such a link.  They explain:
If natural variability dominates the recent decline, and if this natural variability exhibits low levels of predictability as suggested by the study of Power et al. (2006), then the frequency of damaging land-falls could increase abruptly to much higher values during the coming decade.

On the other hand, some unknown fraction of the recent decline in the SOI might instead reflect a weakening of the Walker Circulation in response to global warming (e.g. Power and Smith 2007). This means that the component of the decline in land-falls linked to the decline in the SOI might also be partially driven by global warming. . .

The extent to which global warming might be contributing to or offsetting the decline in land-falling TCs over eastern Australia is unknown.
Callaghan and Power note prudently that the unresolved issues of global warming need not stand in the way of making effective judgments of risk -- landfall rates have been higher in the past and that observers should not be misled by recent experience:
Given that natural variability is at least partially responsible for the observed interdecadal decline documented here, and that this natural variability might be dominated by unpredictable changes (Power et al. 2006), it would be imprudent to suppose that the very low landfalling rates observed in recent decades will necessarily continue. Planning should therefore reflect the possibility of a rapid return to much higher landfall rates.
Best wishes to all those in Queensland affected by Cylone Yasi for a speedy recovery, and what inevitably will be storms to follow.