A new research paper providing scientific underpinning for links between global warming and the increasing regularity of hurricanes is a challenge to policy-makers to think afresh, report Dave Frame & Dáithí Stone.
The devastation wrought by hurricane Katrina and the subsequent development of hurricane Rita have invited speculation about the role of greenhouse warming in the frequency and intensity of hurricanes. Over the last decade many scientific studies have attempted to examine the links between the two. The evidence has generally been mixed: many studies have asserted increases in intensity or frequency of hurricanes, but none have convinced the research community of a definite link.
However, a recent, eerily timed paper by Kerry Emanuel from MIT shows compelling evidence that there has been an increase in the destructiveness of tropical cyclones over the last thirty years. While the paper has its critics, it has also captured the research communitys attention as a potentially important line of inquiry.
In the paper, Emanuel develops an index of hurricane destructiveness based on the dissipation of energy (over the lifetime of the hurricane). This index is plausibly linked with observed climate signals, including multi-decadal cycles in the north Atlantic and north Pacific, and global warming. The results suggest that continued warming may lead to an increase in hurricane destructiveness, which, combined with the accumulation of coastal risk that has resulted from urbanisation along vulnerable coasts, may lead to a substantial increase in socio-economic damage over the coming century.
But the establishment of a highly credible link between climate change and hurricane destructiveness, while extremely important, is only part of the story. To know what we ought to do in response to this growing threat, we need to get some sort of quantification of the increase in risk. Its one thing to say the risk is rising or decreasing; its another to say how much.
openDemocracy writers examine the fallout of Hurricane Katrina:
Mariano Aguirre, The Hurricane and the Empire
Ian Christie, When the levee breaks
Godfrey Hodgson, After Katrina, a government adrift
Michael Thieren, Katrinas triple failure: technical, ethical, political
If you find this material valuable please consider supporting openDemocracy by sending us a donation so that we can continue our work and keep it free for all
Measuring the risks
New techniques available to researchers which exploit the probabilistic turn in climate science can, in some instances, allow us to examine the degree to which an event becomes more or less common in a world with increased greenhouse gases. Peter Stott and co-workers applied this technique to the European summer 2003 heatwave and concluded that the risk of such a heatwave was strongly amplified in our 2003 (378 parts per million [ppm] CO2) world, when compared to a non-industrial world (278ppm). This is a softer sort of causal link than the sort of deterministic cause-effect chain that people customarily associate with science, but the complexity which characterises the earth system make such determinism elusive.
As many others in the climate research community have noted it is and will remain impossible to say this particular flood or drought was caused by global warming (just as it is impossible to say that this particular case of cancer was caused by smoking) but we may be able to discuss changes in the likelihood or risk of such events. Attempting such a fractional change in risk or fractional attribution study for hurricane Katrina is a theoretically inviting idea, but given the difficulties the current generation of climate models have in simulating processes smaller than several hundred kilometres, one that may have to wait for modelling to catch up (at least a little more) with the real world.
Even once the risks of such events are quantified, decision-makers have to plan around them. This introduces another set of complexities: climate events and political decisions are taken at a wide variety of scales. The most literal sense of global greenhouse warming is that the Earths average surface temperature is expected to rise by a few degrees over the coming century because of increasing atmospheric abundances of greenhouse gases. But climate impacts are felt on a range of scales beneath this: El Niño affects much of the Pacific region; the heatwave that beset Europe in 2003 was pretty much continent-wide; and hurricane Katrina affected one large city and the surrounding regions.
Political decisions occur on a similarly wide range of scales. Decisions to try to prevent disaster, ameliorate it or insure against its presence are made all the time, at all scales. The United Nations is involved in trying to reduce global emissions of greenhouse gases; the levees of Louisiana are a way of trying to alleviate climate threats. So is the consumers purchase of an air-conditioner. Where risks remain constant we can continue as before; when risks change, we need to try to quantify the magnitude of those changes and adapt accordingly by figuring out what the new odds are. We may need a more sophisticated air-conditioner, or more levees. But these adaptations only deal with the symptoms of climate change: on the prevention side we may need to develop agreements and incentives to emit less carbon into the atmosphere.
Planning for disaster
It is important to recognise that these decisions take place on a very wide range of scales, and that focusing exclusively on any one scale leaves most of the problem unaddressed. Its also important to note that these decisions change from private-good problems at localised scales (skimping on your air-conditioning affects only you) to much more public-good-type problems at larger scales (such as levees). The provision of public goods usually falls to local or national governments because markets alone usually underprovide them.
Polities face resource pressures, and if something doesnt happen very often, or isnt fresh in the memory, then there is a temptation to cut corners and to skimp on prevention, preferring to insure a cure. This is exacerbated when risks change over time, because the communitys precautions no longer fit the real risks they face. A hundred-year flood in a pre-industrial climate may become a forty-year flood in a 21st-century climate, and though modelling can help identify some of these risks, most communities will find out about changes in climate risks the hard way.
Climate scientists may be a little way off being able to quantify these sorts of changes in risk for all but the largest-scale events, but the identification of plausible physical mechanisms linking the statistical properties of extreme events to long-term fluctuations and trends in climate is an important and necessary step along the way. The next steps, and the ones that are most crucial to our long-term risk exposure, are to deepen our understanding of the systems in which such extreme events are embedded, and to communicate the right sort of information to people in language they can base decisions on. That way, even if scientists cant always predict the precise details of heatwaves and hurricanes, at least they can provide people with the right odds.
As hurricane Rita bore down on the US Gulf coast, responses ranged from individuals driving to higher ground, to organised evacuation of vulnerable areas, to deployment of emergency services. In the event, the scale of the impact was smaller than feared, and in any case such responses helped on this occasion to avoid the major loss of life and deleterious economic impacts associated with hurricane Katrina.
But even if societys reactions to major climatic events go smoothly and we manage to avoid the horrors witnessed in and around New Orleans after Katrina, such reactive endeavours can only ever be part of the story: as increases in atmospheric concentrations of greenhouse gases cause various aspects of climate to change rapidly over the coming decades, we should expect to be surprised by climatic phenomena on a more regular basis. The only certainty is that we dont yet know exactly how.