Saturday, March 17, 2012

24 November 2010 – Weather-related mortality in England and Wales

"Is not their climate raw, foggy, and cold / On whom, as in disdain, the Sun looks pale?"
(The High Constable of France, Shakespeare, Henry V)


There's a new paper out in Climatic Change that might be of some relevance to those of us charged with figuring out what the world of the future might look like.

In assessing the potential impact of climate change on human health, the authors of the report of the IPCC's Working Group II (which produced the 2007 Report entitled "Climate Change 2007: Impacts, Adaptation, and Vulnerability") stated, with "high confidence", that climate change would "bring some benefits to health, including fewer deaths from cold, although it is expected that these will be outweighed by the negative effects of rising temperatures worldwide, especially in developing countries", and also that "even high-income countries are not well prepared to cope with extreme weather events" (IPCC 2007, Report of WGII, Chapter 8, 393).  The authors went on to state - this time with "very high confidence" - that "Economic development is an important component of adaptation, but on its own will not insulate the world’s population from disease and injury due to climate change" (idem.) 

Various factors, the IPCC argues, can alter the human impact of changes in weather patterns, ranging from poverty and endemic disease to wealth and urbanization (e.g., "Heatwaves," the report notes on p. 406, "are exacerbated by the urban heat-island effect" - itself a puzzling conclusion given that the report of Working Group I, "The Physical Science Basis", argues that the urban heat island effect has had "a negligible influence" on surface temperatures over the past century), but overall, the IPCC concludes (a) that climate change will result in an increase in deaths from more heat and a numerically smaller decrease in deaths from less cold, and (b) that even developed nations are unlikely to be able to cope with the impact of adverse weather on human mortality.

The numbers involved, of course, have never been large; even using the most generous modelled projections for deaths due to climate change, the projected increase in mortality for the US from climate change in the year 2050 will be on the order of 700 deaths per annum, roughly the same number of Americans as die every year from the accidental discharge of firearms, and fewer than one-quarter the number of Americans who die of malnutrition or from accidental falls.  Recent research, however - based on historical and medical data rather than model projections - suggests that the projections offered by the IPCC may not only be wrong, but entirely backwards.

In their new paper in Climatic Change, a trio of researchers (Christidis, Donaldson, and Stott) looked at death statistics for England and Wales during the period 1976-2005, the most recent period of warming (the average localized warming over the period in question, they report, was 0.47 degrees C per decade), with a view to isolating the cause of death per million and determining whether there was a correlation with temperature trends.  Their results show that:

...during the hottest portion of the year, warming led to increases in death rates, while during the coldest portion of the year it lead to decreases in death rates. More specifically, the three scientists report that if no adaptation had taken place, there would have been 1.6 additional deaths per million people per year due to warming in the hottest part of the year over the period 1976-2005, but there would have been 47 fewer deaths per million people per year due to warming in the coldest part of the year, for a lives-saved to life-lost ratio of 29.4, which represents a huge net benefit of the warming experienced in England and Wales over the three-decade period of warming. And when adaptation was included in the analysis, as was the case in the data they analyzed, they found there were only 0.7 death [sic] per million people per year due to warming in the hottest part of the year, but a decrease of fully 85 deaths per million people per year due to warming in the coldest part of the year, for a phenomenal lives-saved to life-lost ratio of 121.4.

These empirical findings overturn both of the IPCC's principal contentions regarding the likely impact of climate change on weather-related mortality.  First, they demonstrate, contrary to the IPCC's contentions, that in regions with more extreme climates (the annual range of daily average temperature in the UK is rather larger than it is in, say, sub-Saharan Africa), the positive impact of warming on cold-related deaths vastly outweighs the negative impact of warming on heat-related deaths, showing that, ceteris paribus, the net impact of warming on weather-related deaths is overwhelmingly positive.  Second, contrary to the IPCC's contentions, the authors demonstrate that even in under-developed countries, warming will not cause more heat-related deaths than will be saved through reductions in cold-related deaths, as even with adjustments for adaptation removed, the ratio remains 29.4:1 in favour of lives saved from reductions in cold.  Finally, they demonstrate, contrary to the IPCC's contentions, that developed countries are in fact more than capable of mitigating the impacts of temperature change on human mortality through adaptation; according to their data, adaptation measures improved the lives-saved to lives-lost ratio from 29.1:1 to 121.4:1, a factor of four.  This probably comes as no surprise to those of us presently engaged in hauling out last year's boots and mittens, putting snow tires on our cars, and laying in stockpiles of road salt (or environmentally friendly Ice-Melter™, a mixture of urea and calcium chloride the principal environmental benefit of which is that it is marginally less likely to kill your grass).

The authors also note that, contrary to what conventional wisdom might suggest, "Benefits from rising winter temperatures would be more pronounced in countries with milder winters, where cold-related mortality is higher." (Christidis et al., 551)  In other words, the positive impact of milder winters on human mortality is likely to be more noticeable in warmer, i.e., developing, countries.  Counterintuitive, perhaps, but science is all about going where the data take you.

Taken altogether, the results reported in the paper place a question mark after some of the IPCC's "high confidence" predictions about how climate change is likely to impact human mortality, which suggests in turn that those of us who are asked to unscrew the inscrutable and predict the future need to be a little cautious - not to say sceptical - when citing what seem to be authoritative sources.  The authors put it this way: would be easy to compare the recent decrease in cold-related mortality with the increase in temperature and make the seemingly logical assumption that fewer people have died because of milder winters. Our work, however, shows that this is not the case. We find that adaptation of the population to colder temperatures can explain much of the observed change. We also show that if adaptation to cold weather had remained unchanged, then the anthropogenic warming would have produced a detectable decrease in winter mortality. Nevertheless, in the real world the effect of adaptation appears to be more important than the impact of the anthropogenic warming. (Christidis, et al., 550)

And it's the "real world" that matters.

You might be wondering why this topic suggested itself this morning.  Well, the Christidis, et al., paper was one reason; another was the fact that it focuses on England and Wales, a region that exemplifies the "North Sea" microclimate, which is only marginally more hospitable than our own, and it's definitely been a little "raw, foggy and cold" these past few days.  A third reason is that another part of that microclimate region - Jutland - got about 20 cm of snow yesterday, a singularly unusual occurrence, particularly for November.  Unlike us, the Danes don't tend have snow shovels, let along snow-blowers.  In short, 'tis the season when excess human mortality due to cold becomes, once again, a matter of some personal interest.




(1) Christidis, N., Donaldson, G.C. and Stott, P.A. "Causes for the recent changes in cold- and heat-related mortality in England and Wales", Climatic Change 102 (2010), 539-553.
(2) For Danish weather, I recommend the websites of Jyllands Posten ( or Ekstra Bladet (  Be cautious when clicking through on the latter one; the Danes seem to have a much more flexible definition of "NSFW" than we do.