Climate and Societal Factors in Future Tropical Cyclone Damages in the ABI Reports

April 24th, 2006

Posted by: Roger Pielke, Jr.

Last year the Association of British Insurers released several reports on damages from extreme weather events and climate change. Since then, I’ve seen the reports cited as evidence of (a) a climate change signal has been seen in the disaster record, and (b) the importance of greenhouse gas reductions as a tool to modulate future disaster losses.

Unfortunately for those citing the ABI reports in these ways, the actual content of the reports supports neither of these claims. A close look at the ABI report shows that it actually supports the arguments that we have made on the relationship of extreme weather events and damage trends. And there is at least one significant error in the report, which serves as a reminder why as valuable as such “grey literature” can be, it is always important to put complex research through a process of rigorous peer reviews. Here are all the gory details (and I do mean all!):

The ABI overview report (PDF) uses catastrophe models (provided by a firm called AIR headquartered in Boston) to project future increases in damage costs related to climate change and extreme events. The report observes:

This study is one of the first to use insurance catastrophe models to examine the potential impacts of climate change on extreme storms. It focuses on one of the most costly aspects of today’s weather – hurricanes, typhoons, and windstorms, because of their potential to cause substantial damage to property and infrastructure.

The report concludes of tropical cyclones in the U.S. and Japan, and European windstorms:

Under these climate change scenarios, total average annual damages from these three major storm types could increase by up to $10.5 bn above a baseline of $16.5 bn today, representing a 65% increase.

My analysis below focuses on the two largest contributors to these costs, tropical cyclones in the U.S. and Japan. First, it needs to be observed that the report does have an enormous oversight:

These loss estimates do not include likely increases in society’s exposure to extreme storms, due to growing, wealthier populations, and increasing assets at risk. For example, if Hurricane Andrew had hit Florida in 2002 rather than 1992, the losses would have been double, due to increased coastal development and rising asset values. Adaptive measures to limit vulnerability could prevent costs escalating.

Why does this matter? Consider Japan, which the report estimates will see its population reduced by half over the next 100 years. By not factoring in this decrease in exposure, the report overestimates the future impacts of climate change by a factor of two. In other words, if today’s population of Japan was half its present value, its current baseline of average tropical cyclone damages would be proportionately lower. It is incomplete, at best, to project future damages by holding society constant and simply changing the climate. Society changes in big ways, and these also must be considered in any projection. The failure to consider societal change sets the stage for misunderstanding or misrepresentation of the ABI reports.

On to the data:

From Table 6.4 we find a baseline of total losses for the US and Japan related to tropical cyclones of 13.5 billion (9.5 and 4.0, respectively). In the ABI Technical Report (PDF, Table 3.7 on p. 38) we see that the total losses increase by the following amounts for a doubling of CO2 by the 2080s:

Wind speed increase US Japan
4% 4.4 1.6

6% 6.8 2.5

9% 11.3 4.1

The ABI also provides data that would allow for a sensitivity analysis of the influence of changing per capita wealth and populations on future damages. Here is what is reported in ABI Technical Report tables 3.8 and 3.9 for population and wealth out to 2100 (2004 = 100):

US Japan

Wealth 262 280

Population 267 50

Since the climate changes impacts are projected to 2085, we need to adjust this table to 2085 through interpolation (2004 = 100):

US Japan

Wealth 236 259

Population 231 57

So this gives a combined change in population and wealth for the US and Japan, 2004-2085, under the assumptions of the ABI as follows:

US = 2.36 * 2.31 = 5.45
Japan = 2.59 * .57 = 1.48

So using the framework we had presented in an earlier post to examine the sensitivity of losses to future climate and societal changes:

Baseline US = 9.5
Increase related to changing climate, society constant (4%, 6%, and 9%) = 4.4, 6.8, 11.3
Percentage increase over baseline = 40%, 63%, 103%
Increasing damage as a function of societal changes (climate constant) = 51.8
Percent increase of societal changes over baseline = 545%

Sensitivity ratio, societal factors to climate factors, US case:

13.6 to 1
8.7 to 1
5.3 to 1

Baseline Japan = 4.0
Increase related to changing climate, society constant (4%, 6%, and 9%) = 1.6, 2.5, 4.1
Percentage increase over baseline = 46%, 72%, 119%
Adjusted for population decrease = 26%, 41%, 68%
Increasing damage as a function of growing wealth (climate constant) = 10.4
Percent increase of societal changes over baseline = 259%

Sensitivity ratio, societal factors to climate factors, Japan case:

10 to 1
6.3 to 1
3.8 to 1

In Table 3.5 in the Technical Report the ABI estimates that climate mitigation could reduce from about 20% to 80% of the projected increases in damages related to climate change, or from $1.2B (0.2 * (4.4+1.6)) to $12.3B (0.8 * (11.3+4.1). Setting aside the inter-related effects of societal and climate change as well as the baseline value (which consequently makes this calculation very conservative), by contrast efforts to reduce vulnerability could reduce future damages by as much as $62.2 billion (51.8 + 10.4). In other words reducing vulnerability is between 5 and 52 times more effective than mitigation. This is the same conclusion that we reached in our published work on this subject.

Unfortunately, the information found in the ABI report is sometimes misrepresented. For instance, in an exchange with me in Science, Evan Mills makes the following claim (here in PDF) about the ABI report:

As an indication of the potential value of emissions reductions, the Association of British Insurers, in collaboration with U.S. catastrophe modelers, estimated that U.S. hurricane or Japanese typhoon losses would vary by a factor of five for scenarios of 40% and 116% increase in pre-industrial atmospheric CO2 concentrations.

Perhaps this is just inartful writing, but this claim is decidedly not what the ABI report says, as can be clearly seen from the tables and analysis above. When assumptions and qualifications are dropped, meanings can change dramatically, sometimes 180 degrees as happened in this instance. Less charitably, Mills claim about what the ABI report says is a complete misrepresentation. Elsewhere I have written about the “laundering of grey literature,” and this instance is one worth watching with respect to the next IPCC report.

Contrary to some citations of the ABI reports on climate change, a close look indicates that the reports actually underscore the points we have been making about the relationship of tropical cyclones and climate change. It is worth noting that the analysis in this post takes the ABI assumptions exactly as they are given in the ABI reports. This exercise indicates that only way to conclude that reducing emissions is an effective tool of disaster management is to rig the analysis to guarantee this result by doing two things: (1) freeze society and only allow climate to change, and (2) to disallow vulnerability reduction as a policy response and allow only greenhouse gas reductions. This is of course not how the real world works. The bottom line is that it would be a mistake to cite the ABI reports as in any way contradictory to the work we have done. In fact, I’m going to start citing them in support of our work!

8 Responses to “Climate and Societal Factors in Future Tropical Cyclone Damages in the ABI Reports”

  1. Benny Peiser Says:


    I noticed an interesting media report today about the Indian Meteorological Agency’s position on climate change and tropical cyclones:

    “The clamour over climate change the world over notwithstanding, the country’s weather agency believes that variation in rain and temperatures over the country being experienced over the years fall within the “natural variability”.

    “We are keeping a watch. We are not denying…. It (the variations) are still under the natural variability,” Director General of the India Meteorological Department Dr B Lal told reporters here today.

    There has been no significant change in terms of temperature and rainfall on year-to-year basis, he said…..

    “Thus, there is no clear cut signal…We are keeping a watch over temperatures,” he said.

    Lal said there had also been no increase in intensity and frequency of tropical cyclones. Since October 1999, there had not been any super cyclone in the country.

    Last year there were only five disturbances, of which two became cyclones of marginal value, he said…..

    It would appear that some of these scientific conclusions are also backed by research into the socio-economic factors in India related to the increase of tropical cyclone damages:

    “Contrary to the common perception that tropical cyclones are on the increase, due perhaps to global warming, studies all over the world show that, although there are decadal variations, there is no definite long-term trend in the frequency or intensity of tropical cyclones over the period of about a century for which data are available. There is, nevertheless, a sharp increase in the socio-economic impact of tropical cyclones in the form of increasing property damage. An analysis of cyclones affecting the state of Andhra Pradesh, India, in the last quarter century by normalizing cyclone damage for economic and demographic factors shows that here, as elsewhere, the greater vulnerability is attributable mainly to these factors and not to any increase in frequency or intensity of cyclones. The decrease of alertness in disaster management that often occurs after a few years’ lull in occurrence of cyclones, known as the “fading memory syndrome,” also contributes to increases in loss of lives and property damage. This distinction between meteorological and socio-economic causes for the increased impact is important to avoid a tendency for political and administrative decision makers to blame natural causes. They have to take these realities into account, not just in developing a vigilant disaster management system, but in land-use planning, development of coastal districts, and insurance measures.”

    What a refreshing soberness compared to the unjustified alarm in some green circles.

  2. 2
  3. Roger Pielke Jr. Says:

    Benny- Thanks much for this link.

  4. 3
  5. Jim Clarke Says:


    Thank you for setting the record straight on the ABI report.

    It is absolutely ridicules to call for a reduction in CO2 emissions as a means to reduce the threat of tropical storms! Vulnerability reduction is a much cheaper, efficient and proven way to reduce the tropical storm threat. There is really no comparison!

    Imagine a fisherman out to sea in his boat to catch as many fish as he can. He has a large net and a system of fishing rods and bait. In the end, he ignores both, dives into the ocean and chases the fish in the hope of catching one bare-handed. Anyone watching would assume that the fisherman has lost his marbles!

    Those promoting a reduction in CO2 emissions as a way to reduce the threat from tropical storms are like the fisherman diving into the ocean to catch a fish.

    Choosing the most costly and least effective means to solve a problem is just crazy. One may argue that there are other reasons to reduce CO2 emissions, but the tropical storm threat is not one of them!

  6. 4
  7. mb Says:

    Seems to me that’s a bad analogy. First – ameliorative action and vulnerability reduction are not mutually exclusive. More importantly, if the issue of of GCC were simply related to cyclone/hurricane events, you’d have a point. But- potential GCC related impacts on food production, migration, security and biodiversity would remain largely untouched by vulnerability reduction alone.

  8. 5
  9. Roger Pielke Jr. Says:

    mb- Thanks for your comments. I’ve addressed this comment many times. My point is not that climate mitigation does not make sense — it does. My point is that climate mitigation is not an effective tool of disaster mitigation, and should not be presented as such … Thanks!

  10. 6
  11. Dano Says:

    I agree with Jim’s and Roger’s point about framing necessary CO2 reductions for single benefits.

    Reducing CO2 emissions – as Roger has indeed said many times – has multiple, quantifiable benefits. It is a no-brainer. If there’s a slightly reduced risk of storm damage, so much the better.

    CO2, of course, being an indicator of fossil fuel consumption.



  12. 7
  13. mb Says:

    I appreciate that – I was addressing my comments more to Jim’s analogy, perhaps misinterpreting it, because it seems to me that given the political situation in the U.S., it’s probably worthwhile to make note all of the impacts of GCC, even those in which mitigation efforts are likely to be of relatively minor consequence.

  14. 8
  15. Jim Clarke Says:


    My analogy was specific to the problem of hurricanes.

    As to the impacts of GCC – You seem to speak as if these are known quantities and not hypothetical possibilities based on speculative computer models that have yet to show a skill level much higher than chance!

    It seems that much of Europes politicians have accepted these speculations as facts and are willing to sacrifice their economies and countrymen for the cause.

    One day we may see the headline:

    “Many Europeans freeze to death in fear of slightly warmer temperatures!”

    My fear is that many will read such a headline and not see any irony in it.