Earlier this week I described an exchange in the October, 2005 Bulletin of the American Meteorological Society (BAMS) on the subject of the attribution of the impacts of recent disasters to climate change. Our letter (PDF), which was prepared in response to a December 2004 paper by Epstein and McCarthy (EM), concluded, “Future research may yet reveal a connection between climate change and trends in disaster costs, but at present it is premature to attribute trends in disaster costs to anything other than characteristics of and changes in societal vulnerability.”

EM prepared a reply to our letter (PDF) which also appears in the October BAMS. Here I offer a point-by-point rejoinder to their reply.

1. Let’s start on where we agree. EM conclude, “The task is to harmonize adaptation with mitigation and do it in such a way as to stimulate the global economy.”

Response: Sounds good to me.

2. EM restate their original argument as, “We wrote that both increasing coastal populations and real estate prices have increased exposure to weather-related disasters and that the number of extreme weather events has risen worldwide… We do not attempt to quantify the relative importance of these factors in the well-documented increase in insured and uninsured losses from weather-related disasters. We believe our statement is balanced and true to the current literature.”

Response: Where is the evidence that the number of extreme weather events has risen worldwide? The IPCC disagrees (with only one exception). Further, there is a literature that has sought to tease out a climate signal in the disaster record, which is simply ignored by EM. Their statement is in fact contrary to the current literature, including that in the most recent IPCC. Saying otherwise doesn’t make it so.


3. EM assert that the IPCC has identified (with various levels of confidence) trends in extreme related to global precipitation and temperature, which is true. In the case of temperature there is no evidence that such increases, e.g., in diurnal temperature ranges, have any relationship to disaster costs. And we have discussed precipitation and floods here in some detail (e.g., see this post). The trends that EM point to in the IPCC are simply off point, and a distraction to this discussion.

4. EM engage in some misdirection when they state, “The World Meteorological Organization (2004) made explicit the connection between an energized climate system and more extreme weather.”

First, they cite the wrong reference here, pointing in their references to the 2003 WMO report instead of the 2004 report. Second, and more importantly, the 2004 report (here in PDF) actually has conclusions contrary to those suggested by EM.

“According to the third IPCC Assessment Report the duration, location and frequency of extreme weather and climate events are likely to change, and would result in mostly adverse impacts on biophysical systems. The IPCC also notes that for certain extreme phenomena there is currently insufficient information to assess recent trends, and climate models currently lack the spatial detail required to make confident projections. Therefore the linkages between climate variability and climate change and patterns of natural hazards remain a research topic that needs to be further investigated by the scientific community.”

I continue to be amazed to see in this area scholarly citations misused to support otherwise unsupportable assertions.

5. EM then get out on some very thin ice when they cite Mills (2005). We discussed Mills (2005) at length, so no need to revisit that critique. However, this is a good example of how a poor analysis is used to carry weight that it cannot support.

6. EM raise the issue of “the costs of biological sequelae of weather extremes.” These costs have not be explored in detail in the literature and I agree with EM that more attention should be paid to them

7. EM state, “We attempt no rank ordering of social vulnerabilities versus climate variability and severity because we think it premature.” Premature? Numerous scholars have been actively studying this issue, (I’ve been involved for close to 10 years), and a robust, peer-reviewed literature has resulted. Clearly, more work should be done in this area, and given its importance, it is hardly premature. It may be that “premature” in this case is just a synonym for “inconvenient.”

Bottom line: There is nothing in the EM response to alter our conclusion that “Future research may yet reveal a connection between climate change and trends in disaster costs, but at present it is premature to attribute trends in disaster costs to anything other than characteristics of and changes in societal vulnerability.”