21 June 2009

Schmidt et al. 2009 Replication of Pielke et al. 2008

Silvio Schmidt and colleagues have a new paper in press that replicates our hurricane loss normalization work in Pielke et al. 2008 (PDF). Here is a comparison of the two normalizations over the Schmidt et al. period of record (which begins in 1950, ours dates to 1900), which should give additional confidence in the fidelity of our work, as Schmidt et al. use a different dataset for losses (Munich Re NatCat) and a different approach to normalization. As you can see, over the period of record the differences are minimal (in fact the statistical analysis in Schmidt et al. turns out identically if you use the results from Pielke et al.).

After a detailed look at the data they conclude quite properly:
There is no evidence yet of any trend in tropical cyclone losses that can be attributed directly to anthropogenic climate change.
They do speculate about a link based on the conclusion of IPCC 2007:
The IPCC states that humans have, “more likely than not”, contributed to the trend towards intense tropical cyclone activity since the 1970s. Therefore, any increase in losses could, more likely than not, be partly related to anthropogenic climate change. . . we advance the premise that if losses are affected by natural climate fluctuations, they are also likely to be affected by additional global warming due to anthropogenic climate change. This premise is supported by indications that the intensity of tropical cyclones is affected by anthropogenic climate change.
This is a valuable paper not just because it replicates our work (but of course that is nice to see). The authors also do a nice job clearly distinguishing what can be shown with available data versus what remains in the area of speculation.

3 comments:

Dean said...

Note that "trend in tropical cyclone losses" and "trend in intense trpoical cyclone activity" are not the same thing. And at the early stag when we are looking for the signal, I would expect the latter to be visible before the former. There are many measures for cyclones: wind speed, barometric pressure, total accumulated energy, moving to some of the msot coarse measures like rating at landfall. We don't yet know which of these (and other) metrics should be the first to show a signal, though maybe you have an opinion.

In any case, congrats on the new study that confirms your previous results.

Chris said...

Intersting that the sentences either side of your quote tell a slightly different story...:
"annual adjusted losses since the beginning of the last cold phase (1971) show a positive trend, with an average annual rise of 4% that cannot be explained by socio-economic components. This increase can at least be interpreted as a climate variability impact. There is no evidence yet of any trend in tropical cyclone losses that can be attributed directly to anthropogenic climate change. But we advance the premise that if losses are affected by natural climate fluctuations, they are also likely to be affected by additional global warming due to anthropogenic climate change. This premise is supported by indications that the intensity of tropical cyclones is affected by anthropogenic climate change."

lucia said...

Chris--
Adding the text further supports Roger's interpretation. The authors of Schmidt say they cannot yet attribute losses to AGW. However, because there is reason to believe climate change will adversely affect weather, it's reasonable to form the hypothesis that that climate change is affecting the losses even though we cannot yet detect this or confirm this.

The authors suggest it is, reasonable to advance the hypothesis that climate change will cause increased losses-- it is reasonable to do so. But advancing a hypothesis is not the same as being able to find data to support it or confirm it. While advancing the hypothesis, the authors flat out say there is no evidence yet of a trend in torpical cyclone losses that can be attributed directly to antropogenic climate change.

That is: They put forward a hypothesis. They test to see if the hypothesis is true. They cannot yet find the trend that would confirm it. But, nevertheless, they still advance a hypothesis that the AGW does affect losses. Presumably, they think this hypothesis will be confirmed as AGW intensifies and further data accumulate.

But Roger is correct: Schmidt agrees that, even if their hypothesis is true, the effect cannot be detected yet.

There is a big difference between saying 'we predict the link will someday be shown" and "we think the link has been shown". Schmidt et al say the former, and also specifically tell readers they do not claim the latter.

Post a Comment