A reader writes... I trust you will critique Ingrid is born; Humberto and Felix--a sign of climate change?. Naturally I'm stupid enough to fall for that sort of a challenge.
On the issue of cat5 records I think he misses the obvious point (stoat passim): cat5 numbers are low, their statistics are unstable, and you can't tell much from the record. He does say "We've now had six Cat 5's in the past three years, and eight in the past five years. Is this an indication climate change is at work? Well, we did have back-to-back years with two Cat 5's each (1960 and 1961), so one can still argue that the Cat 5 activity of recent years is a statistical abnormality. In addition, recent work done studying sediment deposits indicates that intense hurricanes have gone through cycles lasting hundreds or even thousands of years long. Periods of high Category 5 activity similar to that observed the past five years could well have occurred in the distant past." but then ends the piece with "But, the events of 2005 and again this year leave me concerned. Eight Cat 5's in five years is an awful lot of severe storms in such a short period. Climate change may be indeed be changing Atlantic hurricanes for the worse." This is all vagueness, but slanted vagueness. You cannot say anything useful *from the numbers alone* but you can spend a long time talking around the numbers in an effort to imply something.
The other thing worth jumping on is this "rapid intensification" thingy. His conclusion is pretty fair: "No scientist has published a paper linking rapid hurricane intensification rates with global warming. While the cases of Humberto and Felix are certainly unique, the year 1969 also had two storms that were very similar in their intensification rates. A quick look I did at historical intensification rates don't show any noticeable trends, and I think that the rapid intensification rates of Felix, Humberto, and Wilma the past three years are not far enough outside the statistical norms that we need to invoke climate change as an explanation. Still, it does leave one wondering, and climate change could be affecting hurricane intensification rates." What makes me suspicious is that its a retrospective sign: only now we've had an otherwise unexciting hurricane distinguished only by rapid intensification is anyone interested in RI.
- Log in to post comments
The vast majority of Atlantic hurricane seasons do not have storms in May or June . (First storm is usually around July 10) .
As has been covered recently - maybe you didn't notice - category 5s are rare in the Atlantic - only about 1 every 3 years, and most weaken before landfall. Two landfalling category 5s is quite exciting. It is true that total number of storms is not extradinory, but since the 9th storm does not usually show until about the 12th of october, it's still above average. This is actually a very exciting season, unless one has a completely ridiculous benchmark like 2005, which overran dozens of records, often by huge margins (28 storms versus 21?).
(Of course it's quite reasonable to argue that there isn't enough RI data to draw conclusions about trends, which seems to be your main point.)
The usual tactic for dealing with sparse data is to average over longer periods (like 10 yrs.) At this point the best strategy is a lot of suspicion and watchful wariness (which I have mis-spelled) and increased efforts at modeling which is probably the best approach for actually getting at the issue before Eli starts pushing up carrots.
Finally, AFAIK cyclone intensity is a marker for increased SST and warming of the oceans over which the storms pass. It seems to me that the record on those issues are fairly clear.
Again, I think there is nothing wrong with finding suggestions of a trend where the data are insufficient to demonstrate one. If you offer the right caveats, if you know you might be seeing mirages in the fog, it does no harm to say you suspect you see something more substantial.
We have never thought about the rate at which these storms spin up, but we have thought about their maximum intensity, and there are reasons to suspect the latter should increase and are doing so.
Is there some theory as to how fast storms spin up? It seems like a modeling study wouldn't be infeasible. This sort of suggestive data, erm, suggests such a study. I'd have a better idea how to go about it than getting an Eemian polar bear population, for instance.
The right balance between worry and insouciance is a judgment call. Either tendency can be and is horribly misinterpreted, but there is no reason to call out the person who inserts the right caveats.
Michael has the right idea, and in a decent adaptive management scenario for our society, we'd have alarm bells going off over our indicators, triggering whatever things we said would be triggered by the numbers generated by the indicators. .
Oh, wait: the insurance industry uses adaptive management and scenarios and is pulling out of certain areas and the RE industry is doing the same. Too bad the IPCC isn't doing a good job (*cough*) at explaining scenarios for the rest of society and for the media to explain to societies...
Best,
D
I think Masters inserts the right caveats. I myself have always tried to follow his model.
Dano, the Bush administration's failure to follow up on the clinton administration's national assessment process for climate change risks is, I believe, more to blame here than the IPCC.
FWIW, Balling and Cerveny (2006) examined Atlantic tropical cyclone intensification rates and SSTs and found little relationship.
Balling, R.C., and R.S. Cerveny, 2006. Analysis of tropical cyclone intensification trends and variability in the North Atlantic Basin over the period 1970-2003. Meteorology and Atmospheric Physics, 93, 45-51.