Terra -- I posted it mainly to diffuse anything before it got started, since it is such a hot-button topic. Thought I had seen something while skimming through the thread as well, but I could be mistaken.

The link posted does indeed take you to the paper as it was published. There are a few primary concerns with the article, which I'll briefly mention...

1) The dataset only extends back to the mid-1970s (pg. 1). The majority (over 2/3) of this dataset includes times both where all of the leading climate indicators, as bore out by storm activity, were for below-normal seasons. Only since 1995 have we seen a major upswing in frequency of storms.
2) The PDI defined in the paper (pg. 1) will be dominated by not just strong storms -- but by storms that last quite a long time. This argues that much of the variation can be explained by an increase in the number of storms -- concurrent with the decadal signals in the data and improved detection techniques to classify storms, particularly in the eastern Atlantic -- as well as an increase in the number of long-track Cape Verde storms, more likely during times of low shear (e.g. La Nina vs El Nino events) given favorable SSTs that appear every season. Thus, this is only tangentially related to the global warming issue, not directly as Emanuel would suggest.
3) "But the large upswing in the last decade is unprecedented, and probably reflects the effect of global warming. We will return to this subject below." (pg. 2) -- again, the data set does not include times of such high activity such as the 1930s-1950s, where activity was similar to that seen in the past ten years, if not higher due to the lack of many eastern Atlantic storms in the data set. Intensity measurements are of dubious nature from that time period as well and if anything are too low. This is where the Hurricane Reanalysis Project (HRP) comes in, helping to provide some confirmation here...bringing things back to Chris Landsea's comments (he's the leader of the HRP). It is quite a large leap of faith to make with little or no evidence to support it.

I cannot make a justified comment about the "bias correction" for data noted in "Supplementary Methods," as it is not available with the article as posted. However, unless many factors have been accounted for -- the Dvorak technique, satellites (both their advent and many improvements since then, including the microwave data now available), greatly increased reconnaisance flights over the past decade, advances in hurricane understanding and knowledge, changes in NHC practices (e.g. subtropical cyclones, classification of storms) over the past decades, and so on -- then I cannot accept these results at face value.

Is global warming having an impact? Perhaps. How much of it can be directly related to hurricanes? Not much. Almost all of the variance can be explained by other factors in our climate. I would expect similar results for the very active times back around the 1930s if we had a similar data set to the one that we have for the past decade. The science in the paper is sound, from the Carnot cycle for hurricane formation/maintenance to the maximum potential intensity to the calculation of the PDI used...it's just the methods are largely incomplete and jump to conclusions, in my view, when much simpler -- if less "trendy" -- ones are more likely the case.

It should be noted that Emanuel's own MPI theory predicts a rapid rise in storm intensity with increasing SST, into the hypercane realm, which are not thought to be possible. It and the work presented here do not take into account compensating factors in the Earth's atmosphere.

BTW, Weatherchef....Kerry Emanuel is a 'he.'