Increasing number of strong hurricanes

I'm reading Changing Planet, Changing Health, and I came across a mention of increased number of more intense storms:

Studies over the last five years have shown that hurricanes are getting stronger. In a 2005 paper in Nature, meteorologist Kerry Emanuel of the Massachusetts Institute of Technology described a new measure, the power dissipation index, that captures the amount of energy that flows through a hurricane over its lifetime. ... The same year, Peter Webster's team at Georgia Tech University reported in Science that while the total number of tropical cyclones per year stayed roughly constant worldwide, the strongest storms -- category 4 and 5 storms ... -- nearly doubled over thirty-five years.

Then, the figure from the Webster et al paper was presented. However, the data that was included was from 1970-2004. Looking at that, I realized that there was enough of a time interval since the publication of the paper to extend it by another lustrum. While waiting for students, I decided to track down the data for the period of 2005-2009. Using the various hurricane records from Wikipedia (and roughly estimating the lustral hurricane values published by Webster et al), I was able to come up with the following (expanded) graphic:

As you can see, the most recent lustrum was characterized by a slight decline in Category 1 storms, a serious decline in Category 2 & 3 hurricanes, and a very small increase in the number of Category 4 &5 hurricanes worldwide.

Recalculating to collapse the categories into decades, we can look at how many more (or fewer) storms were occurring worldwide during the 2000s relative to the 1970s. From the data, we find that there were 25% fewer Category 1 storms, 10% fewer Category 2 &3 storms, and 91% more Category 4 &5 storms worldwide in the 2000s compared to the 1970s! To quote further from the book about the significance of this trend:

Top hurricane researchers do agree that as the globe warms, hurricanes will intensify, though the total number of hurricanes will remain relatively constant, or perhaps even drop.

While I don't have worldwide data on this, it seems to track with the trend in the record-high-temp-to-record-low-temp ratio over the continental United States over the same period of time does indicate continued warming:

The trends of an increasing number of high power storms corresponds to the expectation of increased magnitudes of storms caused by increasing global temperatures, and the high-record-temperatures across the continental United States corresponds with expectations of larger-scale warming trends that are being seen throughout the world.

What does all this mean? Well:

In 2010, Science published a study by Morris Bender, Tom Knutson, and Robert Tuleya ... [that] concluded from modeling studies that category 4 and category 5 hurricanes will double in frequency in the Atlantic this century because of ocean warming.

And with an increased number of hurricanes globally comes increased incidence of disease outbreaks, famine, and price spikes in commodities (e.g., food, fuel) increasing the likelihood of war (or at least some sort of punitive action by one group of people against another; state-sanctioned or not).

ADDENDUM: The number of storms in the 1975-1979 and the 2005-2009 lustra were both well below the lustral average of ~240 storms/lustrum; '75-'79 had ~206 storms, and the '05-'09 had 217. All the other lustra experience between ~232 and ~270 storms.

ADDENDUM: Webster et al use the term "pentad" to refer to a period of five years, while I prefer the use of the term "lustrum". Looking at the Wikipedia entry for "lustrum", one finds that "pentad" is synonymous, as is "quinquennium." However, "pentad" is a general term that means a group of five things, while "lustrum" specifically means a period of five years. Of course, "quinquennium" also means a period of five years. However, it's a lot more difficult to spell than either "lustrum" or "pentad".

ADDENDUM (06/30/11): I tried to extend the graph to earlier lustra, but soon realized that this might be difficult, since gaps in the data exist on Wikipedia. (The gaps might not be problematic ... if I actually wanted to wade through all the raw data.) Therefore, since I don't want to commit too much more time on this, I'll be leaving the analysis (as much as it's an analysis) as is.