The Intersection

21 Cat 4-5 Storms for 2006?

i-814642326fd6e71132ab1764a3abfa94-Monica April 24.jpg

Cyclone Monica on April 24, 2006, possibly the strongest recorded hurricane in the Southern Hemisphere.

So: According to Jeff Masters, there were 21 Category 4 or 5 hurricanes globally in 2006. Zero of them were in the Atlantic, which makes the total number sound even more staggering.

I actually disagree with Masters on this; after doing my own count using Wikipedia, I only get 20 of these storms; and after doing my own count with the “best track” data provided by Unisys Weather, I get 19. But we’ll get to that in a minute. First, let’s talk about what the number, whatever it is, actually signifies.

According to Masters, the long term average for the number of Category 4 and 5 storms globally in a year is 17. So whatever the correct number is, 2006 sounds like an above average year (albeit perhaps only slightly). And if you’ve got enough above average years, well, you might have a trend. A scary one.

But as you may know, there’s still a big debate about whether such a trend exists, and whether these strongest and deadliest hurricanes are actually increasing globally. We can’t identify a trend without knowing what the past was like, but our past data on intense storms has been called into question by scientists like Chris Landsea, who think these storms were undercounted in previous years because our technologies for measuring storm intensity weren’t as dependable back then.

Because of this ongoing debate, we can’t really know what to make of the fact that there were 21 (or 19, or 20) Cat 4 and 5 storms this year. But we can say one thing: Clearly, you can have a dead year for the most intense hurricanes in the Atlantic but still have a very active year for them globally.

But now, let’s get into the numbers. First, I did a search using Wikipedia. This isn’t the same as going straight to the “best track” data, but I find that the Wikipedia people are generally pretty good with hurricane basics. Plus this allows me to provide links.

Based upon my searching of Wikipedia, I only count 20 Cat 4 and 5 hurricanes for 2006. I have listed them alphabetically by basin:

Northwest Pacific (9):
Supertyphoon Chanchu, Cat 4
Typhoon Chebi, Cat 4
Supertyphoon Cimaron, Cat 5
Supertyphoon Durian, Cat 4
Supertyphoon Ewiniar, Cat 4
Supertyphoon Saomai, Cat 4
Typhoon Shanshan, Cat 4
Supertyphoon Yagi, Cat 5
Typhoon Xangsane, Cat 4

Northeast Pacific/Central Pacific (3):
Hurricane Daniel, Cat 4
Hurricane/Supertyphoon Ioke, Cat 5 (Central Pacific/moved into Northwest Pacific)
Hurricane John, Cat 4

Southwest Pacific (3):
Severe Tropical Cyclone Larry, Cat 4 [?]
Severe Tropical Cyclone Monica, Cat 5
Cyclone Xavier, Cat 4

South Indian (4):
Cyclone Bondo, Cat 4
Cyclone Carina, Cat 4
Severe Tropical Cyclone Floyd, Cat 4
Severe Tropical Cyclone Glenda, Cat 5

North Indian (1):
Cyclone Mala, Cat 4

So, why do I get 20 storms and Masters 21?

First, I don’t know whether Masters is using Wikipedia or some other source. But one possible reason for confusion is Ioke, which crossed basins from the Central Pacific to the Northwest Pacific. On Wikipedia, Ioke is listed both under 2006 Pacific Hurricanes and under 2006 Pacific Typhoons. Similarly, Unisys Weather lists Ioke for both the East Pacific and the West Pacific. Masters counts 3 Cat 4-5 Northeast Pacific storms and 10 Cat 4-5 Northwest Pacific storms, but that’s only possible if you count Ioke twice, in my opinion.

There are other possible sources of confusion in the data as well. For example, in addition to double-listing Ioke, Unisys Weather lists Cyclone Monica in both the South Indian and the South Pacific. This too could lead to double counting. Finally, Wikipedia gets smart and categorizes Cyclone Larry as a Cat 4 on the Saffir-Simpson scale, whereas Unisys classifies the storm as a Cat 3. I’m not sure who’s right, but again, this could introduce confusion.

Anyway, I’m going to go with the number 19, trusting the Unisys data on Larry rather than Wikipedia’s interpretation, and only counting Ioke and Monica once each. By these lights, both 2004 and 2005 were more staggering global years than 2006. According to my quick Unisys count, there were 23 Cat 4 and 5 storms in 2004, and 22 in 2005. In each case, that has a lot to do with the fact that there were plenty of Cat 4 and 5 storms in the Atlantic during these years.

Of course, this method of intense storm counting is itself simplistic. As scientists like Kerry Emanuel have pointed out, it’s not at all clear why we should be counting a record-breaking long lived intense storm like Ioke on a par with a much shorter-lived Cat 4 storm like, say, Bondo or Xavier. But that opens up a whole different can of worms….

P.S.: Masters lists 101 total tropical cyclones. As the average is 80-90 (he puts it at 86), this also seems significantly high.

P.P.S.: It’s easy to make little bone-headed errors with this kind of counting, so I may well have done so….anything caught will of course be corrected.

P.P.P.S.: The National Climatic Data Center also categorizes Larry as a Saffir-Simpson Cat 3. Wikipedia seems to be improvising.


  1. #1 Dark Tent
    January 4, 2007

    For determining whether the number of intense storms is increasing, I’d say a simple count of Cat 4 and 5 hurricanes based on the saffir-simpson scale is of limited value and not just becasue of the point about not factoring in storm duration.

    A storm can be at the upper reaches of a Cat 3 and not get counted with this method, while it has virtually the same intensity as one that just barely made Cat 4 and did get counted.

    Also, the cat 4 and cat 5 storms only tell part of the story.

    Presumably, increased SST’s would boost the intensity of all storms and do something else as well:

    I am particularly intrigued by annomalies like the two named storms (one tropical storm Rosa, the other a hurricane sergio) that formed in November in the eastern north pacific, the first time since 1961 that this occurred.

    Sergio, though only a Cat 2, was the longest lasting tropical cyclone on record in the eastern north pacific in November.


    If there is ideed a minimum temp at which hurricanes form (and the indication is that there is), it would make sense that if SST’s are increasing, one would start to see hurricanes forming outside the normal tropical storm season.

  2. #2 Chris Mooney
    January 5, 2007

    Hey Dark Tent,
    Yeah, well, there were two late season atlantic storms last year, Hurricane Epsilon (December) and Tropical Storm Zeta (December-January). And in 2003 there was a radically early season Atlantic storm, Tropical Storm Ana, in April.

    What to make of all of this is another matter. No one record breaking storm proves anything. But Peter Webster, Judith Curry, et al have been arguing that season lengths are indeed increasing…..

  3. #3 Lion Kuntz
    January 7, 2007

    Cyclones/Hurricanes happen at sea. Nsautical measures are used by default. 100 knots is impressive to sailors and landgrubbers alike, so 100 knots is the dividing line for MAJOR Hurricanes. That’s 115 mph, or 185 kph.

    The US Navy has to operate in all seas at any time, so the Navy site is the authoritarian one, where wikipedia gets it’s data. It keeps extensive archives going back years, especially the current (recent) year, and is less complete as one goes back in time.

    On average there has been one major hurricane every 13-14 days somewhere in the world since Katrina leads the list. That is 35 current major hurricanes.

    The navy uses uniform measuring systems, so it is comparing apples to apples, not apples to oranges. It is also better equipped than most entire nations with access to over 20 satellite platforms daily. Some nations only record the values for cyclones that hit them, not stronger ones that pass them by. They can only tell you the speeds when the storm got into radar range, not elsewhere for the same storm.

    The navy website records 75 named storms for 2006, including tropical storms and five categories of hurricanes:

    Using Naval windspeed maximums 19 category 4 & 5 storms were active in 2006.

    The parameter has very limited usefulness by itself. Cyclone Shanshan hit Japan at category 4 but hit the USA many days later below tropical storm strength but causing 70 tornadoes in the Southeast USA. More people died in the USA than died in Japan from the same weather event.

    The USA and territories was struck by six asian cyclones in 2006, and IOKE deposited a calculated 27 megatons of heat energy to melt ice in the Beaufort Sea. That heat storage is now responsible for most of the bitter winter in the west USA, and is killing 340,000 cattle as you read this now.

  4. #4 RickD
    January 9, 2007

    Like you saying, counting storms is a bit silly. Suppose the entire Pacific basin were covered by a typhoon that lasted all year long. That would count as one storm?

    If you wanted to do this a bit more scientifically, you would do something like weigh each storm according to its duration. So I guess I’m agreeing with Kerry Emannuel. You could do something like (square miles of storm) * (hours of storm duration) as a measure of impact. And you could also add in a factor relating to wind speed.

New comments have been disabled.