Mighty quiet out there in the Atlantic basin and it would not surprise me if it stayed that way for another week or two, but this is not a post about the current state of the basin. Climatology - the study of past weather conditions - is often thought of as a rather boring branch of Meteorology, however, it can provide interesting insight into limits and expectations for weather events - including tropical storms and hurricanes.

I've taken a look at the past 126 seasons (1880-2005) and captured some interesting statistics. First, the easy stuff: 25% of a season's total number of named storms occurs on or before August 15th, so statistically this would yield 12 named storms for this season since we've already had 3 tropical storms this year. But I also found that if the 3rd named storm occurred on or before August 1st, on average the season would have 13 named storms - but is this really an average season?

Just for fun, I learned that there have been 14 seasons prior to this one that started off with 3 tropical storms - and for all of those seasons, the average was 10 named storms including 4 hurricanes of which 2 were major hurricanes. The last season that started this way was 2002 - as a matter of fact 2002 started with six straight tropical storms and had a season total of 12/4/2. The 4th storm of that year was named on August 29th.

So I started to wonder what Climatology would think about this season if the next storm this year was not named until after August 25th - and the answers are interesting (if you like numbers). In the past 126 seasons, 74 of them did not have the 4th named storm until after August 25th. Nothing unusual there - seems like a common event, but what do the season totals look like when that happens?

Another interesting answer. The average season total is 8/5/2, with the highest seasons at 14/6/4 (1953) and 13/10/8 (1950). This means that in order for Colorado State (Dr Gray et al) to hit its revised forecast of 15 named storms this season, the next named storm must occur before August 26th...OR...the Atlantic basin must set a new record for season activity after August 25th.

I decided to take this statistical investigation one step further. What happens (no matter when the 4th storm occurs) if the 5th storm does not get named until after September 10th (the climatological peak of the season)? It turns out that this happens one-third of the time (42 of the last 126 seasons) - still somewhat common. When this happens, the season totals decrease to 7/4/1 on average, with the highest seasons at 11/5/1 (1898) and 10/5/2 (1943). I also discovered that there has only been one season (1968) where the 4th storm occurred before August 26th and the 5th storm occurred after September 10th. In other words, if the 4th storm happens late, usually the 5th storm does as well, but if the 4th storm happens earlier that August 26th, the 5th storm happens before September 10th.

If the next storm occurs before August 26th, then these statistics are meaningless - at least for this year. But will it...? In our Storm Forum contest for season totals this year, my first forecast on 11/21 was 13/8/3 - amended to 14/8/2 on April 4th and changed again to 15/9/2 on May 13th. One of these years I'm going to learn that your first forecast is usually your best forecast - although if the basin stays as quite as it has been, even 13 named storms may turn out to be a bit too high this season.
Cheers,
ED