Statistics behind late-starting seasons

[Anonymous]

We're approaching Jul 20 and still have yet to see a classified TC in the Atlantic basin. 60-year climatology says that we should have had about 1 named storm by now...so we're running slightly below schedule. This means the season won't be as active as originally thought...right? Wrong. First off, the amount of activity we see during the first 2 months of the season has no correlation with how active the season is itself.

Here are some interesting climatology statistics (1949-2003)...comparing the mean amount of hurricane activity before Jul 20 in above average, average, and below average seasons.

Above Average Years:
1.00 NS / 0.38 H / 0.13 MH

Near Average Years:
1.10 NS / 0.45 H / 0.03 MH

Below Average Years:
0.96 NS / 0.43 H / 0.04 MH

Again, these numbers represent the mean amount of TC activity in the Atlantic basin before Jul 20 for years 1949-2003. As you can see, there is NO significant variation between pre-Jul 20 activity in all 3 types of seasons. We're dealing with only slight differences in the nearest hundreths...statistically insignificant.

And for curiosity purposes, I looked at all the years that had 0 NS / 0 H / 0 MH before Jul 20 as looking to be the case this year. Turns out there were 21 of them (about 1 every 3 seasons on average)...10 went on to be above average seasons, 9 below average, and 2 near average. Once again, that's an extremely minute correlation...technically it's slightly in favor of above average seasons.

Here are some examples of hurricane seasons that got off to a slow start and yet ended up active...

1949
First Storm: Aug 21
Season Total: 13/7/3

1950
First Storm: Aug 12
Season Total: 13/11/8

1955
First Storm: Jul 31
Season Total: 12/9/6

1961
First Storm: Jul 20
(Second Storm: Sep 2...had to mention this)
Season Total: 11/8/7

1969
First Storm: Jul 25
Season Total: 18/12/5

1980
First Storm: Jul 31
Season Total: 11/9/2

1984
First Storm: Aug 28
Season Total: 12/5/1 (H and MH were lacking but caught up big time in NS)

1988
First Storm: Aug 5
Season Total: 12/5/3

1990
First Storm: Jul 22
Season Total: 14/8/1 (same as 1984, big NS catch-up)

1998
First Storm: Jul 27
Season Total: 14/10/3

2000
First Storm: Aug 3
Season Total: 15/8/3

As you can see, there have been many seasons in the past that have had slow starts only to have more named storms (and in most cases hurricanes and major hurricanes) than normal. And as the statistics above show, there have also been seasons with slow starts and slow seasons. Putting the slow start aside, as far as the 2004 hurricane season is concerned...atmospheric setup still favors an above average year. ENSO is neutral, so that's not going to have any noticeable influence period...this leaves us with a very strong ATC and westerly QBO. Both are favorable for hurricane development (especially in the Mean Development Region) this year. The slightly positive phase of the NAO combined with latest SLPA and SSTA patterns still suggest that the east coast will most likely escape a direct hurricane landfall...though the GOM coast (particularly the FL Panhandle and southern TX) looks to be in for a rough ride if the current steering patterns persist through Aug and Sep. Of course, as some have pointed in another thread...since the general pattern often "hiccups", timing can be important and lead a storm on a whole different path than what the dominant pattern would indicate.

So back to the topic of this post, for those of you thinking of lowering your hurricane predictions JUST because we haven't seen anything yet...I'd think twice.

[cycloneye]

That was a great post SC and shows that the season really starts in august especially from mid august when the CV season gets going.For my part I will no change my forecast numbers 13/8/3 as I never do so during the whole season.

[caneman]

Excellent Rob

[Hurricanehink]

Wow, great job and great post!

[elw]

Great analysis Rob!

[frederic79]

but read with great interest the stats on slow starting seasons. Here's some that I dug up that seem interesting in light of the date we're at without an A-named storm. Since 1994, there has been three years in which the A-named tropical storm formed after July 13. In 1998, 4 tropical storms and 2 hurricanes affected the Gulf Coast between Galveston and the Florida Keys. During 2000, 3 tropical storms and 1 hurricane affected the Gulf Coasts of Mexico and Florida. And in 2002, 4 tropical storms and 1 hurricane affected the Gulf Coast from Galveston to the Big Bend area. In contrast, the total Gulf Coast strikes for all seven years where the first named storm formed prior to July 13 are 7 hurricanes and 12 tropical storms. By comparison, years that the A-storm formed July 14 or later, the Gulf Coast saw an average of 5 named strikes per season. In seasons that started earlier in June, the Gulf Coast saw an average of just 2.7 named strikes per season. Now that we’ve passed July 13 without an A-named storm, I'm curious to see if this trend continues for the Gulf Coast. Just an aside, I got curious so I looked up memorable years where names were retired for Gulf storms: 1965, 1969, 1979, 1985 and 1992. Guess what. A storms for those years were late too, with the exception of 1979, where Ana formed June 19 near Cape Verde and was considered a once a century occurance. In 1965 when Betsy hit NO, the first named storm was Aug 21. In 1969 with Camille, Anna formed July 23. In 1985 with Elena, Ana formed July 15. And in 1992, infamous Andrew formed Aug 16. I wish I had the time to go back 50 years and look for a definite trend. Let's just say many years in which A storm started mid-July or later saw increased Gulf strikes including many whose names are now retired. That's an attention-getter in my book.

[Anonymous]

but read with great interest the stats on slow starting seasons. Here's some that I dug up that seem interesting in light of the date we're at without an A-named storm. Since 1994, there has been three years in which the A-named tropical storm formed after July 13. In 1998, 4 tropical storms and 2 hurricanes affected the Gulf Coast between Galveston and the Florida Keys. During 2000, 3 tropical storms and 1 hurricane affected the Gulf Coasts of Mexico and Florida. And in 2002, 4 tropical storms and 1 hurricane affected the Gulf Coast from Galveston to the Big Bend area. In contrast, the total Gulf Coast strikes for all seven years where the first named storm formed prior to July 13 are 7 hurricanes and 12 tropical storms. By comparison, years that the A-storm formed July 14 or later, the Gulf Coast saw an average of 5 named strikes per season. In seasons that started earlier in June, the Gulf Coast saw an average of just 2.7 named strikes per season. Now that we’ve passed July 13 without an A-named storm, I'm curious to see if this trend continues for the Gulf Coast. Just an aside, I got curious so I looked up memorable years where names were retired for Gulf storms: 1965, 1969, 1979, 1985 and 1992. Guess what. A storms for those years were late too, with the exception of 1979, where Ana formed June 19 near Cape Verde and was considered a once a century occurance. In 1965 when Betsy hit NO, the first named storm was Aug 21. In 1969 with Camille, Anna formed July 23. In 1985 with Elena, Ana formed July 15. And in 1992, infamous Andrew formed Aug 16. I wish I had the time to go back 50 years and look for a definite trend. Let's just say many years in which A storm started mid-July or later saw increased Gulf strikes including many whose names are now retired. That's an attention-getter in my book.

Nice observation...goes along with our GOM 04 thoughs...interesting. This could become another research project.

[Guest]

Great post SC but you should copy it & paste it & keep it in your files for future reference because no doubt the next time there is a slow start the same people will start whining.

[stormchazer]

Thanks for the info.