A 53-day headstart over 2004

[Ixolib]

If this was last year at this time, we'd still have another 53 days before a TD formed.

http://www.nhc.noaa.gov/2004alex.shtml

[Josephine96]

May mean a really busy season ahead..

[senorpepr]

Statistically, based on the past fifty years, the first tropical storm of the year forms on the 20th of June.

[Ixolib]

Statistically, based on the past fifty years, the first tropical storm of the year forms on the 20th of June.

So, this is even ahead of the "statistics". Sure does make 'ya wonder what the rest of the season has in store.....

[dhweather]

Well, there's so many ways to analyze statistical data.

Mean, median, mode, outliers, what's the standard deviation, I could go
on and on, but I'd look like a huge geek. Wait, I am.

[canegrl04]

"mean,median,mode,outliers"

HUH???

[Ixolib]

"mean,median,mode,outliers"

HUH???

Here's to hoping "DH the Geek" and others like him will continue educating us throughout the season!!

[Brent]

TWC was saying the average is 1 June storm every other year.(About 60 storms since 1886). The last one was 2 years ago, although it was in Late June(June 30th, Bill).

[dhweather]

Well, let me try to make an easy summary of the three:

Mean - it's what most people think of as the average.
In tropical weather, September 10th (or so) is, on average (the mean date)
for tropical systems. A brief example would be you have one storm
on September 10, another on September 20 - so the mean (average) is
Septermber 15th.

Median - It's the numbers in the middle of a series of numbers.
Lets use these days, for example:

Aug 1
Aug 5
Sep 5
Sep 13
Sep 22

The median would be Sep 5 - it's in the middle (or media) of the numbers
laid out in sequence.

Modal - It's the most frequently occuring group of numbers
For example - take a distribution of dates and the numbers of
storms on those dates:

Sep 1 - 3 storms
Sep 3 - 4 storms
Sep 6 - 2 storms
Sep 10 - 8 storms
Sep 17 - 6 storms
Sep 19 - 5 storms

The modal (most commonly occuring group) number is Sep 10 - 8 storms.
If another date (say September 23) had 8 storms also, then you would
have a bimodal distribution.

Another common statistical term is the range. That would, in tropical terms,
be the date of the first storm to the date of the last storm.
That's essentially how the NHC picks the date for hurricane season.
A range of June 1 to November 30 has covered, historically, most
of the systems we've ever had. There are a few anomolies (Ana in April),
but that's part of statistics.

Sorry if you are bored with this geek topic.

[crazycajuncane]

Wouldn't it freak everyone out if this was the first and last storm??

[robjay]

Where typically do June storms originate? Isn't the GOM kind of rare?

[dhweather]

Where typically do June storms originate? Isn't the GOM kind of rare?

The GOM and Carribean are the most popular origins for storms in June.

[Brent]

Where typically do June storms originate? Isn't the GOM kind of rare?

GOM, Western Caribbean and the Western Atlantic around the Bahamas and off the Carolina Coast.

June actually has a higher number of landfalls even though July is more likely to see activity because June's development is always close to land whereas July development can be farther out to sea.

[Derek Ortt]

actually,

dr gray found that there is NEGATIVE correlation between June activity and that for the rest of the season. A very active June and early July often means a very inactive peak of the season

[HurricaneBill]

actually,

dr gray found that there is NEGATIVE correlation between June activity and that for the rest of the season. A very active June and early July often means a very inactive peak of the season

Kinda like 2003. 4 named storms including 2 hurricanes by mid-July. Then nothing until Erika in late August. But then came Fabian, Grace, Henri, and Isabel.

1995 had 12 storms by the end of August. However, during September, the most active month for activity, only Marilyn formed.

However, October brought another rush of activity.

[Derek Ortt]

the correlation was only about -.4, meaning only 16% of the variance can be explained.

There are some glaring examples where this did hold: 1997, 5 by July 20, then only 3 more
1986 is another

[HurricaneBill]

the correlation was only about -.4, meaning only 16% of the variance can be explained.

There are some glaring examples where this did hold: 1997, 5 by July 20, then only 3 more
1986 is another

Both of which were moderate/strong El Nino years.