2005 Hurricane Season: Early Thoughts

[donsutherland1]

The 2005 hurricane season promises to be another active one. However, it might be somewhat less active than 2004. A higher than normal percentage of named storms should develop into hurricanes. Moreover, a larger than normal percentage of hurricanes should reach major status. At this point in time, early indications are that the season could have a classic profile for above normal major hurricane activity (average of 4.2 per season in those that met such criteria). 8 of 9 (89%) of seasons that met such criteria saw 3 or more major hurricanes.

The PDO could be a critical player. If the PDO averages positive during the 2005 hurricane season, the potential would then be in place for possible excessive major hurricane development (last season met this profile). If not, such development should be above normal. For now, I’m taking the more conservative route, especially as the PDO has had a negative tendency over the past few months. Whether any major hurricanes threaten U.S. landfall is another issue, but that will be examined more closely in a few months (probably early April). For those who desire to know right now, signals are mixed but there is a significant cluster of analogs where such hurricanes made U.S. landfall. Until it becomes clearer as to which analogs are proving strongest, it’s too soon for me to make such a call.

Finally, in another thread, Luis (Cycloneye) observed that the tropical Atlantic is unseasonably warm at this point in time. This could be another hint that reinforces the early analog idea of another active hurricane season.

The December 2004 Tropical North Atlantic Index (TNA) was the highest on record for December at +0.75. FWIW, there were four previous years when the TNA came to +0.40 or above in December and 7/8 (87.5%) had 10 or more named storms in the following hurricane season:

Named Storms:
1953: 14
1963: 9
1970: 10
1988: 12
1996: 13
1998: 14
2002: 12
2004: 15

Average: 12.4

Even more disturbing just as a cluster of emerging analogs pointed to the risk of a major hurricane making landfall, the same held true from the above-noted seasons.

[MWatkins]

The December 2004 Tropical North Atlantic Index (TNA) was the highest on record for December at +0.75. FWIW, there were four previous years when the TNA came to +0.40 or above in December and 7/8 (87.5%) had 10 or more named storms in the following hurricane season:

Named Storms:
1953: 14
1963: 9
1970: 10
1988: 12
1996: 13
1998: 14
2002: 12
2004: 15

Average: 12.4

Don...

Once again excellent work. Well thought out and easy to read...I wanted to get your thoughts on overlaying ENSO trends against the analog years...although the data is probably connected to the very good data posted above...and may also support your concern.

As you know...we are expected to transition into a neutral ENSO pattern by the peak of the hurricane season (if we aren't already in one) coming off of a weak ENSO event in 2004.

Going from the heart of the previous season to your analog seasons:

1953: Moderate La Nina in 1952 to Neutral Conditions in 1953
1963: Weak/Moderate La Nina in 1962 to Moderate Nino in 1963
1970: Moderate to Strong Nino in 1969 to Neutral/Weak Nina in 1970
1988: Strong Nino in 1987 to weak warm ENSO in 1988
1996: Weak warm ENSO in 1995 (yep chek the data) to Weak/Warm Enso in 1996
1998: Strongest Nino Ever in 1997...weak/neutral nino in 1998
2002: Weak Nina in 2001 to Weak/Mod Nino in 2002
2004: Weak positive anoms in 2003 to neutral conditions in 2004

So, right away you can chuck 1953, 1963 and 2002 from the data sets...in both years we moved from cool to neutral or warm conditions. That leaves 1970, 1988, 1996, 1998 and 2004 as the 5 remaining analogs.

That takes the average up to 12.8 (13 to significant digits) with a min of 10 and a max of 15.

Then..filtering more you could also chuck 1988 and 1998 because both years came off extreme nino events...the number actually drops to 12.7.

And even then the 1970 season doesn't really fit this situation either as the 1969 temp anoms were pretty warm compared to 1995/96 and 2003/2004.

So to me filtering on ENSO gives a couple of strong representatives to this upcoming season...1996 and 2004...unfortunately.

Just thought I'd throw my $.02 in...interested to see what your thoughts were on this.

MW

[HurricaneBill]

Named Storms:
1953: 14
1963: 9
1970: 10
1988: 12
1996: 13
1998: 14
2002: 12
2004: 15

Average: 12.4

Even more disturbing just as a cluster of emerging analogs pointed to the risk of a major hurricane making landfall, the same held true from the above-noted seasons.

1953:
U.S. Hurricane landfalls: 3 (Barbara, Carol, Florence)
Major U.S. landfalls: None

1963:
U.S. Hurricane landfalls: 1 (Cindy)
Major U.S. landfalls: None

However, Hurricane Flora plowed through the Caribbean as a Category 4 hurricane, killing 8000.

1970:
U.S. hurricane landfalls: 1 (Celia)
Major U.S. landfalls: 1 (Celia)

1988:
U.S. hurricane landfalls: 1 (Florence)
Major U.S. landfalls: None

However, Gilbert hit Jamaica as a Category 3, the Yucatan Peninsula as a Category 5, and Mexico as a Category 3.

Joan slammed into Nicaragua as a Category 4.

1996:
U.S. Hurricane landfalls: 2 (Bertha and Fran)
Major U.S. landfalls: 1 (Fran)

1998:
U.S. hurricane landfalls: 3 (Bonnie, Earl, and Georges)
Major U.S. landfalls: None

However, Georges moved through the Caribbean as a major hurricane.

Mitch devastated Honduras despite being only a Category 1 at landfall.

2002:
U.S. hurricane landfalls: 1 (Lili)
Major U.S. landfalls: None

Isidore struck the Yucatan as a Category 3.

2004:
U.S. Hurricane landfalls: 5 (Charley, Frances, Gaston, Ivan, and Jeanne) (Although Alex did not make landfall, the Outer Banks took a direct hit)
Major U.S. landfalls: 3 (Charley, Ivan, and Jeanne)

Charley struck Cuba as a Category 3.

Frances struck some of the Bahamas as a Category 3 and Category 4.

Ivan struck Grenada as a Category 3. Jamaica and the Cayman Islands took direct hits while Ivan was a Category 4. The western tip of Cuba took a direct hit while Ivan was a Category 5.

Jeanne struck the Bahamas as a Category 3.

[MWatkins]

Thanks for the updated info Bill...4 of the 5 major hurricane landfalls in the data set you provided occured in 1996 and 2004 (1 and 3 respectively). The remaining 1 major, in 1970, was the last storm chucked from the dataset. That means all of the landfalls occur in the top 3 enso filtered seasons...

Hmmm...pending Don's agreement with methodology/conclusions...you've added more support for Don's concern of an increased likelyhood of a US landfall.

Thanks for running down all of those landfall specifics. Good work.

MW

[HurricaneBill]

In many of those years, it seems the Caribbean was vulnerable as well.

In fact, on at least 3 occasions, there were incidences of hurricanes in the Caribbean within less than 2 weeks of each other.

1. Hurricane Edith and Hurricane Flora 1963

2. Hurricane Isidore and Hurricane Lili 2002

3. Hurricane Ivan and Hurricane Jeanne 2004

Also note that many of the years had a one-month period where many storms formed:

1953:
Mid-September to early October (8 storms)

1963:
September (5 storms)

1970:
Mid-July to Mid-August (4 storms)

1988:
Mid-August to Mid-September (7 storms)

1996:
Mid-August to Mid-September (5 storms)

1998:
Late-August to Late September (8 storms)

2002:
Late-August to Late-September (9 storms)

2004:
August (8 storms)

[donsutherland1]

Thanks, Mike.

FWIW, those years with a very high TNA Index in December are not all among the mix of possible analogs. But the conclusions are not too different. I really haven't reached the point of taking a closer look at the possible analogs as the mix might well change somewhat over the next few months.

FWIW, 2004 appears to rank quite high in the early mix.

Have a great weekend.

[MWatkins]

Thanks, Mike.

FWIW, those years with a very high TNA Index in December are not all among the mix of possible analogs. But the conclusions are not too different. I really haven't reached the point of taking a closer look at the possible analogs as the mix might well change somewhat over the next few months.

FWIW, 2004 appears to rank quite high in the early mix.

Have a great weekend.

Ah ha...I see exactly where you're going I think...

FWIW, those years with a very high TNA Index in December are not all among the mix of possible analogs

Yes...but the subsequent years are.

Year....TNA....Subsequent Year
2004...+0.75...2005 ?
2001...+0.66...2002 Active
1952...+0.55...1953 Active
1987...+0.52...1988 Active
1962...+0.51...1963 Active
2003...+0.49...2004 Very Active...see ENSO post above
1995...+0.47...1996 Very Active...see ENSO post above
1969...+0.44...1970 Active...see ENSO post above

So...coupling positive Atlantic SST anoms with ENSO patterns...we may have something here.

I am buying what you are selling. You have a great weekend...too.

MW

[cycloneye]

As always don an excellent brief but to the point anaylisis.I agree don with you early anaylisis of the 2005 season.It looks like another headeche for the US coastline in terms of landfalls and also for the caribbean as the mean development region will be active.Let's see how the ENSO pattern evolves in the next 3 months but right now I think that it is close to neutral if not already is.

[krysof]

I think this year, the first tropical storm or hurricane will occur much earlier than last year, when a tropical depression formed on July 31st and a tropical storm on August 1st. This year a May storm or Early June is possible.

[donsutherland1]

Krysof,

Statistically-speaking, just over 5 out of 8 hurricane seasons see their first named storm develop prior to last year's initial development on July 31. In the 1851-2004 period, 97 of 154 (63.0%) seasons saw their first named storm develop July 30 or earlier.

Nevertheless, if the unusually warm conditions persist over the Atlantic Ocean as measured by the Tropical North Atlantic (TNA) Index, a strong argument can be made for another season where the first named storm could develop July 1 or later. Such seasons, particularly when the TNA is much warmer than normal during the March-May period, have a greater likelihood than the historic average of seeing their first storm occur July 1 or afterward. The remainder of seasons has a far greater probability of seeing the first named storm develop prior to July 1.

Historic Averages (1851-2004):
• First storm earlier than July 1: 67/154 (43.5%) seasons
• First storm July 1 or later: 87/154 (56.5%) seasons

Going back to 1948 when the TNA was first regularly recorded and examining all seasons where the TNA averaged +0.300 or above during the March-May and April-June periods, the following comparisons can be made:

Here's a comparison based on "High" TNA (Tropical North Atlantic Index: +0.300 or above) and All Other seasons fared:

March-May TNA:

TNA: +0.300 or Above:
• First storm before July 1: 4/13 (31%) seasons
• First storm July 1 or later: 9/13 (69%) seasons

TNA: .Below +0.300:
• First storm before July 1: 27/44 (61%) seasons
• First storm July 1 or later: 17/44 (39%) seasons

April-June TNA:

TNA: +0.300 or Above:
• First storm before July 1: 5/12 (42%) seasons
• First storm July 1 or later: 7/12 (58%) seasons

TNA: .Below +0.300:
• First storm before July 1: 27/45 (60%) seasons
• First storm July 1 or later: 18/45 (40%) seasons

[donsutherland1]

Luis,

I agree with you. The El Niño is now borderline at best and slowly fading. Most or all of the upcoming hurricane season should see neutral ENSO conditions.

[rainstorm]

i am going with a 1980/88 type season. i expect a warmer than normal and dry summer inthe plains to the east coast, which will recurve canes out to sea, or allow for an allen/gilbert scenario

[Derek Ortt]

The Carib does not need another Allen or Gilbert.

They are going to be spending many years recovering from the Charley, Ivan and Jeanne of 2004

[hurricanetrack]

Derek:

Can you see anything now in the pattern that would give a hint to the steering flow? If Florida remains dry through May- and abnormally so- wouldn't that satisfy the ole Lushine Rule of dry May= hurricanes in the Fall?

Certainly there are indicators as to whether or not there will be a persistant east coast trough or ridge? I think we know the hurricanes are going to form- now the big question becomes- where will they go?

Anyone with some insight, please reply.

[MGC]

Hurricane Allen, what a classic storm. I remember having a discussion with a WSO Met from the Mobile Ala office when Allen was still out in the Atlantic. As you may remember, a large high was in place over the US most of the summer resulting in a very hot summer. Anyway, in my discussion with the met I predicted that Allen would go into Northern Mexico or south Texas since the high showed no signs of breaking down. The Met was rather reluctant to make a prediction. Turns out I nailed Allen on the nose.....MGC

[mobilebay]

From all the data sources I'm looking at and reading, it is going to be a busy August, September, and October. Thanks for the great info and insight on this thread. It is great.