THE OFFICIAL HURRICANE SEASON FOR THE ATLANTIC BASIN (THE ATLANTIC, CARIBBEAN SEA AND THE GULF OF MEXICO) RUNS FROM JUNE 1 TO NOVEMBER 30 EACH YEAR

Following image displays estimated wind field at 9:30p SatSep1, when Felix was passing N of LosRoques & Aves off the Venezuelan Coast.

Following image is a close-up of estimated wind field at 3:30a SunSep2, when Felix was passing N of Bonaire off the Venezuelan Coast.

Following image is a wider-angle of estimated wind field at 3:30a SunSep2, when Felix was passing N of Bonaire off the Venezuelan Coast.

Following image displays estimated wind field at 9:30a SunSep2, when Felix was passing N Aruba & Curacao off the Venezuelan Coast.

Following image is a close-up of estimated wind field at 8:00a TueSep4 when Felix was making landfall along Nicaraguan Coast.

Following image is a wide-angle of estimated wind field at 8:00a TueSep4 when Felix was making landfall along Nicaraguan Coast.

July 17, 2006

It's now mid-July, and time for an update on the Tropical situation.  After one system early in June, we've had a quiet 5 or 6 weeks.  Observing the large-scale weather pattern in place across the Atlantic, this quiet period is no surprise!  Trade winds were relatively-light all winter, and sea surface temperatures near the Eastern Caribbean were nearly 1-degree C above normal for much of the winter & spring.  But in June, the Atlantic RIDGE (Azores-Bermuda HI) built much stronger, and Trades increased to nominally 20 kts across the E Caribbean, at times stronger especially E of the Islands, and with frequent GALEs in the Central Caribbean.  June sailing in the E Caribbean was more like a typical March or April, and even the wind direction was closer to E than the usual ESE.

In addition to the strength of RIDGE, its position was farther N than usual, and even allowed for numerous areas of lower pressure to lie in the sub-Tropics, between the Caribbean & RIDGE.  These upper-LOs / TROFs & occasional surface TROFs literally sucked much of the nasty, squally weather into this area between 20N-30N, leaving the E Caribbean with WAVEs lacking some of the strong squalls we often see.

Enhanced Trade Winds in June & early July mixed the water at various depths to a greater extent than normal in summer, and stirred-up cooler water from beneath the sea surface - we've seen SSTs remain unchanged from May & June, near 84F (29C), which is about normal for this time of year (this temp in May was about 1-deg C above normal).  Normal SSTs provide only normal fuel for potential Tropical systems, which is different from 2005, when SSTs were 2-degree C above normal for much of the season.

Stronger Trades these past 6 weeks have helped produce stronger wind shear (net speed difference between wind near in the lower 5000' of atmosphere vs that near 30,000')...wind shear literally rips the strongest squalls away from any clusters of convection which might otherwise go on to develop into Tropical LOs.  Stronger Trades also move WAVEs along at a pace often too rapid to allow focused clusters of squalls to develop.

Looking ahead to the remainder of July, there is some chance wind shear will decrease, or areas of lighter wind shear will become more widespread, allowing a somewhat better chance for a Tropical LO to develop somewhere in the region.  Some areas of higher wind shear are likely to remain, so any developing system would not likely experience favorable conditions for a long-enough period of time to become a major Hurricane.

As for potential tracks...Azores-Bermuda RIDGE remains mostly N&E of Bermuda, with little RIDGE-ing in W Atlantic...so any systems which might develop in the Tropical Atlantic would have a high probability of tracking NW, in the general direction of Bermuda, then recurving N past Bermuda, and NE into the Central North Atlantic.  Only risk for a Tropical LO to strike E Caribbean would come from development of a strong Tropical WAVE passing thru the area, or a Tropical LO which forms very far S in the Atlantic, so its W-NW track is across the E Caribbean.  Long-range outlook does not anticipate any such activity for the remainder of July.

There may be Tropical development elsewhere, but geographical areas with light shear / favorable conditions will be small, so as any developing systems move, they are not likely to remain in favorable conditions for long periods of time.

A note about risk of impact in any given location: 

Meteorologists & the media often hype nasty Tropical weather, and indeed, the destruction caused by an intense Hurricane is awesome.  But I can't tell you how many emails we receive from folks after a storm, asking if a certain Island they're traveling to sustained damage...when the Hurricane did not pass anywhere near that Island.  Probably many more folks cancel their vacation plans, or simply don't travel to the Caribbean in summer or fall.  How risky is it to travel to the Caribbean in summer & fall?

Let's just examine the popular vacation grounds from the VirginIslands to the Leewards & Windwards, which encompasses the entire "Lesser Antilles" of the E Caribbean.  These Islands scribe an arc across the E Caribbean approximately 600 mi long and 60 mi wide, covering about 36,000 sq mi of ocean.  The Caribbean Basin is much larger, stretching over 1500 mi from E to W, and about 600 mi N to S, or an area about 900,000 sq mi.  The average Tropical LO (if there is such a thing), may cover an area at any given time of about 200 mi diameter with Tropical Storm Force winds, damage from which can usually be cleaned-up in a matter of hours or days, and stronger Hurricane Force winds over a smaller area, maybe a 40 mi diameter.

If you hear that "a Hurricane passed thru the Eastern Caribbean", there's about a 1-in-3 chance your vacation destination in the Eastern Caribbean saw Tropical Strom Force winds, which caused little-or-no damage, and less than 1-chance-in-10 your destination saw Hurricane Force winds, and a much lower chance, say 1-in-30 that a specific location saw winds over 100 kts which cause severe & lasting damage.  On average, the Eastern Caribbean only sees less-than-1 such event each year, lasting less-than-1 day as it passes.

So...if you're planning a trip to the Eastern Caribbean from June to December, the risk of seeing Tropical Strom Force winds on any given day, at any given location is less than 1-in-500, with a good chance for pleasant weather on the other days of your trip.  Your risk of seeing Hurricane Force winds on any given day, at any given location may be about 1-in-2000, with a good chance for pleasant weather on the other days of your trip.  Your risk of seeing strong & destructive winds over 100 kts is less than 1-in-5000.

If you hear that "a Hurricane passed thru the Caribbean", remember the Caribbean is a huge area, most Hurricanes are relatively-small, and there's only a slight chance your destination was impacted.  Contact your destination for more details if you're think it may have been in the effected area.

Given the small risk that a Hurricane will impact your plans, you might consider taking advantage of off-season rates, and using some of your savings from the lower rates to purchase travel insurance.  And of course, check our Marine Forecast daily for the latest weather forecast.

******************

June 2, 2006

We are now beginning the 2006 Atlantic Hurricane season, so I suppose it's time for an update on what we can expect this year.  Before we can even begin to make sense of 2006 forecasts, we need to understand a bit about forecasts vs climatology.  You've undoubtedly heard 2006 will be another active year.  But how will it compare to 2005's remarkable season?  Most forecasts suggest 2006 may be about as active as 2005 was predicted to be.  But will 2006 be the record-beraker 2005 was?  The numbers give us no guidance.

Forecasts vs climatology: Let's say we had reason to believe it was going to be 120 degrees tomorrow in Tortola, BVI...if we had reason to believe the factors coming together tomorrow would support temperatures like that, we would forecast it, even though such an event has never occurred.  That's weather forecasting...we can be fairly specific, even predicting rare or record-braking events.  Seasonal forecasts, however, can only suggest how conditions may turn out relative to cumulative climatological averages.  So...all we can expect from Hurricane forecasts is to suggest whether this will be an above-normal season, or an extremely-active season, or, possibly, among the top 5% of seasons.  As you'll read below, there is some disagreement about whether this will be an extremely-active season, or one of the busiest seasons on record. 

2005's forecast was the most-active forecast I've seen...2005 was expected, with a high degree of confidence, to have about 150% to 170% of an average year's activity, and that's a pretty strong statement for a climatological forecast.  The actual numbers came in at 250% to 300% of average, breaking nearly every record I can think of.

2006 forecasts again call for about 150% to 170% of normal activity, which is about all analysis vs climatology can justify.  It's possible we'll see another year with 250% to 300% of normal activity, but the available data can not support a forecast for more than 150% - 180% of normal activity.  I suspect this will not be as active a season as 2005, but more on that later.

Here are the numbers...analysis follows...

2006 forecasts

2005 forecasts

* CSU uses NTC (Net Tropical Cyclone activity).  Though not comparable to ACE, I have roughly converted their NTC to an ACE-equivalent so we can compare forecasts.  NOAA's original ACE numbers are expressed versus the median level of activity (where as many years were more active as were less active than the median).  I've converted their ACE values to be relative to the mathematical average (where all the activity over the years is added together, then divided by the number of years).  UCL presents their ACE relative to the mathematical average, which I prefer.  Median ACE over the 56 years 1950 to 2005 is 87.5.  Mathematical average ACE for the same 56 years is 102.  So...all numbers above for all time periods are adjusted to provide a fair comparison.  Numbers on comparisons found on the Archive page for previous seasons may NOT have been adjusted for fair comparison.

CSU (Colorado State University, Drs Gray & Klotzbach) takes a clinical approach, back-testing nearly every weather statistic available to determine which would have anticipated Tropical activity during previous years.  Previously, I have not been comfortable with their reliance on predictors mostly in the fall of the previous year.  This year, their new set of 4 predictors includes 3 from March-May of the current year, and 1 from the previous fall.  It's hard to argue with the proof they provide of the past predictive power of these parameters.  Here's what they say about this year's season:

SSTs (Sea Surface Temperatures) in the Eastern Tropical Pacific, May, 2006: This is the ElNino / LaNina variable.  Warm, ElNino conditions tend to cause stronger W-to-E wind flow in the Tropical Atlantic, shearing-apart most forming Tropical systems.  SSTs are currently near-normal, with just a slight ElNino, suggesting no enhanced Tropical activity this season.

SSTs near Canary Islands April & May, 2006: Above-average SSTs here suggest the Atlantic sub-Tropical surface RIDGE is weaker than normal, and surface Trade winds in the Tropics may be weaker than normal this summer.  Weaker Trades support warmer SSTs in the Tropical Atlantic, lower surface pressures, greater tendency for surface LO formation, and less wind shear due to lighter upper-air winds.  This variable is giving a strong indication for an active Tropical season.

Surface pressure just N of Equator from Africa to Brazil, March & April, 2006: Lower than normal surface pressure here during Spring is a strong predictor of the same factors as SSTs near the Canary Islands above.  This variable suggests slightly below-normal Tropical activity this season.

Upper-air pressures from Greenland to England November, 2005: In addition to its suggestions about all the factors above, this variable may suggest more RIDGE-ing in the Central Atlantic.  This variable is sending a strong signal for an active Tropical season.

CSU also uses an analog comparison technique to years with similar patterns.  Similar years are 1961, 1996, 2001 and 2004.  Average activity for these 4 years is very close to CSU predictions for 2006.

CSU also offers guidance on where Tropical systems may tend to go this season.  To summarize, the strong predictive signal is for weaker W-to-E upper-air winds in the mid-Latitudes of the Atlantic, so Tropical systems may have less tendency to recurve to the E as they turn from a W-bound track beneath the RIDGE to a more N-ward track along the W periphery of the RIDGE.  This could increase risk for a strike along the Eastern US.  CSU also uses a predictor for RIDGE-ing near SE US, which would tend to drive Tropical systems into the GulfOfMexico...and this variable predicts a below-normal tendency to drive Tropical systems into the GulfOfMexico.

What does this mean for the Caribbean?  The sum of these factors suggest this will be an active season for Tropical systems in the Eastern Caribbean, as systems are likely to form in the Tropical Atlantic.  Normal RIDGE-ing N of the Tropical Atlantic may tend to steer storms toward or just N of the Eastern Caribbean, and drive them in the general direction of the Eastern US.

NOAA's analysis emphasizes the tendency for about 20 years of active Tropical seasons, followed by 20 inactive years, then 20 more active years...known as the Multi-Decadal Signal.  We are in a period of active years, and about 10 more active years are likely before we return to less-active years.  Patterns we tend to see during active years under the Multi-Decadal Signal are higher SSTs in the Tropical Atlantic, lower surface pressures, more moisture, stronger upper-air RIDGE N of the Tropical Atlantic, reduced shear, lighter Trade winds, and lighter mid-level winds in the Tropics which also reduces shear...each of these factors supports above-normal Tropical activity.

SSTs in the Tropical Atlantic, while above average, are not as far above average as they were in 2005.

ElNino should neither enhance nor reduce Tropical activity in the Atlantic this season.

Interestingly, NOAA is predicting a Tropical season which is farther above-average than either UCL or CSU.

UCL (University College of London) and the Tropical Strom Risk group rely mainly on 2 predictors: SSTs & Trade winds in the Tropical Atlantic.  Warmer SSTs support more Tropical activity.  Lighter Trade winds allow developing Tropical systems the ability to spin up into LO pressure systems more readily.  Both predictors are signaling a slightly above-average Tropical season.  UCL also considers the Pacific ElNino / LaNina influence, and believes a slight warming (ElNino) may provide a bit more shear in the Atlantic later this season.

Summary: Though it's difficult to argue with CSU's research (and I don't), I do prefer predictors whose signal is based on currently-observable factors, such as the present SSTs, Trade winds, and other factors.  But for this season, it doesn't really matter which methodology you use...CSU, NOAA and UCL all predict 2006 will be an active Tropical season, probably among the top 5 most active of the past 56 years, but probably not anywhere near the 2005 season in terms of the number of systems or the ACE.

What's the bottom line for the Eastern Caribbean?  Based on statistical research from the University College London, there will be 0.6 systems impacting the Lesser Antilles with sustained winds of Hurricane force (74 mph or greater), versus the 56-year average of 0.4 storms, or about 50% more than the "average" risk.  Of course, you can have either 0 storms or 1 storm, and both figures fall between the two.  If the forecast is correct, we may-or-may-not experience one hurricane.  JUST REMEMBER, WHATEVER THE NUMBERS….IT ONLY TAKES ONE!

Sea Surface Temperatures in the Tropics have a major impact on Tropical activity, with warmer Sea Surface Temperatures increasing atmospheric potential energy by supporting higher air temperatures and greater evaporation of water; both key ingredients for Tropical systems.  SST is one of the factors in the 20-25 year pattern we discussed above.  The image below shows startling correlation between SSTs and Accumulated Cyclone Energy (ACE).

The following chart shows strong correlation between SST and Tropical activity

The following image shows current SST departure from normal.  Notice the overall bright-yellow coloration in the Tropical Atlantic between North Africa and the Eastern Caribbean?  There are areas of pale-yellow (not as much above-average) and yellow / orange (very much above normal), with the average nearly 1 deg C above normal.

The Caribbean Weather Center Ltd

EMail: weather@caribwx.com

Tel/Voicemail: 1-284-494-7559 Fax: 1-284-494-5358
Columbus Centre, P.O. Box 3069, Road Town, Tortola, British Virgin Islands