It wasn’t in our plans to spend the Hurricane Season in the Caribbean. By now, we should be enjoying good food and good weather in the Western Med. Our plan was to cross “the pond” in May, so that we could be in Portugal by June 15th. Then Covid-19 happened, and we thought it would be better to cancel the crossing, since the situation in Europe was so bad.
The Hurricane Season in the Caribbean runs officially from June 1st through November 30th. The Hurricane Box is considered by most insurance companies to be the area between 12 degrees 40 minutes North and 23 degrees 30 minutes North Latitude, and 55 degrees West to 85 degrees West Longitude. The truth is, that if you look at the hurricanes tracks recorded by NOAA since 1990, the area is much bigger and covers all the north Atlantic basin:
Tropical Waves forms on the north west coast of Africa at a rhythm of one every three days in the June/November time frame, and it takes them almost one week to cross the Atlantic. This is how Meteo France identifies them:
NOAA uses a different naming convention than Meteo France, and when a Wave seems to organize in a convection, they call it Invest 9X, where “Invest” stays for “Investigation” and the number goes from 90 to 99. When convection becomes abundant, focused, and persistent, the low pressure system is rated as Tropical Depression and given a number:
This is the last stage before the low is classified as Tropical Storm and given a name.
Contrary to popular belief, Tropical storms are not named after the meteorologist who spots them first. NOAA has a list of names and the 2020 list is this one:
NOAA come out with 6 lists of names that are used in rotation and re-cycled every six years, i.e., the 2020 list will be used again in 2026. The only time that there is a change in the list is if a storm is so deadly that its name for reasons of sensitivity is taken off the list forever.
We were having a great time kiteboarding in Union Island when NOAA reported that there were some showers and thunderstorms associated with a tropical wave located a few hundred miles south south-west of Cabo Verde Islands showing signs of organization.
This is a picture of Invest 99, which will become Tropical Storm Gonzalo, and which scared the hell of us:
NOAA gives these depressions a probability to develop in a Named Storm and further into a Hurricane: <40%, 40/60% and >60%.
When that happens and a Low pressure system turns into a Named Storm, NOAA tries to model its speed and course drawing a cone on a chart, called “the cone of concern”:
To have a better understanding of how the “cone of concern” is created, what its limitations are, and what it can be used for, you can watch this video.
At this point, the clock started ticking for us!
By Tuesday it become clear that a Tropical Storm or Hurricane would pass somewhere between Trinidad and Martinique. Oroboro was just in the middle, between the two islands.
More specifically, the GFS model, which is the American one, predicted that the eye of the tropical storm would pass right over Bequia:
Whereas the ECMWF model, which is the European one, predicted that the eye of Gonzalo would pass over Grenada.
The problem with these forecast models is this: There are so many of them! For now, the more accurate one has always been the European model. Perhaps because NOAA has undergone deep budget cuts by the current administration, and it doesn’t have the computer power of the European agency. But as the say goes “Past performance is no guarantee of future results”. After all, if one model would always be right, there wouldn’t be the need to have so many!
On top of the popular GFS and ECMWF models that everybody knows, there are many others, such as:
- 12zGFS Ensemble
- 12zGFS Operation
- 12z Operational
These models ca be simple enough to run in a few seconds on an ordinary computer, or so complex to require many hours on a supercomputer. All of these models belongs to the following categories:
Dynamical models are the most complex and use high-speed computers to solve the physical equations of motion governing the atmosphere.
Statistical models, do not consider the physics of the atmosphere but instead are based on historical relationships between storm behavior and storm-specific details such as location and date.
Statistical/dynamical models blend both dynamical and statistical models by making a forecast based on historical relationships between storm behavior and atmospheric variables provided by dynamical models.
Trajectory models move a tropical cyclone along based on the prevailing flow obtained from a separate dynamical model.
Ensemble or consensus models are created by combining the forecasts from a collection of other models.
If you want to learn more about forecast models, please click here.
So, based on all these predictions, we had to make a call and decide to sail either South or North, since Gonzalo was predicted to pass right over us. We decided to wait until 48 hours before Gonzalo was predicted to cross the Caribbean to make a call.
The problem with Gonzalo was that being a “relatively small” system, was therefore more difficult to predict its track. If it was going to intensify it would turn north, otherwise it would keep going west.
It was a bet: would Gonzalo intensify, turn into a Category 1 Hurricane and turn north? Or would it stay a Tropical Storm and keep going west?
This was the question that kept us awake at night.
If we went north, we run the risk that if Gonzalo become a Hurricane we would be caught in the northern quadrant (north of the eye), which is the one with stronger winds. If we went south, we would have been in the southern quadrant, with less wind.
The call was made on Friday at 8 am. We were the last boat leaving the anchorage in Bequia, heading south to Trinidad (some 140 nm away). We had a beautiful sail with flat seas and 15 knots of wind, with speed over ground of 7.5 knots. We thought that if this condition persisted all the way to Trinidad, it would have been a very nice sail.
However, later in the afternoon the wind backed north and we were running dead downwind. The wind become light and the sail was not enjoyable anymore.
Then NOAA finally sent their specially equipped aircraft “hurricane hunter” flying over Gonzalo, and when the NOAA 6pm update come out, everybody was surprised that the aircraft found Gonzalo much more south than it was predicted by all mathematical and statistical models. With the data gathered by the aircraft, the models were now very accurate. At this point Gonzalo was predicted to pass between Trinidad and Tobago on Saturday at 2:00 pm.
So it was now clear that the Gonzalo had kept is track west exactly along latitude 10 degrees all the time, and it had accelerated from 12 to 20 knots. Since its formation into a Tropical Storm, Gonzalo didn’t gain even one degree of latitude!
Crazy as it may appear, we would have been better off in Bequia, where we were. We were almost in sight of Grenada’s northern shore when we decided to turn around. At this point the wind was strangely coming from 20 degrees true, so we had to motor-sail head into the wind. We cranked both engines at 2,400 RPM and we were doing 6.5 knots SOG. Half way to Bequia, the wind got stronger as predicted. Now it was blowing 20 gusting 25 knots. At this pace we would have reached Bequia around 3 am.
We knew that some of our friends took shelter in a marina in Canuan, which was half way to Bequia at this point. So I called them up and all of them assured me that the marina was very well sheltered and that it could have easily survived winds up to 60 knots sustained. So we broke the long standing rule to not enter an unknown marina at night and we were really surprised by how well lit the marina was. All our friends were waiting for us on the pontoon to pick up our docking lines. I didn’t have to do anything more than putting the engine in neutral and let the boys work the lines!
Gonzalo’s tropical storm winds were expected the next day, from 11 am to 3 pm. We were expecting 60 knots sustained winds. The next morning, around 8 am, the wind went up to 30 knots.
I thought it was the start. But instead this was it. It lasted only 20 minutes.
Oroboro was safely and peacefully docked.
We were ready to take the promised 60 knots. Luckily, that did never happened. And by 4 pm, we were at the bar of the marina with all the other cruisers celebrating!
What a lesson learned! We enjoyed two beautiful days in a 5 stars marina.