Getting across the Gulf Stream without undue drama is often a challenge in late Fall. Our tytpical approach is to get down to Lake Worth and hang out till conditions fit our criteria. More on that later.
Privateer plans
Normally we are in no particular hurry to get across. Usually our first schedule commitment is to meet up with our sons in the Exumas for Christmas. But this year we planned a major family reunion for Thanksgiving in Spanish Wells. Still not a problem if there were no boat mechanical issues and the weather pattern was typical. Unfortunately, that is not how it worked out. We spent a week on the hard in early November dealing with a propeller vibration issue.
As the work progressed we watched the weather, hoping the boat would be ready when a weather window opened up. By the 9th we were seeing a possible window for the 12th, but at that point we had neither props or shafts installed. Since this post is about the GS and not our mechanical issues, I will skip the details and just say that the last technician left the boat at 6pm on the 11th and we departed the morning of the 12th. Our sea trial would be conducted as we motored out Lake Worth inlet.
Gulf stream crossing strategy
Back to our criteria for GS crossings. Our usual navigation approach is to cross perpendicular to the current, with constant course over ground (COG) and speed through the water (STW, while varying heading (HDG) as required to hold course and letting speed over ground (SOG) be what it will be. We let the autopilot drive the boat.
This method, sailing the shortest distance from one side of the stream to the other, is the most efficient for our boat speed of 8+ knots, versus the typical GS current of 2- 2.5 knots. Normally we cruise a bit slower for fuel efficiency, but the greater the speed differential between the boat and the current, the less effect the current has on the overall trip. With a slower boat, or significantly stronger current, the shortest time approach, holding HDG and accepting the set, might be more efficient. Back to our criteria for crossing conditions.
Weather and seas
A forecast of NE 5-10G15 and 3′ seas will keep us in port. These conditions will create a rather nasty set of steep, short period seas near the nose, often much higher than the forecast suggests. Change the wind direction to NW and we will accept a little more wind. Winds with a southerly component still seem to stir up the stream more than they would on no-current waters so even with a southerly wind we avoid anything over 20 knots.
So lets get back to this crossing. By Wednesday the forecast for Friday was looking better. We use Windy and Predict Wind for their presentation of the various GRIB models. If the GRIBS are not in agreement, we tend to place more stock in the Euro model. We rely on NOAA for a summary and Chris Parker for a sanity check. Chris’ approach is to forecast the worst conditions one would likely encounter, so if his forecast meets our criteria we are good to go. We usually start looking a week out and then watch the trends. On Tuesday, NOAA was predicting NW10-15G20 and 4-5′, a no-go, but by Wednesday had dropped back to 5-10G15, a favorable trend.
Here is what the GRIB models were forecasting, courtesy of Windy:
Windy forecast with Euro model as of Tuesday for Friday
A minor risk of squalls and maybe a thunderstorm but otherwise a great forecast. This forecast pretty much held through Friday morning and was supported by both NOAA and Chris Parker. All the forecasts suggested a NE swell at 3′ and 9 seconds, coming down to 2′ later in the day. We were a GO!
Geluf stream current forecasts
The GS forecast for current speeds also looked pretty good. We use Passage Weather for these stream forecasts. The max current of 2.5-3 knots looked to be for a fairly narrow slice of the stream, with mostly 1 to 2 knots. We have encountered as much as 4 knots but that is more typical closer to the Keys.
GS forecast with our route added
We cleared the break wall at Lake Worth inlet at about 8am and in less than five miles we began to pick up a little current. Luckily, the swell was already 2′ or less and just got better from then on. The following table lays out our trip with mileage, speeds courses and headings. From this data I was able to calculate the set and drift of the current. Using the Passage Weather chart I estimated the current at intervals along the route and compared them to actual performance.
Actual Set and drift
I was surprised to see that despite a maximum crab angle of 21 degrees at 1000am, our speed over ground was still pretty good at 7.9 knots. That can be explained by the NNE set during the strongest portion of the stream. All this assumes our speed through the water was in fact a steady 8.3 knots, which may have been a bit low. We had essentially flat seas and at times a bit of a tail wind which might have improved speed through the water.
Once we were within sight of West End the current was minimal so we cut the corner a bit and headed straight for Spanish Wells, arriving at about 9am on Saturday. I had been so rushed when we were leaving that I never checked on the moon phase. I was happy to see a half moon near overhead just after sunset that lasted until the early ours of the next day. Plenty of tankers and cruise ships (yes, they are back) on this route but they all played nice and dodging them gave us something to do.
Sunrise as Privateer approaches Spanish Wells
Now that we are here we appreciate how lucky we were to find any window at all. Wishing all of you will find equally pleasant weather windows.
A while back I posted about some of our experiences crossing the gulf stream, good and bad. That post discussed various approaches to getting across as quickly and easily as possible. For this trip we were heading to West End from Lake Worth inlet, using all of a fairly narrow weather window.
For planning purposes, I calculated 55 nm at 94 degrees (T). I estimated the current to be an average of 2.5 knots for the 40 nm of the stream. That would equate to an average speed of 6.5-7 knots and 8-9 hours. Since the route is roughly perpendicular to the stream I chose to take the rhumb line route rather than play the current.
I used NOAA gulf stream data, sourced from the Passage Weather App and weather from NOAA, Windy, and Chris Parker. The weather forecasts were generally in agreement that conditions would be decent, but within that broad comment, there was a fair amount of variability, as you will see in a minute. Since we were looking at about 12 hours of relative calm between two cold fronts, that would likely explain the range in the forecasts.
For this post, I will first present the forecasts and planning data, then the actual results. We will cover both weather and GS current forecasts. So lets start with the NOAA GS forecast, courtesy of Passage Weather, for Friday, 15 Jan 2021:
This forecast for GS current shows a mostly yellow (2-2.5 knots). From NOAA data the west boundary of the stream was 7 nm from Lake Worth Inlet. A rough estimate of the eastern boundary put the GS about 13 nm off West End for a total width of 35 nm.
The GRIB models are presented next:
The top two rows of numbers are the gradient and gust winds from the Euro model. The next two rows are the US model, them Meteo and NAM. Note that the Euro shows a spike in early afternoon. (that spike never happened) Otherwise, the winds were predicted to be light out of the west.
Weather Forecasts:
NOAA:
For Friday, as of Wednesday:
West winds around 5 knots becoming south southwest in the afternoon. Seas around 2 feet. Period 5 seconds. Intracoastal waters a light chop. Slight chance of showers in the morning, then chance of showers in the afternoon.
For Friday as of Thursday:
Northwest winds 5 to 10 knots becoming west southwest in the afternoon. Seas around 2 feet. Period 5 seconds. Intracoastal waters a light chop. Slight chance of showers.
For Friday as of Friday Morning: No change from Thursday
Chris Parker: Northern Crossing Route
For Friday as of Wednesday:
NW-N up to 15k/2-5′ Fri15
For Friday as of Thursday:
L&V<WNW-NW up to 15k/1-4′ tonight-Fri15;
The actual weather and sea conditions were pretty close to both NOAA and Chris Parker as of the day before. As you will see in the table coming up, we started off with NW winds in the 15 knot range, with some gusts to 20. The winds died off during the morning and did back to the SW just as we arrived in early afternoon. Note that these human-generated forecasts were closer to actual than the GRIBS.
Sea conditions started out a bit higher than NOAA predicted. We saw 3′-4′ for a while, before calming to 2′ then 1′. Chris got it right with his 1′-4′ forecast, but I suspect he was thinking the 4′ would have come at the end of the trip and not the beginning.
To build the table below I recorded course, heading, speed through the water and ground speed every half hour. After the trip I calculated current set and drift from that data. Despite my attempts at accuracy, I had to estimate speed through the water based on RPM settings and experience with the boat. Our paddle wheels work but are not always accurate. And the calculator I used did not carry current drift speed to the first decimal so that too is a bit rough.
Lets first show a picture of the gulf stream with our course plotted. The blue arrows represent the strength and direction of the current. Note it was strongest near the west wall and nearly on our nose near the east wall. The numbered arrows, 1 through 5, relate to points on the table, below.
At an average speed through the water of 8.2 knots and a speed over ground of 7.8 knots, we made much better time than planned. As you can see from the table, the current speed was much below predicted, and the lowest I have ever experienced. One note, with light winds from generally astern, we might have picked up a tenth or two of STW do to the reduced wind drag. That might increase the current drift respectively so I entered some of the current speeds as a range.
In the continuing discussion about the best ways to navigate the GS, in this case the straight shot approach was the best.
We are in Stuart, FL and will move down to Lake Worth tomorrow to stage for crossing the Gulf Stream on Friday, headed to the Bahamas. This will be crossing number ten or so, but I really do not feel I have a tactic to make the best of the Stream’s current.
(edit: I first released this on 9 Nov. just after the crossing. Then realized maybe doing math after no sleep is not a good idea. So if you read this yesterday and had some issues with the math, I apologize. Hopefully my edits have corrected the math errors)
On the sailboat we were usually between New England and the Caribbean and the Stream was just a small distraction on a 1500 mile trip. Our tactic was to cross at right angles to the current and just maintain our general heading. The current would set us twenty miles or so, one way or the other. By the time we exited, we would just adjust our heading by a degree or two and keep on sailing. The word was, do not crab into the stream.
But sitting here in Florida, and wanting to go to Bimini, for example, the problem is a little different. I don’t really want to get set twenty miles north on a sixty mile trip. And I don’t want to run all the way down to the Keys, just to get a better angle. So I have been wondering just what is the most efficient method of crossing the Stream.
In central and southern Florida the Stream is about forty miles wide and has current from 2-4 knots, averaging 2.5 knots for the forty miles. A little internet searching will bring up as many tactics as there are sailors, but one can winnow those tactics down to just a few.
The first approach is to go south on the Florida coast in order to improve the angle, say Angelfish Creek to Bimini. This approach does provide a relatively quick, day hop, and if your goal is shortest possible time on the passage, this works. But proponents of this approach tend to forget about the days spent moving down the Florida coast, not to mention the pain of all the bridges if you do the trip inside on the ICW.
For the other approaches, lets use a trip from Lake Worth inlet to West End, Bahamas. About fifty-five miles on a bearing of 95 degrees. The Explorer Charts recommend the classic navigation technique of calculating the required offset for the current and then setting a heading to accommodate that offset. See below.
Assuming one holds a constant heading, this tactic will generate a ground track roughly “S” shaped as the current first increases then decreases during the trip. Not sure why they say, “Dont angle into the current” because that is just what you are doing here.
This approach has lead to some folks suggesting sailing a purposefull “S” track. Proponents of this tactic suggest leaving Lake Worth and turning SE even further than necessary to hold the rhumb line. Getting ahead of the current, so to speak.
Waterway Guide provides headings for various boat speeds and destinations, essentially the same as the Explorer chart Again, holding a constant heading to counteract the average current will produce an “s” curve track.
Then there is what the old salts told me way back in my sailing days. And that was to take up a heading perpendicular to the stream and let the stream set you. Then angle back to your destination once clear. This does have the advantage of getting you out of the stream the fastest, even if it does drop you off fifteen miles north of your desired track.
Finally, there is the approach of letting the autopilot carry you across the stream exactly on the rhumb line. This approach will get you across the stream in the shortest distance, as compared to the previous, shortest time approach.
As an aside, think about what would happen if boat speed were four knots and if the Stream were four knots. If you let the autopilot drive, you will stand still, with speed through the water of four knots and speed made good of ZERO! On the other hand, if you took the perpendicular heading approach, you would still get across the stream, albeit far from your intended destination.
So what is the modern mariner to do! Time for a little math. Using a speed through the water of 8 knots and a current of 2.5 knots for a distance of 40 miles, I would estimate about 12 miles of drift north. That would require an 11 degree offset and reduce my velocity made good (VMG) by one knot. That would increase my time by 45 minutes or so.
Alternatively, I could cross at 8 knots by not crabbing, then come back south 12 miles, which would increase my time by 1.5 hrs.
In the final analysis, all of the options fall in between those two options; the “S” curve approach is a compromise between the shortest time and shortest distance approach.
I plotted a course across the Gulf Stream and estimated the speed of the current every 5 miles. Here is my slice across the Stream, crossing from Lake Worth to West End
And then I built a table of headings and speeds for each 5 mile segment of the trip.
NM
0
5
10
15
20
25
30
35
40
Leg
0
5
5
5
5
5
5
5
5
Knots
1
2
2.5
3
2
2
2
1.5
1
Course
95
95
95
95
95
95
95
95
95
STW
8
8
8
8
8
8
8
8
8
HDG
102
109
113
117
109
109
109
106
102
Corr.
7
14
18
22
14
14
14
11
7
SOG
7.7
7.5
7
7
7.5
7.5
7.5
7.5
7.7
Hrs1
0
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
Hrs2
0
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.6
Set (NM)
0
1.3
1.6
1.9
1.3
1.3
1.3
0.9
0.6
If this is accurate, it should take me 5 hours, 45 minutes to cross, versus the 5 hrs with no current. Frankly, I doubt we will do this well.
My plan is to capture our actual headings and ground speeds every 5 miles and compare them to the calculated values.
My hypothesis is that the fastest way to get from point A to point B, in the presence of current is the shortest distance route, not the shortest time route. And that means running the rhumb line. The easiest way to run the rhumb line is to let the autopilot drive the boat, so that is what I will do.
Note that our ultimate destination is Palm Cay on New Providence. So the full question gets a bit more complex. The route we will take, crossing the stream at right angles, results in a total distance to our destination of 201 nm. If we were to go by the most direct route, the course would cross the stream on a course of about 125 degrees, angling into the stream. So the bigger question becomes: is it better to cross at right angles to the stream, even if that takes you out of your way. In this case, 10 miles out of your way.
I will update the blog after our crossing. We will do the 40 mile test run, then turn south and run overnight to Palm Cay marina, so give me a few days to do the data analysis.
Part 2 Trip complete:
Good news is, the trip was great. Stream was smooth and for the rest of the trip the water was like glass. And a 3/4 moon to boot. So how did all the calculations work out? Below is a table that shows the calculated GS current and the actual current derived mathematically from my actual headings and SOG. (speed over ground). (Best I can do to get all the numbers on the chart)
Miles
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
Leg
0.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
Estimated current
1.0
2.0
2.5
3.0
2.0
2.0
2.0
1.5
1.0
Course
95.0
95.0
95.0
95.0
95.0
95.0
95.0
95.0
95.0
Speed Through Water
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
Estimated Heading
102.0
109.0
113.0
117.0
109.0
109.0
109.0
106.0
102.0
Estimated Speed Over Ground
7.7
7.5
7.0
7.0
7.5
7.5
7.5
7.5
7.7
Set
0.0
1.3
1.6
1.9
1.3
1.3
1.3
0.9
0.6
Actual Heading
110.0
116.0
115.0
116.0
109.0
103.0
102.0
101.0
100.0
Actual Speed Over Ground
7.7
7.5
7.5
7.5
7.7
7.8
7.8
7.9
7.9
Actual Current
2.0
2.7
2.6
2.7
2.0
1.5
1.0
1.0
1.0
As you can see, the actual current in the Stream was a bit less than calculated, so our speeds were better than planned. With a no-current crossing time of 5 hours, and an estimated crossing time of 5hours, 45 minutes with the calculated current, our crossing time was 5 hours, 25 minutes. Our total time for the 201 mile trip was 25.5 hours, exactly as predicted, which means we were slower than 8 knots for the portion of the trip once we exited the stream.
Now lets compare this to the direct route. The no-current total time for the 191 miles at 8 knots would be 23.9 hours. At a 125 degree heading across the stream we would be in the Stream for 50 miles at an average speed of 6 knots. At that speed and distance it would take us 8.33 hours to cross the stream. We would complete the remaining 141 miles at 8 knots in 17.6 hrs for a total time of 26 hrs, all that using the original calculated estimate of Gulf Stream current.
However!! The stream was not as strong as estimated. If we use the actual stream data calculated from our actual trip, then the direct route crossing would not have taken 8.33 hours but just 7.8 hours. And that would result in a total trip time of 25.5 hours. Exactly the same!
A brief note on geometry here. As angle A, in this case the angle between a route perpendicular to the current and one angling into the current, increases, the hypotenuse, in this case, the time/distance in the Stream, increases rapidly (geometrically, actually). So in our case, with a direct route angle of 122 degrees and the estimated current of 2.5 knots, it is better to take the perpendicular/shortest distance route. At the actual current of about 2.2 knots, the two trips are identical in time. At this angle, slower current favors the direct route and stronger current favors the perpendicular route. With lesser currents a larger angle of attack can be taken. With stronger current, say the more typical 3 knots, then the perpendicular route will be favored.
And a last note about boat speed. The slower the boat, the more favored the perpendicular route will be – up to a point. If your boat only goes 4 knots, and the current is 4 knots, take a plane. Just kidding, but in this whole experiment we used a constant 8 knots and variable angles of attack and current level. Varying boat speed leads to the same conclusion, the more time you spend in the stream, the larger the penalty.
With all this, my conclusion is: the fastest way to cross the Gulf Stream at typical current levels is to cross at right angles and crab into the current sufficiently to maintain a course perpendicular to the current. And the best way to accomplish that is to use a capable autopilot.
I understand the “S” curve idea, but I believe it is a hold over from when autopilots just held a course. Using old fashioned navigation skills to calculate a heading to counteract set and drift, a constant heading would result in an “S” curve track with the Gulf Stream current. But that does not make an “S” course efficient.
Of course this over-simplifies the problem. If you are trying to make a day hop of it, and want to check in at Bimini or West End, then you might need to pick and choose your departure point. Lake Worth works well for West End, Fort Lauderdale or Miami for Bimini. And the slower your boat the more you might need a lift from the stream to improve your ground speed. And finally, if the weather stinks in the stream, moving south in the meantime might improve things. But I wouldn’t waste time going to Miami for the angle if the weather was decent when I was still in Lake Worth.
Finally, a note on calculating speeds, angles and distances. The easiest and quickest way is with an old fashioned E6B flight computer, nowadays available on a computer or as a phone app. Using the tool you can quickly and easily test various alternatives and choose the best combination of launch point, course, etc. So, try this experiment yourself and see what you find.
Nothing like a cold beer after a Gulf Stream crossing. As crossings go, this one was just fine, if a bit bumpy at first. A different sort of crossing in that we were coming NE from Key West so crossing the stream at about a 30 degree angle. We left Key West at about 8am, figuring on covering the 225 NM at a fast average speed over the ground of 8.6 knots, 1.1 knots above our actual speed through the water, thanks to the boost from the current in the stream.
We originally planned to enter the banks at South Rock with an option to choose a more northerly course to enter the banks above Bimini if we didn’t like the conditions. This would add about 12 miles to the trip but would also give us a bigger boost from the stream.
In my planning I estimated the GS current at 3.5 knots and with our angle of attack figured about a 15 degree offset angle to hold course and a 2 knot resultant boost in speed. As it turned out it was taking closer to a 25 degree offset to hold course and we were not getting quite 2 knots of boost. At a boat speed of 7.5 knots and a steam speed of 3.5 knots it doesn’t require much of a miscalculation to throw off prior estimates. In this case, I figure the angle of attack was closer to 45 degrees so the required crab was much greater. Had I seen that coming I might have entered the stream a bit further south. Maybe next time. Hard to calculate the best approach between taking the shortest route and getting the best speed boost.
With lots of unknowns regarding what was going to happen later that night with winds and sea state we decided to turn north to improve the ride, reduce the crab angle and maybe pick up enough extra speed to absorb the extra 12 miles of total distance. Also, the wind was forecast to clock further to the west in the early morning hours which would put it more astern after our turn east at North Bimini. As it turned out, the wind never clocked and stayed at about 225 degrees throughout the night and morning, but still behind the beem, so just fine.
We can see out 64 miles with our radar so I began tracking and timing the advance of the front i knew was coming some time Saturday morning. At 3am it was moving at about 12 knots so would arrive right on time at noon. With that estimate we planned on standing off until the front passed, possibly testing our new lightning protection system. By 10 am, when we were 2 hours out, the front had slowed to 8 knots and would not arrive until 1pm. Plenty of time for us to get in and tied up before it it. As it turned out, the first slip designated for us didn’t agree with our boarding doors so, with the lightning and thunder near, we moved to a more agreeable slip. We had the last line secured when the front passed over. Lucky timing on our part.
We will stay here a week, with plans for kayaking, bicycle touring, and maybe some fishing and conch hunting. And maybe a few more Kaliks on the beach. Then we will begin moving towards the Exumas. The Exuma Land and Sea Park is expecting us on 1 February for our one month stint as mooring field hosts at Cambridge Cay. After that, no real plans.
We left Great Harbour Cay Marina on Monday, 2 April, anchoring out in order to get an early start for the run to Bimini. We pulled anchor around 6am and crossed over to Bimini, about 75 miles. Wind and a light chop from behind us made for a pleasant trip.
We anchored just southwest of the cruise ship dock in nice white sand. Plenty of boat wakes in the evening, followed by a mild swell overnight that rocked us to sleep. At first light we upped anchor and headed out for West Palm Beach. Almost immediately we picked up a knot or so of GS current. Some Dolphins arrived to play off our bow. Always a good omen for a safe passage.
No idea why they do this. We often see them off in the distance, changing their course when they see us, then coming over to race ahead of our bow. They will do this for 5-10 minutes then go on their way. I hesitate to anthropomorphize, but I assume its some sort of game for them.
About half way across a USCG cutter approached us and called for information about us. They ran along side us for a few minutes while they checked out who we were. Apparently we are good guys so they released us to go on our way.
The GS is much closer to the Florida coast than I am used to. For most of the trip we were getting about a 2.5 kn boost. At our angle I would estimate the current to be about 3.5 kn. About 10 miles off the coast the current increase to over 4 kn and was still at least 2 kn as we approached the sea buoy at Lake Worth inlet.
Our plan was to anchor in the north end of Lake Worth, but the weather was a bit inconvenient, a 270 degree wind shift over night and some rain predicted, so we took a berth at North Palm Beach marina, a favorite of ours. This closes the loop, so to speak for our trip to the Bahamas this year as we left from this marina in December.
Our reentry to the US was painless. We used the CBP Small Vessel Reporting System (SRVS) and although the system requires lots of complicated preparation, it works. We were able to clear in with just a phone call. In past arrivals we have been boarded and inspected, or have been required to report in person. Much better this way.
We will sit here for a day, then run up the ICW to Stuart. In Stuart we have a long list of boat maintenance actions, some warranty related, some just repairs and clean up after being gone for 4 months. Plus a road trip to Tallahassee to watch our son Grayson graduate from FSU.
When we get back to the boat we will prep for the trip back to Newport. This year Lisa will step off and I, plus a few friends, will take the boat back. The goal will be to stay outside as much as possible, weather permitting. We shall see.
You must be logged in to post a comment.