Gulf Stream currents are an extremely important factor in finding the quickest route to Bermuda, but they have to be balanced against the wind distribution. And, the balance will be different for every boat, partly because different boats sail more effectively in a given range of wind angles, and partly because slower boats will arrive at a given point hours, even days, later, and by that time the wind will be different from the wind experienced by the leading boats.

For this reason, it is pointless to spend time and money on studying the Gulf Stream weeks in advance. The only Stream that counts is the one that exists the days of the race, and you can't apply it to determine a route without knowing actual winds.


Results are now in for the race. While most of the fleet went west of the rhumbline, the winners followed our east route below.

In Division A, Kathleen and Venture went east and held the lead to the end, finishing 1-2.

In Division B, Vision Quest went east and held the early lead, but was overtaken by Triple Lindy which stayed close to the rhumbline. Second went to Babe who left the rhumbline midway in the race to go far to the east.

In Division C, Avalon went east and held the lead till the finish. Swift first took the rhumbline but then east to come in second.

In Division D, Fiona Rois went east and held the early lead. Silhouette first took the rhumbline but then went further east and overtook Fiona, which finished second.

In Division E,  Chase went E and held the lead most of the race, to be overtaken by Cordelia (which went further east) and Bermuda Oyster, which came from behind during the final slow approach to Bermuda.


Below the red streamers show the Gulf Stream the day of the start, June 15, 2007. The main flow in the vicinity of the rhumbline to Bermuda is pretty much due east, ranging to 4 knots, and there are several counter-clockwise eddies further south. This first example routing assumed winds were constant at 10 knots and directed from Hamiton to the start (so the race would be dead upwind). All relevant routes are then within the rhumbline, and no time is lost by sailing any sequence of tacks east and west as you tack your way south. The optimal route is then determined solely by the current distribution.

The LK router is unique, incorporated in all LK tactical programs including the new Advantage Racing Software. It computes not only an optimal (red) route, but several other possible routes for comparison. The comparison routes often provide more insight to the significance of the optimal route, and help you make a more informed strategic decision. In this calculation the optimal route takes advantage of the west edge of the eddy (lower part of image) and reaches Bermuda in a little under 100 hours (these examples assumed Farr 46 polars). The magenta route somewhat to the west also uses the eddy, plus the easterly-flow of the Stream, and takes about 33 minutes longer, and the black route which deviates slightly from the red route about 25 mintues longer. The other comparison routes take up to 7 hours longer.

mbcurrs.gif (62908 bytes)


The routing changes significantly when the computation includes our pre-race wind forecast (high-resolution long-range wind prediction file). The image below shows the start of the calculation at the entrance to Buzzard's Bay (waypoint "3"). The black vectors show predicted wind in the early hours of the race from NNE. The combined effect of wind and current leads to an optimal route that takes a little over 74 hours.

mberm1.gif (98778 bytes)
The graphic below shows an intermediate time point at 0625 UTC on 6/16 (0225 EDT), at which time boats on the various routes are predicted to reach the points shown by the black dots. At this time the predicted winds have substantially reduced. Winds to the east continue to be stronger, and the boat on the red route is already almost 20 miles ahead of the boat on the yellow route (the right-hand column shows distance from the black dot to the destination).
mberm2.gif (96281 bytes)
The graphic below shows a later timepoint (0057 UTC on 6/17), at which time the boats on the various routes are predicted to have reached the main flow of the Gulf Stream. Comparing to the constant-wind computation above, it is striking that the optimal red route (and the easterly green and blue comparison routes, which take almost identical times) does not take advantage of the eddy currents. As shown below, in this case the wind distribution trumps the Gulf Stream in routing.
mberm3.gif (90241 bytes)
The graphic below shows more detail of the Gulf Stream. In our displays the current and wind vectors are not located at the underlying modeling points, and are unrelated to the resolution of the underlying models. The more you zoom in on the display, the more detail you see.
mberm4.gif (50862 bytes)
What drives this routing is the wind. The graphic below shows the wind pattern a few hours later, after the boats have passed through the main flow of the Gulf Stream. The magnitude of the wind is proportional to the length of the wind vectors, and is also indicated by stronger colors. There is a signicant "hole" in the wind to the west of the rhumbline that sharply decreases speed for the westerly (yellow and magenta) routes.
mberm5.gif (62249 bytes)
The final graphic shows the approach to Bermuda. Now the winds have filled in across the region, but by this time the boats on the westerly routes have fallen well behind (60 - 80 miles).  A word on comparison routes: sometimes all of the comparison routes take substantially different times; other times, as in this race, there are a band of adjacent routes which take almost identical times. This allows you the freedom to take advantage of local wind or weather variations, or to stay close to a competitor, by deviating from the optimal route without paying a signficant time penalty. Only LK routing gives this information.
mberm6.gif (101332 bytes)
These computations were performed a few hour prior to the start of the race, estimating when the boats would reach the entrance to Buzzard's  Bay. It is important in all such calculations to use the latest wind predictions available, and also to recompute routing solutions periodically when underway. This is particularly important after the start, because the wind files are least accurate close to shore and may not reliably predict when your boat will reach open ocean (and more predictable winds). If possible, obtain several independent wind files and compare predicted to actual positions to see which is better.