2002 Stamford to Vineyard Race - Posted 9/2/02 This race began at 1500 hours (EDT) on August 30, 2002, from just outside the entrance to Stamford harbor (the "Cows"), east up Long Island Sound past Block Island to the Buzzards Bay Light Tower off Cuttyhunk, leaving the Tower to starboard, then returning west down LI Sound, leaving Block Island to starboard, to Stamford. This takes the fleet through many tidal cycles and a couple days of shifting winds. Routing is critical, and depends on accurate knowledge of the wind and currents. Local Knowledge has developed a detailed current model of the Sound, all the way past Block through Vineyard Sound to Nantucket, and also produces high-resolution wind models. This becomes input to LK's routing software (see Bermuda Race analysis) and also supports thirdparty tactical software such as Deckman, Raytech, Maxsea and others. |
|
| Routing Solution for start at 1600 Aug 30, with Blue Yankee polars. The black vectors show wind at the start, basically from the NE (a beat all the way to the end of LI). The red line shows the optimal (quickest) route, a bit over 16.25 hrs, to the turning point west of Cuttyhunk. Other routes are calculated for comparison (times are color-coded to the route in the 'Best Course' box). The time differences are primarily due to different routes from the start to Orient Pt, as shown, and all solutions prefer the final approach to the south of Block Island. The time differences are for this very fast boat and will be considerably larger for slower boats in the fleet. | |
 |
|
| Recomputed routing solution from predicted position of boat at 2230, Aug 30, with winds lightening and backing to become more northerly. The optimal (red) course to the south of Block Island is 40 minutes faster than the "blue" route more along the rhumbline, and takes advantage of the ebb current through Plum Gut and stronger winds to the south of Block Island. | |
 |
|
| Coming back, assuming arrival at the turning point at 0830 on Aug 31, the boats must return keeping Block Island to starboard, and the routing solution has a slight preference for the southerly route, continuing through Plum Gut back into the Sound. Winds are now stronger, and from behind (jybing conditions), with a predicted time about half of the time elapsed coming out. Overall, for a boat as fast as Blue Yankee, the race could be over in about 25 hours. For slower boats, the differences between the optimal and comparison courses can be significantly larger, as they are more sensitive to differences in current and windspeed. In terms of corrected time, winds will continue to increase throughout the day on Aug 31, and the slower boats will spend a greater percentage of their time in stronger winds than the fast boats. This could help them to correct out ahead. | |
 |
|
| The discussion above is based on our actual "Local Knowledge" wind forecast for the 2002 race, and our current-prediction model for Long Island Sound and Block Island Sound. Below is our pre-race sample analysis with different wind assumptions and more routing examples. | |
| What if the race started August 3, 2002? The
graphic shows our "Force 4" routing solution, assuming a start at 1500 EDT on
8/3, using a 3-day wind-prediction file and J35 polars. Solutions depends on the boat's
performance charactertistic (polars), partly because different boats sail effectively at
different angle to the wind, and partly because with different boat speeds, different
boats reach a given point at different times - and hence see different wind and current
when they get there. [SCROLL TO RIGHT TO SEE COMPLETE GRAPHIC] The optimal (quickest) route is shown in red, with alternative routes in other colors for comparison. Times for each (to turning point at Buzzard's Bay) are shown in the 'Best Course' box at the right (in hh:mm, color-coded to the route). Here the red route is a bit over 3 hours faster than the green route, which hugs the north edge of the Sound. Wind vectors are also shown at the time of the start, running SSW and increasing to the south and east (but see below). |
|
 |
|
| The above solution looks at the entire distance from start to Buzzard's Bay. In practice one recomputes every hour or so from actual position. The alternative routes are originally confined to sectors (say to far right of rhumbline) and then are allowed to find the optimal route from there (this point is adjustable). Below a solution is recomputed from the point where the boat has reached the vicinity of Plum Gut. The optimal (red) solution is the same as above, but now the comparison routes are also computed from the boat position. From here the difference in time isn't as dramatic (up to a half hour out of some 12 hours), but the router definitely prefers not to go through Plum Gut and to stay north of Block Island. | |
 |
|
| Not surprisingly, the solution depends critically on the wind. If we use a constant wind of 10 knots from 210 degrees T (the wind predicted above for BI Sound), the routes look quite different from the above. In steady reaching conditions it is possible to sail pretty much on the rhumbline,the only deviations are due to current. Since the average boat speed is faster (note time of 16 compared to 21 hours), the boat will arrive earlier and hence see different current at any given point. | |
 |
|
| For comparison, note solution obtained when the course is upwind (assumed wind of 10 knots from 60 degrees T). Here the boat is beating all the way and can sail any route without sailing any extra distance. The route variations shown are due solely to current effects. On this day the currents are relatively modest and the effects not large, but 20 minutes is often significant in a race | |
 |
|
| Looking at above solution in more detail from the point where the "boat" (right-most black dot along the red route) has passed through the Race, at 0900 on 8/4. Note how the route tries to pick up an assist from the ebb current (green streamers flowing left to right) without going too far south and sailing extra distance. The controls on the 'Best Course' box allow you to step through the comparison solutions simultaneously and see predicted positions at any time | |
 |
|
| Same spot for prior solution (10 knot winds at 210 degT). The boats reach the Race earlier, at 0600, in flood (red streamers) instead of ebb, and the faster route is pretty much the rhumbline to Buzzard's Bay tower. This minimizes adverse current without sailing extra miles. | |
 |
|
| DIFFERENCES IN WIND FILE RESOLUTION. Public wind
(grib) files are often quite accurate, but do not have the resolution required for a
relative shortly race such as the Vineyard. For example, the free wind gribs on the
navcenter.com site have a resolution (spacing) of 1.25 deg in longitude, and 1 deg in
latitude, which means that only four (4) points surround the area of interest for the
Vineyard race. LK's high-resolution wind models have a spacing of 0.33 deg in longitude
and less than 10 deg in latitude. This is fine enough to pick up local differences in wind
in the Sound, for more accurate routing. FOR DETAILED COMPARISON OF WIND MODELS, CLICK HERE AND, compare the graphic below, with wind vectors generated from LK's high-resolution model, with the next graphic, a public source. |
|
 |
|
| Public wind model. Note that the differences in predicted wind from one point to another are next to inperceptible. Worse, they result from an interpolation over a distance of some 60 nm spacing between data points. In the graphic above, you can clearly see variations over the scale of the chart, with winds shifting in direction and diminishing along the north shore. | |
 |
|
| LARGE SCALE: Local Knowledge wind model shows local variations over scale of the race. | |
 |
|
| Public wind model for overall course. Values just interpolate over large distance. | |
 |
|
VINEYARD RACE PROGRAMS AND DATA "Offchart Racer" routing software (list price $695), Local Knowledge current model and charts for Long Island Sound and Block Island to Nantucket, and high-resolution wind file for the Vineyard Race - PACKAGE PRICE $850 [see more details about OR at Bermuda package] Long Island Sound OR Block Island Regional Curent Models & Charts as ADD-ON: $100 Current File for the Vineyard Race: $250 (Grib format) Wind Grib File (included with purchase of OR format) for the Vineyard Race: $250 |
|
| CONTACT LOCAL KNOWLEDGE FOR MORE DETAILS | goto GOFLOW HOME PAGE |