Excerpts from comments on "routing" by Brice Pryszo, founder of MaxSea.
As the designer of the first onboard routing system in 1984, I would like to clarify that technically speaking "course optimization" and "routing" are two totally different methods which people very often confuse (calling "routing" what is in fact "course optimization".) Both methods use the speed of the boat as input (generally recorded as “polar curves”), the weather forecast and current predictions (tidal current or ocean current).
“Course optimization” optimizes the course of the vessel along the great circle route. Alternately, for short legs, optimization can be calculated using the vessel's rhumb line: The resulting course will most probably be significantly different than the great circle route.
- “Routing” or “weather routing” enters weather data into the algorithm, and maximizes input to take advantage of weather and/or current forecasts that may take a course a significant distance from the great circle route.
As with most mathematical problems, this one can be solved (or not solved) in several ways, with some methods being better than others. All “routing” solutions are not equal. Accuracy of the forecast is crucial; the quality of the software is equally crucial.
To illustrate the dramatic results of utilizing routing, I will describe two events which occurred at the inception of the routing concept for sail boats. In 1984, when I designed the “isochrones” routing algorithm method, the only existing routing solutions required the use of powerful shore-based computers which basically tested millions of routes to choose the best one. This required power and time then not available on board sail boats. The MaxSea routing algorithm was designed to be executed on a personal computer.
The very first time that I tested the MaxSea routing algorithm was during the trans-Atlantic race “La Route de la Decouverte” with Philippe Jeantot onboard Credit Agricole. Philippe was late arriving at the Canary Islands due to equipment failure. The fleet leaders, 24 hours ahead, were sailing WSW in nice 10-15 Kt trade winds, south of a low pressure system (as any smart sailor would do). Meanwhile, the MaxSea routing algorithm sent Credit Agricole NW, straight into unsettled weather and unstable wind, up to the cold front of a deep low pressure system. The first hours were not welcoming but behind the cold front, a 20 Kt North wind pushed the large catamaran at twice the speed of the remaining fleet fleet. Two days later, Credit Agricole was leading the race by several hundred miles. At that time no experienced navigator would have chosen such a solution!
Another astonishing example of the power of routing occurred when we tried to optimize an Atlantic passage West to East with regular high pressure mid Atlantic. As any good sailor knows, routing should begin by passing around the high pressure using a northerly course. Instead of following the quiet sailing to the east, the routing plunged the course to the south, straight into the low pressure center. Before reaching ineffective wind, the routing jibbed the boat port into an increasing breeze perfect for reaching, making her route much faster than if she had taken the northern route with a downwind slog.
These two incidents had a major effect on sail boat racing results. Today, using routing is absolutely necessary to be competitive in ocean and long distance races. Routing has also proven to be great tool for cruisers because it greatly enhances safety and comfort. Routing can exploit user-defined characteristics and create what we call “cruising” or “safety” polar curves. Utilizing such polar curves the routing will do all it can to find a path which avoids strong wind. This is beneficial because strong wind is generally surrounded by medium wind in which the boat is faster allowing her to “escape”.
Finally I have to say that the weather prediction centers are becoming much more efficient, and increasingly accurate data is now available. Some data is very good for “Longer Offshore Races” such as the Fastnet race. The “Arpege-Aladin” model from the French Meteorological Office takes in account land influences, has a 8 km x 8 km resolution grid and is relatively accurate for 36 or 48 hours. This makes it especially useful for offshore races other than ocean passages. We can expect major improvements in the near future, while organizations such as NOAA continue to work with new technologies.