Sailing weather prediction

Sailing weather prediction or Sailing weather routing is the art of weather forecasting and routing for adventurers and competitive sailors.

Weather forecasting for sailing involves several activities such as weather training and coaching, dissemination of data for use in navigation and route planning software, race modeling which involves historical weather and sea state analysis for yacht and sail design, trip and adventure planning for distance races and record attempts, monitoring for departure and trip weather windows. It involves several type of events such as day races, long-distance races, around-the-world-races, and record attempts. It is routinely used in races such as Volvo Ocean Race, America's Cup campaigns, and olympic classes regattas.

Long-distance races

Weather forecasting for long distance races is based on dissemination of meteorological data, most often in GRIB format, for use in navigation and route planning software and yacht characteristics (polars), providing guidance, as well as analysis of historical weather and sea state data.

GRIB

GRIB (GRIdded Binary) is a concise data format commonly used in meteorology to disseminate forecast weather data. For sailing purposes the GRIBs are transmitted and received at sea. These GRIBs contain only small subset of surface data, usually winds (direction and wind speed), information about wave strength (proportional to significant wave height) and direction, surface pressure. The data is further reduced by providing its subset around the position of a yacht. The data is transmitted over satellite phones and single side band radios.

Software for optimal weather routing

Modern sailing weather forecasting involves transmission of weather forecasts which are used in on-board software which simulates optimal (and safest) routing in distance races. The data is often transmitted in form of GRIB files or similar which are customized for specific areas. These files are suitable for use in popular routing and tactical racing software such as Expedition, MaxSea TimeZero, Adrena, Deckman for Windows, Raytech Navigator/Sail Racer, Sailplanner, SailFast or OpenCPN.

Weather routing applications

There are several applications that calculate a yacht's point to point route based on weather prediction and polars (the boat's performance specs):

Olympic sailing

Weather forecasting for olympic class sailing is a form of nowcasting predicting weather and currents in approximately 0–6 hours timeframe. Even though understanding of synoptic weather conditions is of importance but mesoscale and local scale events take precedence. The forecast includes predictions of the sea-breeze onset, turbulent winds shifts, coastal jets, changes in tidal currents, fog, as well as wind acceleration and directional changes associated with clouds.

Yacht design

Weather routing has been used as a simulation tool to assist yacht designers to produce the fastest design for the anticipated conditions of the event, in conjunction with VPP software to predict the boat's performance.

The key issue here is that the designer is interested in the outcome using a wide range of likely weather conditions, whereas the navigator on board a yacht is just interested in using the latest weather forecast. Initial attempts in this area used the mean wind speed and direction data from sources such as Pilot Charts, but this is not truly representative, and the correct route is to run the weather routing a large number of times from a set of historical weather records, and then carry out a statistical analysis of the results.

Two examples of this follow, both developed by Tim Thornton of Smartcom Software (previously TT Designs).

For the 1977 Whitbread Round the World Race, Tracy Edwards' Maiden campaign was hoping to commission a new boat to be built for the event, and went to Tony Castro as the designer. Tony used Tim Thornton to do a comparison of a number of candidate designs, and this was done using 100 different sets of weather conditions for each leg of the race. In the end, the money could not be raised for a new boat, but the weather routing simulation was repeated for the boat she bought, both to optimise the design for the anticipated conditions, and to assist the navigator in developing a strategic plan for the race.

For the 1988 America's Cup, the UK's Blue Arrow Challenge followed two routes: designing a monohull strictly in accordance with the Deed of Gift, or a multihull like the American entry. Tim Thornton was engaged by the monohull team to produce an analysis of the optimal design under the Deed of Gift. To do this, he combined his weather routing simulation software with a model that automatically generated possible hull designs and predicted their performance, thereby allowing a large number of design parameters to be investigated rapidly. This showed that, for the anticipated light airs of San Diego, the best design would be a small, light, and heavily canvassed design like an overgrown 18' skiff.

The Math

This section is not detailed enough to show the statistical tests to determine accuracy and confidence level: for example, with iscochrones or fixed grid points no method to evaluate probability of forecast success is implied or stated. This section seems to be written by non scientists, hence the section name "The Maths". Weather routing is an optimisation problem, whereby it is required to determine the lowest cost (i.e. shortest time) route across a network, where the cost of traversing each link in the network varies with time, as the wind speed and angle changes, changing the boat's speed. This is generally solved using Dynamic Programming.

In most implementations, the boat travels in fixed time steps (e.g. 3 hourly) and at fixed angles (e.g. 5 degree intervals), with distance being the free variable. This results in the creation of isochrones, showing the set of possible positions of the boat at a given time, which are a useful tool for the navigator. However, with this approach it is not easy to compare multiple runs, and so another approach used for design purposes is to navigate across a fixed grid of points, with time being the free variable. This is more amenable to statistical analysis of the ensuing routes.

Bibliography

This article is issued from Wikipedia - version of the 2/14/2016. The text is available under the Creative Commons Attribution/Share Alike but additional terms may apply for the media files.