I spent my first 10 years as a weather forecaster in Alaska and Europe, which forced me to learn the basics like few forecasters do today. Instead of using wildy powerful computer models, we had to analyze our own surface and upper-air charts from raw data — which came to us by way of fax machines and teletypes that had been in use since World War II.
While it’s hard to beat Sirius XM Weather for ease of use for coastal and canyon use, serious passagemakers usually turn to professional weather routers for help. And even if you’re a skipper who subscribes to every weather service under the sun, with constant feeds flowing through the multifunction displays at your helm, learning the basics can help you better plan for weather.
When I’m planning a cruise, I look at just three things: the National Weather Service surface analysis, the NWS forecast discussion and windy.com.
The surface analysis is the one information stream that an NWS forecaster actually makes or modifies. It’s massaged by hand instead of by algorithms, which today are the basis for virtually everything else. When I was crossing the Atlantic, I had friends text me the positions of lows and highs from this map only. It’s that reliable.
Next is the forecast discussion. If you are going offshore, this is the one item that is absolutely crucial to have. Every regional forecast office issues one, and you’ll find slight differences among different regions of the country. In the Northeast, the discussions are particularity descriptive. Other regions, not so much. The forecasts also vary based on the forecaster doing the writing.
Last is what I consider to be the sexiest source: windy.com. Click on the little symbol above the +/- symbols, and you can choose winds at different altitudes, waves, swells, wave period, currents and more. You can also choose the numerical model. The default is the NWS worldwide model, also known as the Global Forecast System, with 13½-mile resolution. Each data point is that distance, or 12 nautical miles, apart. And across the United States and off the coasts, a 3-mile (5-kilometer) North American version is available. The third model is the European Centre for Medium-Range Weather Forecasts model, at about 5½ miles. It’s also a worldwide model and has a better resolution than the Global Forecast System.
No one model is better than the others at any given time, place or season. One may do particularly well during the summer over New England, but the same model may not do as well in the Southeast at summertime, or with high-pressure areas versus lows. Depending on where you do your cruising, you can test them and see which works best for you. Look at aspects such as the currents layer; it will show you everything from wave heights (which I don’t always find accurate) to winds (which are usually right on the money).
The thing about forecasts is that they’re seldom wrong in all facets. A forecast may have a timing or location error, but usually not both. Look at trends, and figure on timing and location loosely, and you should be able to compensate for these typical forecast shortcomings.
Don’t give much credence to forecasts for days beyond about a week to 10 days because they’re not based in science. And if you’re relying on GRIB (binary) weather files, know that they’re disseminated without any humans involved. (This means no one is checking to see whether the underlying data is right or wrong.)
Last but not least, if you’re in a situation where you have to make a go/no-go decision, use what I call the 50 percent rule for wind direction, wind speed and the resulting sea state. Begin with the assumption that whatever is forecast could be off by 50 percent in either direction. So if I am hoping for winds at 180 degrees at 8 knots or less, then the wind direction, +/- 45 degrees could be from 135 degrees to 225 degrees, and I have to take that into consideration. Likewise, a forecasted 8-knot wind could materialize anywhere between 4 and 12 knots. This means that while I could be OK with 12 knots, if the forecast is actually for 12 knots, it’s a no-go because I assume that 12 could be as much as 18.
Remember: The most important aspect of using forecasts for route planning is factoring in information you trust to make sound cruising decisions before you set out.
This article originally appeared in the April 2018 issue.