ECMWF, the European Centre's forecast
Days ahead, which forecast puts the lows and the fronts in the right place? ECMWF's model, the forecasters' reference, now gives a route of its own.
Three days out, a passage is decided by the large picture: when the front comes through, where the low deepens, how far the high reaches. Get it wrong and the finest route sails straight into the gale it meant to dodge. That is where forecasts differ most, and where a sailor most needs the one forecasters trust.
The European Centre for Medium-Range Weather Forecasts, ECMWF, pooled by more than twenty European states, runs the model forecasters look to first for the days ahead. Saily WR reads it four times a day on its finest open grid, a tenth of a degree, about 11 km, and gives it a route on every passage, beside GFS's.
It forecasts the whole world, every three hours for six days, then every six up to fifteen days ahead from its midnight and midday runs; those of the early morning and the evening stop at six days. Its strength is the medium range: the position and depth of the lows, the extent of a high, the timing of a front several days ahead.
Say you wait in Falmouth for a window to cross the Bay of Biscay to Bayona. GFS brings a low across the bay on the third day, ECMWF a day later and further north, and their routes part: one says leave tonight, the other that tomorrow is soon enough. Where they agree, the window is real; where they part, a careful skipper leaves the margin to the less forgiving of the two, or waits for the next run.
Fine as it is for a global model, ECMWF only sketches what the coast does to the wind: a sea breeze, the acceleration off a cape, the funnelling between two islands. The European Centre itself warns that it underestimates the strongest gusts in squalls and just offshore, downwind of the land, and the peak winds of tropical cyclones, where the warning centres' advisories come first. Beyond a week, its route, like any other, tells a trend more than a track.
Sources
- Open data, ECMWF
- Known IFS forecasting issues, ECMWF