Live data, last pull 6 Nov, 19:50
How I compute it
Every word on the board takes the same road. Here is one hour at one beach, start to finish, with live numbers.
Choppyscore 6Beren Koyu, Monday 9 Nov 08:00: the best daylight hour of the days ahead.
1 / 8
Where the data comes from
Every 3 hours I pull an hour-by-hour forecast for 98 beaches, 14 days ahead: 32,928 hours each time.
- Waves
- Three wave models through Open-Meteo. Swell from far away and the sea the wind is raising right now come separately.
- Wind
- Two wind models, ECMWF and GFS: mean, gusts, direction.
- Sea floor
- EMODnet depth map: a cross-section from each beach out to sea.
- Coastline
- OpenStreetMap: how open each beach is in every direction.
- Right now
- Nearby airport observations (METAR), every half hour. 28 stations at the moment.
- Logbook
- 9 sessions written by visitors. That is what I tune against.
2 / 8
The point offshore
Models give the wave out at sea, not on the beach. For Beren Koyu that point is 1.5 km off the beach, in 25 m of water.
At the example hour it reads 0.35 m, 2.8 s, coming from NNW.
3 / 8
Two separate waves
The wave offshore is usually two things: orderly swell from far away and the short, messy sea the wind is raising right now. I carry each to the beach on its own, because they behave differently in shallow water.
Swell: 0.32 m · 3.0 s · from NNW
Wind sea: 0.06 m · 1.0 s · from NNE
The lines run at their real periods: each crest takes as long to pass as the wave at sea.
4 / 8
Into shallow water
As the floor rises the wave slows, stands taller and turns toward the beach. It breaks once the water is about 1.3 times as deep as the wave is high.
In the example it breaks at 0.37 m (ankle high): about 10 m off the beach, in 0.5 m of water. Spilling, mushy.
The wave meets the beach at 23 degrees; the wider the angle, the less power reaches the sand.
Power reaching the beach: 0.12 kW for every metre of crest.
5 / 8
Is the beach open that way?
From the beach I draw 360 lines, one in every direction, and measure each until it hits land. Then I look 20 degrees either side of where the wave comes from: if that slice is closed, the wave is hiding behind an island or a cape.
In the example the wave comes from NNW; that slice has 300+ km of open water on average and 100% of it is open.
6 / 8
Wind makes the surface
Wind off the land smooths the face of the wave, wind off the sea breaks it up, cross-shore is in between. If the gusts run far above the mean I trim once more. When the wind is weak its direction hardly matters.
At the example hour the wind is from NE at 4 kt, gusting 6 kt: cross-shore.
Try it: change the direction and the speed.
7 / 8
The score
The power reaching the beach sets the base score. Wind, openness and surface trim it; if the models do not back each other I trim a little more. If everything were perfect the score would be 100.
ChoppyThe period is under 3.2 s: that is chop, not a wave, so Choppy.
Worth a look starts at 30, Go at 55.
The curve is tuned on the logbook and is the same on every sea: very weak waves score little, 5-10 kW/m scores full, very powerful waves drop again.
From power to base score
8 / 8
When I am not sure, I say so
To say Go I look for four things. If one is missing I say Worth a look, whatever the score.
This hour does not score enough for Go anyway.
Height offshore from two separate models: Météo-France 0.34 m, ECMWF 0.36 m. The gap is under the warning line.
The logbook holds 9 sessions. As the gap between forecast and what people wrote adds up, I retune the curve; the logbook trains the board.
- breaking height at least 50 cm0.37 m
- the models agree0.34 / 0.36 m
- no more than 5 days aheadMon 9
- confidence 70% or more84%