Reading a wave forecast: what 0.5 m actually feels like

Every forecast gives you a wave height in metres, and almost nobody knows what the number means. It is not the height of the waves. It is a statistic about them, and the difference is the reason a 0.5 m forecast sometimes produces a flat sea and sometimes produces something that knocks a child over.

The number is an average of the biggest third

Forecasts publish significant wave height: the average height of the highest third of the waves in a sample. Two-thirds of the waves are therefore smaller than the number you were given — and a few are considerably larger.

As a rule of thumb, the largest wave in a long session is somewhere near twice the significant height. A 0.5 m forecast will produce plenty of waves around 0.3 m and the occasional one near a metre. That occasional one is the one that arrives while you are carrying a toddler.

This is not a flaw in the forecast. Waves are not uniform, and any single number describing them has to be a statistic. It just means the figure is a centre of gravity, not a ceiling.

Period matters as much as height

The second number, if your forecast shows it, is the period: the seconds between one wave and the next. It changes the character of the sea completely.

Half a metre of short-period chop from an afternoon sea breeze is more unpleasant to swim in than half a metre of long-period swell, which may barely register until it breaks. If you only look at height, these two days look identical.

The thresholds Garbí uses

Rather than leave this vague, here are the actual numbers this app scores against. They are ours — a judgement about swimming comfort, not an official standard — and they are used consistently everywhere in the product, so a beach card and a notification can never disagree.

The best-day feature, which picks the one day in the coming week worth planning around, is deliberately stricter than all of that. It will not name a day unless the sea is at least 20 °C, the wind is 20 km/h or less, waves are no more than 0.5 m, and no rain is forecast. It also stays quiet when no day clearly beats the others, because naming a best day out of five identical ones is not information.

Where forecasts are weakest

A wave model divides the sea into a grid, and the grid is much coarser than a coastline. It knows about the open water near your beach; it does not know about the headland, the breakwater, the sandbar or the orientation of the bay.

Two beaches a few kilometres apart can share a forecast and have completely different water — one sheltered from the afternoon breeze, the other facing straight into it. A beach angled away from the swell can be calm on a day the model calls rough, and a beach with a steep shore break can throw up an uncomfortable dumping wave on a day the model calls flat.

This is why Garbí prefers a report from a person standing on the beach over the model whenever it has one. A model is a good prior. Somebody looking at the water is evidence.

A practical way to read it

And when the forecast and the flag disagree, the flag wins. The forecast is a model of the sea; the flag is somebody watching it.