Barrier 2 — Lamb Holm to Glimps Holm — is the most exposed of the four causeways. Scapa Flow itself is too small a body of water to build a real sea, so a westerly wind is almost never a problem here. The danger comes from the East through South-Southeast, off the open North Sea via Holm Sound — that, stacked on a high spring tide, is what puts water over the road.
Tide, wind and sea state, aligned together — the window where all three forecasts overlap.
| When | Tide | Wind | Gust | From | Wave | Swell | Risk |
|---|
Tide-only — no wind, wave or air-pressure forecast reaches this far out yet. Heights below are astronomical only.
Built in 1944 from rubble, concrete blocks and sunken WWII block ships, the barriers sit permanently above normal high water — this isn't a ford that floods at high tide. But the road surface on Barrier 2 is low enough that when the sea is already running high, a gust off the North Sea can throw a wave clean over the top.
The predictor combines three things: tide height, wind, and actual wave height. Tide is predictable weeks ahead. Wind and waves are only forecast reliably a few days out, and both wind and wave direction matter as much as their size — a gale, or a swell, from the west has little fetch to work with on Scapa Flow, but the same size arriving from the East through South-Southeast comes straight off the North Sea and hits the causeway broadside. Wave height and direction both come from a separate sea-state model, and only wave energy actually arriving from the East–South-Southeast counts toward risk — the same discount applied to wind. Wave period matters too: a long, slow swell carries more energy up a slope than a short, choppy sea of the same height, so a longer period scales the risk up and a shorter one scales it down. Risk compounds: any one factor running high is worth watching; two or more together is when it gets dangerous.
The tide number itself is also adjusted for air pressure. A published tide table only predicts the astronomical tide — the moon and sun's pull, nothing else — but real sea level runs higher than that in low pressure and lower in high pressure: roughly 1cm of rise for every 1 hectopascal below average (the "inverse barometer effect", the same mechanism, alongside wind, behind a storm surge). That's folded into every tide figure on this page; the tide chart above shows both lines so you can see how much difference it's making.
Tide predictions are primarily for Burray Ness, an official UK Admiralty station on the exposed East side of the barriers (~5km from Barrier 2), via UKHO EasyTide. That free feed only reaches about a week ahead, so days beyond it fall back to Kirkwall (~13km away, scraped daily from tide-forecast.com) — each row in the log below is labelled with whichever station it actually came from. Heights are relative to chart datum (approx. Mean Lower Low Water). Wind and air pressure are the ICON-EU model, and waves come from a marine sea-state model, both via Open-Meteo, centred on Barrier 2, refreshed every 30 minutes.