Natural Apps › Tides Pro › Where we know the tide is right
Coverage and verification
Tides Pro computes the tide for any coast on the planet. But computing it and having checked it against a country's official authority are two different things. This is the map of the second one: the zones where the prediction has been measured against official tables, against which source, and with what result.
Highlighted in blue, the zones where the engine has been compared with official predictions. Orange dots are the app's curated ports; blue dots, a sample of the official stations it ships with.
Scroll the map sideways to see all of it.
Figures from the actual verification runs, with their source. Where there is no number we can stand behind, there is no number.
Mainland, Balearics and Canary Islands
Median time difference against IHM predictions at 52 mainland ports in June 2026 (4.1 in October); in the Canary Islands, 3.1 minutes at 15 ports (2.8 in October). Method and data in the methodology, in Spanish on tablamareas.es.
Mainland, Azores and Madeira
There is no run against the Portuguese Hydrographic Institute: the figure comes from comparing the engine with reference tide websites. The Azores and Madeira fall outside the verified box — there the app tells you it is using the global model.
Metropolitan and overseas
The broadest verification we have: SHOM's full census, port by port, against the official predictions.
Atlantic, Pacific, St. Lawrence and Arctic
Each station's harmonic constants are derived from CHS's own official prediction, so at those points the app reproduces the agency's table.
See the full verificationContiguous, Alaska, Hawaii, Puerto Rico and the USVI
The best result of the set, and the territory with the most official stations built in. In bays and rivers where the global model struggles, the station correction brings the offset down from 18 to 6.5 minutes.
See the full verificationBaja California, Pacific, Gulf and Caribbean
Let's be blunt here: there is no Mexican agency with a public tide prediction API to measure against. The verification compares the engine with three independent websites and judges it against how much they disagree with each other — which at several ports is wider than the engine's own error. The range covers both measurement runs.
Two routes, and either one is enough: your point falls inside one of the eight zones on the map, or it has an official station within 2.5 km. In the second case the app uses harmonic constants derived from that station, so times and heights match its official table. Anywhere else, the app tells you it is working with the global model — in the port list, at download time and on the tide screen itself.
The median time offset of every high and low water against the official prediction, matching each extreme with its counterpart. It is a clock error, not a height error: near an extreme the curve is nearly flat, so a few minutes of offset usually amount to millimetres at the shore.
The United Kingdom, northern Europe, the Baltic, the non-French Mediterranean, the non-U.S. Caribbean, Central America, the entire southern hemisphere and Asia: the app computes the tide there, but we have no measurement of our own against official tables, and we won't imply otherwise. The boxes are deliberately conservative: staying quiet costs more than warning too often.
harmonic analysis with 34 constituents, running on the iPhone: the tide is computed offline. The whole planet is served through eight downloadable regional packs, and the first one is free.
Sources: official predictions from NOAA CO-OPS (public domain) and from the Canadian CHS (Canada Open Government Licence), SHOM's census and predictions, Instituto Hidrográfico de la Marina tables and reference tide websites. This page is informational: for navigation, always use the official tables.
Tides Pro is on the App Store, free and ad-free. If your beach falls inside one of the eight zones you'll know exactly what it was measured against; if it falls outside, the app will say so.