Every number on this site comes from somewhere, and this page says where, when it last updated, and — for the forecast — how far it is from the instruments that measure the same thing. It also says what is missing, because a source list that only shows what works is not worth much. Back to the map.
Issued by Environment and Climate Change Canada. These are official alerts, not something this site computed, and they appear at the top of every affected location page.
| Warning | Applies to | Zone | Issued |
|---|---|---|---|
| Strong wind warning | Juan de Fuca Strait - east entrance | Juan de Fuca Strait | Thursday September 17, 2026 at 10:30 PDT |
| Strong wind warning | Juan de Fuca Strait - central strait | Juan de Fuca Strait | Thursday September 17, 2026 at 10:30 PDT |
Numerical weather prediction. Not an observation — this is a model’s best guess, and the Measured section below states how far it is from the instruments today.
| Feed | What it provides | Source | Licence | Refresh | Last update | Status |
|---|---|---|---|---|---|---|
station_weather.json899 KB |
Hourly wind, gusts, temperature, pressure at 57 locations | Open-Meteo (GEM) | CC BY 4.0 | every 3 h | 1.6 h ago | current |
station_marine.json829 KB |
Hourly wave height, period, direction at 57 locations | Open-Meteo marine | CC BY 4.0 | every 3 h | 1.6 h ago | current |
forecast_7day.json67 KB |
7-day daily outlook per location | Open-Meteo (GEM) | CC BY 4.0 | every 3 h | 1.6 h ago | current |
wind_grid.json9869 KB |
40 × 25 wind grid over 48–55°N, 122–134°W for the map animation | Open-Meteo (GEM) | CC BY 4.0 | every 3 h | 1.6 h ago | current |
historical.json279 KB |
365 days of daily wind and wave aggregates, for the Historical panel | Open-Meteo ERA5 archive | CC BY 4.0 | daily | 19.7 h ago | current |
The only real observations on this site. Instruments in the water, reporting hourly. Real-time data with automated quality control only — not the reviewed archive.
| Feed | What it provides | Source | Licence | Refresh | Last update | Status |
|---|---|---|---|---|---|---|
buoys.json246 KB |
Wind, gusts, wave height, pressure, air and sea temperature at 8 buoys, plus a 30-day archive and the forecast-verification record. Canadian buoys are read from ECCC\u2019s own datamart, which publishes about an hour ahead of NDBC and without its rounding; the two US buoys come from NDBC. | Environment and Climate Change Canada / NOAA NDBC | OGL-Canada / US public domain | hourly | 10 min ago | current |
Tides and tidal currents are computed from the motion of the moon and sun, not forecast from weather. They do not go stale the way a forecast does, which is why they refresh once a day rather than every run.
| Feed | What it provides | Source | Licence | Refresh | Last update | Status |
|---|---|---|---|---|---|---|
tides.json1304 KB |
Water level curve and high/low times for 27 CHS stations, ±7 days | CHS IWLS | OGL-Canada | every 20 h | 7.6 h ago | current |
currents.json198 KB |
Slack and maximum-flow times, speed curve and set bearings for 12 passes, plus wind sampled at each pass | CHS IWLS | OGL-Canada | every 20 h | 1.6 h ago | current |
A forecaster\u2019s judgement, not a model run \u2014 and a warning here is an official act by Environment and Climate Change Canada. When ECCC has a warning up and the model beside it reads mild, the warning is the thing that matters. Shown word for word.
| Feed | What it provides | Source | Licence | Refresh | Last update | Status |
|---|---|---|---|---|---|---|
marine.json10 KB |
Worded marine forecast, sub-zone wind, and warnings in effect for 9 zones | ECCC marine weather | OGL-Canada | hourly | 40 min ago | current |
Read from DFO’s own pages. This site is not an official source and openings change in-season — always confirm against DFO before you go.
| Feed | What it provides | Source | Licence | Refresh | Last update | Status |
|---|---|---|---|---|---|---|
regs.json930 KB |
Species limits and closure areas for 17 DFO management areas | Fisheries and Oceans Canada | OGL-Canada | every 20 h | 1.6 h ago | current |
dfo_subareas.json3065 KB |
Pacific Fishery Management subarea polygons for the map overlay | DFO via public ArcGIS mirror | OGL-Canada | every 30 days | 9 days ago | current |
The forecast and the buoy describe the same place at the same hour, so the gap between them is the model’s error, stated rather than assumed. A buoy reading several knots from the forecast is information, not a fault — it is exactly what a forecast cannot tell you.
| Buoy | Forecast | Measured | Difference | Observed |
|---|---|---|---|---|
| Neah Bay 46087 | 5.1 kt | 9.7 kt | model low by 4.6 kt | 40 min ago NDBC |
| New Dungeness 46088 | 1.6 kt | 1.9 kt | model low by 0.3 kt | 1.7 h ago NDBC |
| Sentry Shoal 46131 | 7.2 kt | 7.4 kt | model low by 0.2 kt | 35 min ago ECCC |
| Central Dixon Entrance 46145 | 1.6 kt | 5.2 kt | model low by 3.6 kt | 35 min ago ECCC |
| Halibut Bank 46146 | 15.9 kt | 14.8 kt | model high by 1.1 kt | 35 min ago ECCC |
| La Perouse Bank 46206 | 5.1 kt | 1.9 kt | model high by 3.2 kt | 35 min ago ECCC |
| Southern Georgia Strait 46303 | 13.0 kt | 12.9 kt | model high by 0.1 kt | 20 min ago ECCC |
| Entrance to English Bay 46304 | 12.1 kt | 14.8 kt | model low by 2.7 kt | 20 min ago ECCC |
How the 24-hour record works. Forecast skill cannot be measured from a model’s own past hours — those come from a run that has already seen the observations, which flatters it. Instead the value the model gives for 24 hours ahead is written down before the weather happens, and scored later once the buoy has measured that hour. A prediction is recorded once and never revised, so the record starts empty and fills in over the following days.
Parsing someone else’s page is only as good as the parse. These are the counts behind the regulations and current panels — a row we cannot date is reported, never guessed at.
DFO last edited each area’s page: Area 1 — 2026-09-01 · Area 2 — 2026-09-16 · Area 4 — 2026-09-02 · Area 7 — 2026-09-01 · Area 8 — 2026-09-01 · Area 13 — 2026-09-15 · Area 14 — 2026-09-04 · Area 15 — 2026-09-01 · Area 16 — 2026-09-16 · Area 17 — 2026-09-16 · Area 18 — 2026-09-11 · Area 19 — 2026-09-16 · Area 20 — 2026-09-05 · Area 23 — 2026-09-16 · Area 24 — 2026-09-16 · Area 28 — 2026-09-01 · Area 29 — 2026-09-05.
These are not cached here. Your browser fetches them directly when you turn the layer on, so there is no update time to report — they are as current as the provider makes them.
| Layer | Source | Licence |
|---|---|---|
| Temperature, cloud and radar overlays | ECCC MSC GeoMet WMS | OGL-Canada |
| Seabed depth (NONNA 10 / 100) | Canadian Hydrographic Service | OGL-Canada — not for navigation |
| Navigation marks | OpenSeaMap | ODbL |
| Basemap | CARTO / OpenStreetMap | © OpenStreetMap contributors © CARTO |
| Tide fallback when the cache is unavailable | CHS IWLS API | OGL-Canada |
Coverage is not uniform and pretending otherwise would be the same mistake as inventing the data. Where a location has no nearby source, the relevant panel hides itself rather than borrowing a reading from somewhere unrelated.
Barkley Sound, Graham Island East, Moresby Island, Bamfield, Prince Rupert, Bella Bella, Bella Coola, Queen Charlotte. The nearest buoy to Prince Rupert is 92 km away and to Bella Coola 177 km; wind and waves do not carry that far, so no measured conditions are shown there.
Tidal currents are a feature of passes and narrows, not of open water. Bindings are verified one by one rather than taken from whatever station is nearest: Bamfield’s closest current station is Dodd Narrows, 102 km away on the other side of Vancouver Island, and a slack time borrowed from there would be worse than none.
Week, month and year are all slices of the same ERA5 reanalysis, so the three agree with each other. Reanalysis smooths exposed headlands — measured against the forecast model over identical days it reads about 3 kt lower at Tofino — so treat it as a climatology, not as a record of what a particular day felt like.