FOR THE LOVE OF IT / CALIFORNIA

Read the forecast. Know what it means.

How PureSurfers combines wave models, surf-face estimates, tides, wind, water temperature and observations—and where the limits are.

Useful numbers. Honest labels. No pretend sensor at every pin.

PureSurfers brings different types of coastal information into one view. It is important to understand which number is modeled, which is measured somewhere else, and which part is a general spot description.

Regional waves versus breaking surf

Regional significant wave height (Hs) describes the sea state represented by a wave model. It is not a direct reading of the face of a breaking wave at your spot. New statewide locations display this regional guidance with an explicit label.

The original forecast-enabled spots use an inherited, uncalibrated transformation of offshore wave data to estimate surf-face height. Local exposure assumptions and wave direction influence that estimate, but detailed reef geometry, changing sandbars and every tide-dependent effect are not dynamically solved.

The original set-height proxy is 1.5 times the estimated face height. It is not a maximum-wave prediction, a guaranteed set size or a safety limit. We keep the two systems distinct even where the interface and number colors are shared.

Models and observation sources

The wave guidance uses NOAA GFS-Wave, ECMWF WAM and Météo-France model families as available through Open-Meteo. These originate separately, but their usual delivery route shares the Open-Meteo gateway. A gateway failure can therefore affect several panels at once. A direct NOAA route for GFS is an alternate source path, not a fourth independent model.

Nearby buoy readings are displayed as observations with their own station and measurement time. They can help you interpret a regional forecast, but the buoy’s exposure and distance from the beach matter. Model spread shows disagreement; it is not a statistically calibrated confidence interval.

Wind, tides and temperature

Wind and gusts come from regional/coastal forecast sampling. Headlands, cliffs, buildings and small coves can create conditions a model does not resolve. Tide curves use a named reference station and the displayed time zone; station predictions can differ from water levels and timing at another location.

Sea-surface temperature may be a regional model estimate or a separate nearby observation. It is not a thermometer attached to your beach. Read the source label, location and time before interpreting a sudden change.

Dates, missing values and refreshes

A retrieval time says when information reached the service; an observation time says when a measurement was taken; a forecast time says the period being predicted. They are not interchangeable. A refreshed page does not turn an old observation into a new measurement.

Where available, the interface shows source time and stale/unavailable states. A failed source should stay missing instead of becoming zero. Sources can fail independently. An alert not arriving is not evidence that conditions are unchanged.

Map pins, difficulty colors and access

Many pins are approximate surf-area references rather than surveyed takeoffs. Some named sections share an area. The catalogue is evolving and is not a claim that every California wave is included or every access point is currently open.

Difficulty colors describe inherited or provisional guide classifications, not a conditions-aware safety assessment. Unrated entries remain neutral. Small displayed waves do not make a location safe. Consult actual conditions, local knowledge, access notices and your own ability.

Cameras and road traffic

Camera links go to their operators’ viewers; a listing is not a promise of working playback or current imagery. The optional TomTom traffic layer describes road conditions, not beach crowding or parking availability. A missing traffic tile is unknown—not a clear road.

Traffic Stability pauses automatic polling while idle or hidden, respects provider/local cooldowns and identifies older snapshots. These protections conserve requests; they do not increase a provider allowance.

Check the source. Help improve the interpretation.

Use a spot’s source information to understand a reading. For a useful discrepancy report, include the spot, selected date/time/model, displayed value and what you observed. One session is valuable feedback, not proof that an entire forecasting method is calibrated.

Report a data or guide issue · Using the atlas: common questions