Map

FAQ

How to read the Snow Likelihood Model map and numbers.

Each station's headline number is the peak day within the forecast window (currently the next 4 days), not necessarily today. If conditions are forecast to get colder or wetter later in that window, the peak date shown will be a few days out rather than today – that's the model reporting when the best chance of snow actually falls, not an error.

The percentage is a blended likelihood score: a synoptic rules-based score (thickness, freezing level, wet-bulb proxy) combined 40/60 with multi-model ensemble agreement (UK Met Office, DWD ICON, NOAA GFS, ECMWF). It maps to a category – Very low (<10%), Low (10–29%), Moderate (30–54%), High (55–74%), Very high (75%+).

Clicking a station shows its peak summary plus one figure per forecast day, so you can see the trend across the window rather than just the single best day.

Shape encodes elevation band: a mountain silhouette for summits, a triangle for mid-elevation and upland stations (201–500m), and a diamond for sea level and lowland stations. A pulsing star marks a location you've checked by postcode.

Live forecast pulls real, current data from Open-Meteo for every station. Example scenario is a fixed demo snapshot (dated 13 January 2026) kept around for trying out the interface without waiting on live data.

The station snapshot refreshes automatically in the background roughly every 30 minutes.

Weather forecast skill drops off quickly past a few days, so SLM limits itself to a short, more reliable window rather than projecting further out with false confidence.

Hold Ctrl (or ⌘ on Mac) while scrolling to zoom, or use the +/- controls – this stops an accidental scroll while reading the page from hijacking it. The map is also capped so you can zoom in but never scroll out past the default UK view.

No. SLM is an independent hobby forecast, not an official warning service – it is not a substitute for Met Office, SAIS, or mountain safety advice.

Weather data from Open-Meteo, postcode geocoding from postcodes.io.