Forecast verification & evidence
Explore cloud inversion case studies comparing saved forecasts with dated photographs and reports. See what matched, what changed and what the evidence can’t confirm.
The short answer
Regional verification means a forecast saved before the event covered the later observation in the relevant area and time window. It does not mean every summit or valley inside that area saw an inversion, or that one forecast point carried the same confidence as the wider map.
The current archive contains selected positive case studies. It shows that useful prior warnings occurred; it is not an overall WeatherDuck accuracy rate.
4 published reports
Newest reports firstDates below refer to the inversion
Replay of saved forecasts
Cloud filled the glen at dawn. Replaying forecasts saved before the event placed the observation within the wider “Likely” area.
A replay using today’s method. This shares a weather event with the Torridon and Tap o’ Noth reports.
Read the reportForecast saved about 21 hours ahead
The forecast highlighted dawn around Cadair Idris. Photographs from Tyrrau Mawr and other dated reports support a wider cloud inversion across Eryri.
The outlook weakened in one update. The evidence cannot confirm an exact cloud-top height.
Read the reportForecast saved about 43 hours ahead
The regional forecast pointed to a dawn inversion. A photograph from Beinn Liath Mhòr at 05:57 shows low cloud and mist lying in the glens.
The outlook was weaker nearest the viewpoint. The photograph cannot confirm an exact cloud-top height.
Read the reportForecast saved about 46 hours ahead
The Eastern Highlands forecast highlighted a dawn opportunity. Dated photographs from Tap o’ Noth show a sea of cloud below the hill at sunrise.
The nearest saved forecast was about 16 km away. Its estimated height was not a measured cloud top.
Read the reportFor each report, we compare the forecast saved before the event with dated photographs or accounts from people who were there. We check the place and time, show how the forecast changed, and link to the original evidence.
The report explains whether the match covers the wider area or a particular viewpoint, including anything the evidence can’t confirm. If we revisit saved forecasts using today’s method, we label it as a replay.
Our weather data comes from the Met Office via Open-Meteo. Forecast cloud heights are estimates; photographs can show a sea of cloud, but can’t measure its exact height.
How our cloud inversion forecast worksRead the record fairly
Good evidence includes the boundary of the claim. These limits apply across the archive.
Evidence into action
Evidence cases show how earlier signals behaved. Your decision should still use the latest inversion window, cloud-layer heights, wind and mountain conditions.
