15/01/2022
Weather case study: Thick Fog missed by Open Weather Maps API.
A perfect example of how internet data alone is sometimes insufficient to protect an observatory. It is 2022-01-15T0230, Canterbury UK. Outside, there is thick ground fog, with actual visibility at ground level of less than 30 metres. Above the fog, it is probably clear and OpenWeatherMaps is fooled. No observatory should be open in this weather. OpenWeatherMap API was just wrong about the visibility of 10 Km (or perhaps they measure it at some height above ground) and they were also giving the summary text "clear sky". Perhaps from a satellite, it looks clear, but not from an observer on the ground.
The humidity was high, but not high enough to cause an observatory shutdown. Ambient temperature was almost 2°C above the reported dew point, which we would consider safe to observe.
The TekNet Clear Skies sensor was reporting a Sky temperature of -1.4°C, which on many days would indicate a clear sky. But with an ambient temperature of just 0.9°C, this indicates a relatively "warm" (moisture rich) sky. The saving data here was a synthetic reading, the Ambient-Sky Delta of only -2.31°C, which was enough to generate the needed unsafe state. Ambient-Sky Delta temperatures should be below about -15 °C to be useful for observing and in good clear skies are often -30°C or more.
Therefore, the TekNet Clear Skies sensor has protected the observatory on this occasion, where internet data alone would have exposed the equipment to thick fog and potential moisture damage.
