The inquest · 78 winters of ARSO's notebook
Who killed the fog?
In the 1950s, the observer at Bežigrad logged fog on 94 days a season. In the 2020s, on 48. Between 1950 and 1990 Ljubljana lost half its fog, and then it stopped losing it. Somebody did this. Or several somebodies.
The evidence
Fog days per season, 1948/49 to 2025/26
- It halved. Until 1990 the season lost 11.7 fog days per decade. Since then it has held at around 48.
- The all-day grey went first. Days with fog and not a minute of sun: 41 a season in the 1950s, 10 now.
- There's more sun. November to January: 116 hours in the 1950s, 180 now.
Ljubljana against the region
How much fog was left.
Fog days, the first span = 100. The region is Brnik, Novo mesto, Murska Sobota, Maribor, Celje, Šmartno pri Slovenj Gradcu and Rateče, each station against its own mean. The spans are split at the steps the test finds. Data: ARSO.
The suspects
There's no single killer.
Most of the fog was taken by something bigger than Ljubljana. Bežigrad added a step of its own, in 1982/83, and it looks more like a change at the station than a growing city.
- Something bigger than Ljubljana guilty of most of it
From the 1950s and 60s to 2007–17, seven ARSO stations outside Ljubljana together lost 35% of their fog days, and Bežigrad 43%. The fog went nearly everywhere, slowly and steadily. Kredarica, in cloud 93 days a season at 2,513 m, didn't change: whatever changed, changed below the peaks.
- The city a suspect, not proven
Town stations log less fog than country ones: over 27 winters, Maribor's town station logged half the fog days of its airport, 7 km away. But Bežigrad's summer nights didn't warm against Rateče's and Kredarica's from the 1960s to the 1980s, while the fog was leaving. If Ljubljana did it, it wasn't obviously through its heat.
- Warmer winters acquitted
The winters only warmed after the mid-1980s, when most of the fog had gone: October to March was 3.3 °C in the 1960s and 3.4 °C in the 1980s, while fog days went from 96 to 60. When the winters did warm, by 1.9 °C from the 1990s to the 2020s, the mist went, and fog days held at 44–54. Wrong timing for a killer.
- The way it's counted a suspect, for the step
In the ten winters before 1982/83, Bežigrad logged 71 fog days a season, and in the ten after, 54. The temperature was 3.7 °C both times, and the morning humidity 91%. Against the region, Bežigrad lost a third of its fog in one winter and only got it back in 2007/08. A step like that usually means a change at the station: the observers, the landmarks they judge a kilometre by, the observing hours, or a new building. The archive doesn't say which.
- Cleaner air the archive can't say
Coal stoves put out soot and sulphur, which droplets formed on. None of the 141 variables in ARSO's archive is about the air. What the archive does show: in the 1960s and 70s the fog thinned more than it left, with 24 fewer fog days and 19 more days of mist. That's the signature of cleaner air: fewer nuclei, fewer and bigger droplets, better visibility. Or of an observer who started judging a kilometre differently.
The first report. What would settle it: ARSO's history of the Bežigrad station around 1982 and 2007, and smoke or SO2 measurements in Ljubljana from the 1960s on. Correlation isn't guilt. inquest.json