A weather site for the least dramatic weather in Europe
I am two months into a sabbatical year, poking at ideas, and one of them has already turned into something I can point at: skaneweather.com.
The tagline is "what the weather will do, for southern Sweden." That do is the whole point. Most weather sites hand you numbers: 14 degrees, 6 m/s from the west, 40% chance of rain. Then they leave the hard part to you, which is working out whether that means anything for your Tuesday. Having spent years on the production side of forecast and climate data, I wanted to try the other half of the job, the part that usually gets skipped: turning the data into an answer.
So the site leads with an outlook that just says Good, Mixed or Poor for the day, weighing rain and wind together, with the ten-day view and a 24-hour rainfall map underneath for anyone who wants to check my working. Around that there are a few more panels:
- Current conditions from live SMHI stations, as cards or a temperature-coloured map.
- Warnings, when there are any, rewritten in plain language with a countdown rather than a colour and a paragraph of hedging.
- Forecast predictability, ranking towns from the easy calls to the ones closer to a coin toss. Coastal Skåne is genuinely harder to forecast than inland, and I thought that was worth showing rather than hiding.
- Climate risk, a decade-out scorecard for heat, cold, drought and the coast.
Everything is open data, pulled live from SMHI and Trafikverket each time you load the page. Nothing is stored, there is no account, and there is no tracking. It runs entirely in the browser.
I picked Skåne partly because I live here and partly because the weather is, day to day, undramatic. Compared with what central Europe has been getting lately, the frequent damaging hail, the flash floods, it is mild, has no real extremes and is mostly grey with the occasional decent day. The real action is confined to a handful of autumn wind and coastal-flooding events a year. That is a feature for a prototype: if the impact-focused approach is worth anything, it should still be useful on the many days when the weather itself has very little to say. If it works here, the same machinery should carry over to other locations, where the weather is anything but undramatic.
It is early and rough. The whole thing can change from one day to another. I would be glad to hear where it is wrong or where the plain-language version is actually less clear than the raw numbers.