Despite never being a great believer in short-term weather forecasts, in the last month, I guess due to the frequent end-of-winter alterations (coming from a place with not much of a winter winter) I tended to take an occasional look at the homepage of the Swiss National Department of Meteorology & Climatology (Meteosuisse), mostly while relaxing and aimlessly browsing around while waiting for water to boil on lunch break.

I never really checked with a thermometer, but the report there never stroke me as very accurate. However one thing I noticed is a very odd inconsistency between the forecast for the same day, and the "current weather" (French Temps Actuel) section. Looking at the measures for Zurich in the early afternoon, the daytime temperature and current temperature often differ by over 7 degrees centigrade. The "local weather" (Prévisions par localité) for Zürich usually doesn't agree with either.

Given that both seem to be updated quite frequently, while you'd expect divergence from the forecast for a few days ahead, wouldn't you normally expect the "forecast" for the current weather to be very similiar to the actual one?

Any ideas where the numbers come from?

Just a curious question...

-- Tom

I find the weather forecast here to be most unreliable, TV, Swiss Meteo and even the (expensive) aviation forecasts. In two words ... they suck.

I have found Meteocentrale http://www.meteocentrale.ch/index.php?id=10&L=0 to be a bit more precise, whether it comes to reporting actual current conditions or forecasting the future.

I also don't find the weather report of Meteosuisse too reliable, but then I'm not sure any of the others really are here. I guess I'll look at Meteocentrale as well as Meteosuisse and see which seems more reliable or consistent with regard to current weather.

I wonder where the differences come from though. The only Swiss meteorological source Meteocentrale lists is Meteosuisse... Do they combine data? Which data come from foreign sources? How many different sources of meteorological information can there be in a small country like Switzerland, or the even smaller regions? Other web sites, like wunderground.com, give you yet again different information. Or perhaps the different providers get the same data from the meteorological stations, but in raw detailed form, much more than is published by Meteosuisse, and then all analyze and perform simulations on their own based on that? That may explain the different forecasts, but still not the different values for the present temperature, precipation, wind, etc.

So far weather forecasting still seems to me mostly like trying to tell the future by looking at the sky.

Meteocentrale has their own weather stations, and they claim that they have a lot more stations than Swiss Meteo, hence allowing them to build a better forecasting model.

If it goes so far as the German Swiss TV (SF) and Italian Swiss TV (RSI) having entirely different forecasts in the same evening news......

I have often found www.wetteronline.de to be quite reliable (menu is in German though).

I agree that Swiss meteorology still has some way to go to catch up with the rest of the world...

I think that today's F1 Grand Prix shows that even with some of the best resources in the world it is still possible to get things wrong, even just a few minutes ahead.

It is because the forecasts are so regularly correct, or very close, that we have come to rely on them.

The actual weather may sometimes seem quite different to a forecast. If analysed afterwards, the variation to cause the change may have been very small in meteorological terms.

If you simply state "the weather tomorrow will be like the weather was today" ... you are already more accurate than Mrs. Boner (weird dyke readhead chick) and Mrs. Baehler (equally weird giraffe chick) or their 2 scary sidekicks on SF Meteo (ex military fighterjock dude and weird nervous twitch dude).

Just to add, geographically CH is not easiest place to forecast the weather, due to influence of mountains, hills, lakes, local fogs, diverting effect of Alps on cloud movement,...

Swiss weather forecasting is based on super-computer simulations obtained from the British Meteo in Reading (England) and the European Center in Darmstadt (Germany). The Reading and Darmstadt forecasts are not congruent. They use several global satellites and then their own computer programs.

Moreover, the Alpine topography causes massive distortions and microclimatic variants. A few angular degrees divergence in predicting wind direction can result in either wet southwest weather, or a mainly southern hot and dry Foehn.

So the Swiss meterologits have a tough job. They puzzle over all the contradicitions to deliver a zonal forecast, e.g. "north of the Alps". Subsequently, local expers reinterpret this to deliver a microclimatic forecast, e.g. for the Rhine valley.

Knowing how tough it is, I am amazed at the relative accuracy, and not distressed at the inaccuracies.

No use shooting the messengers. They only deliver the experts' forecasts.

Hence why they use the grey clouds/sun/possible rain symbol a lot- this way they are never wrong and have all bases covered.

I find the messengers as much of a laughing matter as the forecasts they present.

Yes, as I said, I don't expect weather forecasts to be a lot more accurate than telling the future by looking at the sky. I know the phenomena tend to be quite chaotic so that the accuracy of forecast for every moment depends on good forecast for the time before that, so that the complexity of forecasting rises to great proportions with any further day, and surely the topology and location Switzerland contributes too.

But as you get closer to the present, don't you expect the current weather to have an increasing weight in the forecast? What about the report for the current weather? I wasn't complaining about the inaccuracy of the forecast, but surprised that the present reports from different sources (all frequently updated within the day) of the same authority are so inconsistent.

For example, today early in the afternoon, in Zurich, different sources, just looking at temperature:

MeteoSuisse/Current Weather (Temps Actuel): 18°
MeteoSuisse/General View present (home page): 9°
MeteoSuisse/Local Forecast (Prévision détaillée/Prévisions per localité): 22°

I don't have a Thermometer, but for comparison, the display at tram station Bahnhofquai/Hauptbahnhof at the time said 16°. These are all figures supposedly refering to the current temperature (at around noon, in the shade) in the area of one city, updated within two or three hours of each other). Suppose you cannot even accurately tell the current temperature - wouldn't you still at least expect the same authority (the Federal Department of Meteorology and Climatology) to show similiar (possibly imprecise) notion of it in different views/reports?

Just a pointless rant, of course - I'm not a sailor, surfer, or pilot, so weather reports don't have much practical significance to me day-to-day. But I find it curious, and very, uhmm, un-swiss?

Anyway, I've been looking at the other sites (wunderground/wetteronline/Meteocentrale) at a few occasions in the last days just for fun when I'd have otherwise looked at MeteoSuisse. Seemed more consistent, but I don't really know how accurate.

Cheers,
-- Tom

The "9°" sounds like a low temperature. The high forecast is not really a noon temperature as typically the highest temperature will be recorded in the late afternoon:

http://www.meteoschweiz.admin.ch/web...tationSMA.html

Their weather station is at 556m therefore not really accurately depicting weather at 408m in the city. If the "22°" was the forecast for today it was accurate indeed. If it was for yesterday it was off by 2° and I don't think this is that bad. The "16°" and the "18°" all sound like plausible figures for noon temperature if you look at the curve.

i know its an old thread but accoring to my iphone now its going to snow for the next 6 days. its never been right yet