martin-esmi wrote:
Interesting wx today in Sweden, none of the models provided in Windy forecasted the showers of snow in the very south of Sweden we got this morning, to the contrary they indicated sunny weather as the best, to some clouds in other models. The TAFs got it wrong as well for ESMS with AMDs twice, only the old and trustworthy meteo.pl got it. Quite unusual.
In the Stockholm area as well. Arlanda (ESSA) had 4 TAF AMD during the past 20 hours. The last METAR before the snow showers began had NOSIG as trend forecast…
Interesting wx today in Sweden, none of the models provided in Windy forecasted the showers of snow in the very south of Sweden we got this morning, to the contrary they indicated sunny weather as the best, to some clouds in other models. The TAFs got it wrong as well for ESMS with AMDs twice, only the old and trustworthy meteo.pl got it. Quite unusual.
What is the most accurate weather model for Eastern Europe – nems, gfs, ecmwf or icon?
Noe wrote:
Can’t seem to find things like Couldbase (which windity has)
Great site. I assume the time scale is in UTC. Correct?
I read somewhere the US Air Force started using the European model as more accurate. Unclear how reliable as google source but the google studies seem to agree
Yesterday and today another opportunity to compare weather models, (using windity). ECMWF and GFS in particular regarding lower and middle clouds on a trip to Colmar (LFGA) back and forth. Both models were quite accurate for low clouds (defined as upto 6400 ft MSL). But GFS did not forecast any middle clouds at all, while ECMWF did, and this turned out to be true. Together with the Meteogram showing a vertical cloud profile for various locations, it was all quite spot on. We all know that this does not always work out. Maybe start giving ECMWF a bit more credit?
I think there is something wrong (inconsistency) with some of the data. In the following 2 images I have selected NEMS model for a specific location. The map shows 17kt at 21:00 and if I click on “forecast for this location” then we can see that the table shows only 9kt (still NEMS). The rest of the models show no more than 10kt on the map, which is fairly consistent with the tabular data.
I think it’s Windy.com (used to be called windyty)
What is that App called Noe? It’s not obvious (to me anyway) from the previous posts…
