New model on Windy: ICON vs. COSMO-EU vs. COSMO-DE
-
Cosmo-EU is no longer available sice 1.12.2016, see https://www.dwd.de/DE/leistungen/nwv_cosmo_eu_modelldokumentation/nwv_cosmo_eu_modelldokumentation.html (unfourtunally german language)
it was replaced with ICON-EU https://www.dwd.de/EN/ourservices/nwp_forecast_data/nwp_forecast_data.html
-
@tomber42 Thomas it seems that ICON-EU is nested inside ICON, so we can download just one set of GRIB files and we have both. Am I correct?
Are you from DWD? Would be nice to have some insider for help.
-
@ivo no, I'm not a dwd employee. But you can see what you can get here https://www.dwd.de/DE/leistungen/opendata/help/inhalt_allgemein/opendata_content_de_en_pdf.pdf?__blob=publicationFile
And there icon and icon EU are different data sets (with different update intervals)
And, as you can see, the WAM global and Europe Models are included now (bottom of document), don't know exactly what this is but it seems to include forecast about waves and swell, interesting for water sport I think
-
Do you have decided which model you will add first?
-
It will be ICON-EU
-
@ivo
Do you plan to add it to Windy as a X-mas gift? -
I have to say, I am very excited to use the new ICON-EU (Former COSMO-EU). Especially because I am going to Europe on two separate trips next year. I will probably use this a lot over there.
-
So far ECMWF rocks over EU a probably also worldwide. But we will see how ICON will look
-
ICON certainly the best in future ...(isocahedric model!)
thanks ! your job is very interisting and fine ! -
ICON (EU) has a good performance, especially in coastal regions. Would be great to have it on windy.com.
Is the implementation already scheduled?
-
Backed is finished thx to @dzuremar . Now we have to implement it to the client
-
@lrg-mitglied I am not sure that form of a grid is so important for model quality.
-
@ivo
in my opinion, the future of weather modelling is MPAS (Model for Prediction Across Scales) which uses a flexible grid configuration that gradually refines to a high spatial grid resolution in the area of interest (see image for USA) or in an area with complex terain (see the Andes mountains, South America). In such a flexible mesh we save computational effort compared to a global model using the same high spatial grid resolution.
