Adding new ICON-D2 model
-
Hi @ivo.
Have you considered adding the new high resolution weather model ICON-D2 for central Europe? It’s the replacement for the COSMO-D2 model.slightly bigger coverage then the COSMO, 2.1km grid, 1h steps, 27h, updated every 3h
See news here:
https://www.dwd.de/EN/specialusers/aviation/teaser/news/39_cosmo_icon_node.htmlData is here:
https://opendata.dwd.de/weather/nwp/Would love to see it on windy.com
Cheers.
-
@justinlimbo seconded - it is the best short-term hires model around and I really recommend adopting it, it would be a major USP
I've been using COSMO-D2 for ground condition forecasts before flights for years and it has been the most spot-on of the whole lot
Unfortunately so far it's been available only via PC-met or the occasional off-beat weather website
NB for aviation folk: the DWD models sport much higher vertical resolution internally (65 levels) than just the standard pressure levels
Michael
-
@mhaberler Thank you for suggestion, we will put it into consideration.
Korina
-
@korina It turns out Kachelmann is on the trail already:
-
@justinlimbo Model looks really good.
We will implement it
-
@ivo 😇
Awesome! Can‘t wait! 👍 -
@ivo Are you planning to add any other global models after ICON, maybe Canadian or Great Britain
-
@ivo Unified Model or GEM model
-
@przemo1986 We of course plan to add more models in the future.
-
@ivo Superb - very good move!
-
@ivo if I may add a suggestion:
The recent ICON/ICON-xx come in several distribution forms, two of which are:
- standard pressure levels (1000hPa, 950/925 etc )
- "half height levels" (= 65 layers which seem to be geodaetic heights, not pressure levels)
for the vertically oriented (=aviation folk of various shapes and forms) the higher resolution would be useful - at least in the "VFR layers" (= below 5000m)
it comes at a storage + processing cost though
-
@ivo
what sort of time frame are we talking? from the outside it‘s so hard to judge how much time and effort goes into implementing a new weather model. is it an advantage that you already have the ICON Global and ICON-EU? Or is it a complete new beast you have to tame?
would be cool, to get some insights into your work. cheers. -
Awesome!
-
This post is deleted! -
@ivo said in Adding new ICON-D2 model:
@justinlimbo Model looks really good.
We will implement it
Do you have an ETA on the ICON-D2 implementation? Thanks :-)
-
This post is deleted! -
@flachkurbler There is no certain date yet.
-
@Flachkurbler asked for ETA - ESTIMATED time of arrival. Are you working on it and will you release it in the next couple of weeks/months? Maybe in autumn? Or maybe by the end of the year? Or is that project not really on your focus and you will leave it for the next couple of years?
I guess we don‘t really need a „certain“ date but maybe just a rough idea where we are getting. Something to look forward to. 😄
Actually that also applies to most feature requests in this forum. Nobody will nail you down or give you a hard time if you give us just a ROUGH IDEA what your thoughts on the subject is. But the repetitive answer „we will consider it“ starts to sound like “we don’t really need any feedback anymore”.
That’s just my two cents on that feedback loop on feature requests here on the forum.
Other than that I really enjoy the app and the service it provides. Thanks heaps and keep going! Once ICON-DE is there I will support you with a subscription. 👍 -
@justinlimbo Hi, I understand your point of view, however we do not really want to give a rough ideas about the possible arrival, since it does not serve much purpose. We may have some plans, but they can change along the way. Every feedback is good and counted, this is how we know the true demand for each feature.
-
@justinlimbo DWD just announced that the forecast span of ICON-D2 will be extended to 48hrs (up from 27) by June 23
I think this is significant because it extends the usability of the model to cover a weekend - e.g. check friday evening, forecast will cover all of saturday and much of sunday; before: saturday only
