Closed alcoat closed 4 months ago
Thanks for bringing this to my attention. This is very exciting news.
Now that ECMWF is providing two models, I'm thinking Herbie should provide access to both. If you follow me, the update would be something like this...
- Herbie(date, model='ecmwf') # deprecate `model=ecmwf`
+ Herbie(date, model='ifs') # rename the class ecmwf to ifs
+ Herbie(date, model='aifs') # new class for aifs
Related to #284 and #285.
(This will need updates to the docs and tutorial notebooks, but I can handle that after we get the new templates merged.)
Hi @alcoat,
First, please forgive me for not being too experienced working on pull requests with others; I'm still figuring this out.
I made some additional changes based on your. Take a look and see what you think about this approach: https://github.com/alcoat/Herbie/compare/main...blaylockbk:Herbie:update_ecmwf_ifs
I want to make sure you get credit for this pull request, so I think if you agree with my changes, you can pull them into your branch??
Whoo-hoo! Herbie found the data from my additional changes
hold on. One thing I didn't think about is accessing historical data (e.g., when a user asks for data from January 2024); that still needs to be handled. Perhaps an if condition by the date??
Hi @blaylockbk I've updated the ecwmf dataset in order to access archives for ifs model: two dates are used, one for the update of the 0p25 vs 0p4 resolution and the other one the the ifs and aifs directories. So I've added some tests which are testing the old datasets
This is awesome! Thanks.
FYI, I'll work on a Herbie release tonight or tomorrow.
From : https://www.ecmwf.int/en/about/media-centre/news/2024/ecmwf-releases-much-larger-open-dataset "Medium-range data of the Integrated Forecasting System (IFS) are provided at a resolution of 0.25 x 0.25 degrees (28 x 28 km), compared to 0.4 x 0.4 degrees before, and some additional parameters will be added."
So update of file paths