-
Notifications
You must be signed in to change notification settings - Fork 20
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Sync with ESCOMP; fix aux history files for use_float=.true. #124
Sync with ESCOMP; fix aux history files for use_float=.true. #124
Conversation
both =0 is not an error
fix the x case
Add new XML variables for flexible controls of GPU configuration
implement ocean albedos for UFS
Add length to logic format descriptor, needed for nag.
UFS-specific modifications
* don't print messages if nstate is not created
…hist Fix aux history files for use_float=.true.
…content Compute and export heat content associated with glc runoffs
nan was picked up in pslv field on first pass
Looks good to me. My expectation is that the atm_ds2s_docn_dice test is b4b when configured to use aux history rather than mediator history. Can we include that here (+ run on more platforms) or would you prefer to keep that separate? |
Right now, this is no-baseline change but it probably makes sense to combine even if there is a baseline change for that one test. So I'd say go ahead and test. |
For atm_ds2s_docn_dice test switched to auxiliary history output (https://github.com/NickSzapiro-NOAA/ufs-weather-model/tree/aux_cplhist)
Both ways seem to be working |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
No comments
@NickSzapiro-NOAA I thought you had made a comment w/rt the S4 issue that ncrcat commands/processing of ouput would not be required once you switched to aux history. I don't see any changes like that in your feature branch. Am I mis-remembering? |
So by setting |
Yes. Since it's nice to run tests with longer forecast lengths and this is hardcoded in the ufs.configure, I left a generous cushion. If a concern, maybe the most proper is to add another atparse variable |
I think if you make this |
It does. We exercise multiple stream files for cdeps in a few tests, like: Meanwhile, I've lost days trying to script the formatting right for flexible forecast lengths and it still not working properly. There are some subtleties in atparse, quotes, and ESMF config format,... like with
and I ran out of time to get it working in the end. It's also unnecessary since we can just write to one file and avoid "error prone" solutions (at least for me). Maybe I'm just missing a main point...why split up the auxiliary history/input stream files? |
I don't think having the capability for "daily" aux history files is required for us to at least enable them in the test and remove the mediator history file use. But I think for general use, it would be good to not rely on one big file if possible. Could you make an issue on UWM (if you haven't already) to describe switching to aux history? I'm fine combining this feature branch into my UWM PR and closing that issue. |
Sure: ufs-community/ufs-weather-model#2398 Yes, I'm happy to work more on it if there's interest |
OK, now can you make a PR to me on my fork (https://github.com/DeniseWorthen/ufs-weather-model/tree/feature/fixaux)? |
Please let me know if you want me to fill out the PR more fully, @DeniseWorthen While I see use_float=.true. in ESCOMP/CMEPS, can we set use_float=.false. (more precision + b4b)? |
@NickSzapiro-NOAA There is currently no way to optionally configure |
@DeniseWorthen testing is complete on WM PR 2395. Please feel free to merge here. |
Description of changes
Syncs with ESCOMP, bringing back fix for aux history files so that
use_float=.true.
works correctly.Specific notes