-
Notifications
You must be signed in to change notification settings - Fork 307
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
add smyle L83 compset and SSP3-7.0 extension #1614
Conversation
…/main/controlMod.F90 l804 (pointed out by @olyson and @billsacks) ESCMP/CTSM#1191: call mpi_bcast (hist_avgflag_pertape, size(hist_avgflag_pertape), MPI_CHARACTER, 0, mpicom, ier) --> call mpi_bcast (hist_avgflag_pertape, len(hist_avgflag_pertape)*size(hist_avgflag_pertape), MPI_CHARACTER, 0, mpicom, ier) mpi_bcast should be replaced with shr_mpi_bcast in the whole suproutine in the future.
@jedwards4b thanks for putting this together and pointing us to it. I have a question that I think is mainly for @billsacks. This has history file settings in the use-case. Overall we've avoided doing that in CTSM, and put history settings either in build-namelist or in user-mod directories rather than use-cases. I'm not sure that putting them in a use-case is a bad thing to do -- just that we've avoided doing that. So I wanted to hear if anyone knows if there are reasons why not to put them in use-cases? It does potentially have problems if certain configurations turn off some of the history fields that are in the list. But, that would be OK, if the SMYLE use-case isn't going to be used in that situation. I asked the same question of the CAM folks also. I do see that we did do this sort of thing for the use-case stdurbpt_pd.xml, but that's an uncommon case. For CMIP6 history settings we put those in user-mod directories. And we have some user-mod directories for high-frequency output as well. |
We decided to use user mods directories for other history file settings mostly because it makes it easier and less error-prone for users to customize the starting point: with a user mods directory, you have a user_nl_clm starting point that you can then change (adding or removing history fields as needed). This is more awkward and error-prone if you are setting history settings under the covers via a use case or namelist defaults. |
Yes, that makes sense, and I think it would apply here. We could still have a use-case for some of the settings, but a user-mod directory for the history part of this. You are right that is the part that users are the most likely to mess with and change from one case to another. |
I have removed the history file changes to user_mods in https://github.com/jedwards4b/CESM2-Realtime-Forecast/ |
I redid the use case as 20thC_smyle_transient.xml, however to avoid having to make changes to other components |
Hey @jedwards4b I just want to send a "Are you sure"? cautionary note. :-) First it looks to me that we'll want to update the last years for some of the transient datasets being used. And actually I think we want to use the SSP scenario for some of these so that we have data beyond 2015. The second concern is that do we really want to have some data transient for CLM, but fixed 1850 for other components? If that's correct, that's fine. If not we may want to change the compset even though it will require other components to change. It's also good to have our compsets follow the standard patterns, otherwise we end up having problems down the line when we forget about the exception we made. It's just good to get this right, before simulations are run, even if that means a little more work up front. |
…SP5-3.4 now works, the extension that fails because of missing ndep file is now SSP1-2.6, switch SSP5-3.4 for SSP1-2.6 for test of extensions
… don't have the right range available
…ge too short, and remove phys for it as the blocks are identical, this also fixes ESCOMP#1631
…r configurations, and add the other use-cases to the namelist testing
…hould be closer to what's desired since they want to go to 2020
Description of changes
Adds a compset for the smyle L83 model. BWsc1850smyle
Specific notes
Contributors other than yourself, if any:
CTSM Issues Fixed (include github issue #):
Fixes #1197
Fixes #1633
Fixes #1632
Fixes #1631
Fixes #1578
Fixes #1131
Fixes #1093
Fixes #1036
Fixes #738
Are answers expected to change (and if so in what way)? SSP3-7.0, SSP2-4.5 and SSP5-3.4 because of ndep files
Except for when using the new smyle use-cases
Spinup cases (such as the SSP tests) change answers because of #738
Any User Interface Changes (namelist or namelist defaults changes)? Addition of new smyle use-case
Testing performed, if any: regular testing will be done.