add LDAS_INCR to AGCM.rc.tmpl for land-atmos coupled das #288
Conversation
gmao-rreichle
added
Contingent - DNA
saraqzhang
removed
the
Contingent - DNA
AGCM.rc.tmpl
Outdated
| # 0 : no (default) | ||
| # 1 : yes | ||
| # --------------------------------------------------- | ||
| LDAS_INCR: @LDAS_INCR |
There was a problem hiding this comment.
@saraqzhang @gmao-rreichle Just a question: should we use gcm_setup to give this a value here? At the moment, AGCM.rc when untemplated will still have @LDAS_INCR after an experiment is set up.
There was a problem hiding this comment.
That is, AGCM.rc when created by gcm_setup. Obviously, fvsetup does its own thing.
There was a problem hiding this comment.
@mathomp4 I see that gcm_setup and fvsetup use the same AGCM.rc.tmpl here. then we need to make #LDAS_ICNR @LDAS_INCR in AGCM.rc.tmpl here. In fvsetup we need to add 'uncomment" if LDAS_INCR is needed. @gmao-rreichle
There was a problem hiding this comment.
@saraqzhang Note: it's possible the GCM just doesn't care about this...so it doesn't matter. I think @sdrabenh is testing that now. If the GCM doesn't even know about LDAS_INCR then this might be moot! I just wanted to make sure the GCM never tries to read this variable.
There was a problem hiding this comment.
Thanks, @mathomp4, for pointing out the gcm_setup connection. I agree that we should have a sensible value (or nothing at all) for LDAS_INCR: after gcm_setup. I'm ok with either solution, ie, comment out LDAS_INCR in AGCM.rc.tmpl and then have fvsetup uncomment it, or having gcm_setup introduce a default of 0.
One thing to consider, though: If someone should want to run the GCM without the DAS but still ingest pre-computed LDAS_INCR data (such as in a replay), commenting out LDAS_INCR may be somewhat less attractive. Or maybe replays are entirely the domain of fvsetup, and nobody would use gcm_setup to do a replay? @saraqzhang, just something to think about in connection with the "replay" PR.
There was a problem hiding this comment.
@gmao-rreichle
No worries. That is why we all do these tests. It's a lot to keep track of. I just tested with it commented out and everything seems happy:
Character Resource Parameter: CATCH_INTERNAL_RESTART_FILE:catch_internal_rst Using parallel NetCDF for file: catch_internal_rst Integer*4 Resource Parameter: LDAS_INCR:0
@saraqzhang would you mind pushing those changes?
There was a problem hiding this comment.
@gmao-rreichle I was not asked to test this PR. But yes, this was at the time I took a week off.
There was a problem hiding this comment.
@sdrabenh, @saraqzhang: I went ahead and committed the change, so should be good for GCM. Sara, I suppose now we'll need to modify fvsetup accordingly in a new GEOSadas PR.
There was a problem hiding this comment.
@sdrabenh, @saraqzhang: With the latest commits, the PR is now trivially 0-diff, and in a silly way it's obsolete. We could conceivably address our needs without any changes to AGCM.rc.tmpl. However, since LDAS_INCR: is an rc variable in the GCM, I think it's best to have the commented-out line in AGCM.rc.tmpl even so, because it provides documentation, so I'd still recommend merging it into develop.
There was a problem hiding this comment.
@gmao-rreichle LDAS_INCR will be needed for future gcm replay from LADAS. it is good to have the addition now.
gmao-rreichle
added
0 diff trivial
fvsetup sets the parameter LDAS_INCR to default 0 , user input can set it to 1 for land-atmos coupled das experiment.