CMEPS

CMEPS copied to clipboard

Places some variables in aoflux routine behind cesm-coupling mode control.

fixes #448

@DeniseWorthen I was also discussing these with @jedwards4b and it seems that there are some development in NOAA-EMC side related with this issue that is not pushed to ESCOMP yet. Please see following links,

https://github.com/ESCOMP/CMEPS/blob/3b1e50baab57c739434a4e62937a96a7bea3faf3/mediator/med_phases_aofluxes_mod.F90#L1604

https://github.com/NOAA-EMC/CMEPS/blob/4e19850cb083bc474b7cde5dc2f8506ec74cc442/mediator/med_phases_aofluxes_mod.F90#L1600C1-L1603C14

@DeniseWorthen I think this PR aims to fix those issues and update ESCOMP. Right? I was also looking to this but maybe you already done it. Let me know if you need help.

@uturuncoglu Moving the rainc inside the add_gusts was added by you in https://github.com/ESCOMP/CMEPS/commit/64e1c276dee0de636b03a9bd1186fb794699b02d. But at the top of main at ESCOMP, it is no longer under an add_gusts conditional https://github.com/ESCOMP/CMEPS/blob/3b1e50baab57c739434a4e62937a96a7bea3faf3/mediator/med_phases_aofluxes_mod.F90#L1602-L1605

@DeniseWorthen Okay. It seems that it is removed with other PR. Right? @jedwards4b Do you know why? Was it creating issue with CESM?

@DeniseWorthen @jedwards4b As I remember we are testing all the PR with CESM.

@uturuncoglu I was testing Jim's alarm fix branch and found that it failed for UFS, because of the change in the aoflux that was in main. This may not be the optimal way of fixing aoflux from CESM's perspective, but I can at least run Jim's alarm fix branch if I include these changes.

@uturuncoglu @jedwards4b I was able to test your PR #447 if I also merged in this fix. All tests passed for UFS. Let me know how you want to proceed---if these changes work for CESM or you need a different fix.

@jedwards4b I've tested this PR again w/ UFS and it passes all our tests. Thanks.

Thank you @DeniseWorthen I'm not going to be able to get to this until after the CESM meeting next week, but I'll make it a high priority then.

@jedwards4b I realized there was a much simpler fix for UFS than I initially thought, requiring only a single change in the med_phases_aoflux module. The remaining required changes can go into our UFS fldsExchange.