search

E3SM-Project / v3atm

Fork of E3SM for testing v3 atm changes
Other
0 stars 5 forks source link

Debug run fix #27

Closed keziming closed 1 year ago

keziming commented 1 year ago

This PR is to fix the debug run issue. The code version is before ZM and P3 merge, as well as MAM5-nonBFB-fix: commit d4e95c1cc69e16805c8ca88925f4ec6d8196a401 (origin/NGD_v3atm, NGD_v3atm) Merge: a7906b2a4d fb53798dc0 Author: Wuyin Lin wlin@bnl.gov Date: Tue Sep 20 12:47:59 2022 -0700

Modifications are small and mostly related to chemUCI.

[BFB] NML diff due to changes in chem_mech.in

keziming commented 1 year ago

The restart test (10x64 layout) is BFB 0a8ea0d27c0f10fbb4e5489328370f46 atm_M_1x10_ndays.txt 0a8ea0d27c0f10fbb4e5489328370f46 atm_S_2x5_ndays.txt

tangq commented 1 year ago

The restart test (10x64 layout) is BFB 0a8ea0d27c0f10fbb4e5489328370f46 atm_M_1x10_ndays.txt 0a8ea0d27c0f10fbb4e5489328370f46 atm_S_2x5_ndays.txt

What runs are compared here? I assume it's for the two runs with the same code (with this PR changes).

We need to check if the normal build simulations with vs. without these changes are BFB or not.

keziming commented 1 year ago

The restart test (10x64 layout) is BFB 0a8ea0d27c0f10fbb4e5489328370f46 atm_M_1x10_ndays.txt 0a8ea0d27c0f10fbb4e5489328370f46 atm_S_2x5_ndays.txt

What runs are compared here? I assume it's for the two runs with the same code (with this PR changes).

We need to check if the normal build simulations with vs. without these changes are BFB or not.

@tangq @wlin7 Under 64X10 laylout, the control (before modification) and test (modified) are BFB.

wlin7 commented 1 year ago

Merged to NGD_v3atm. Tests after merging confirmed results are BFB with baselines, except for NML diff in chem_mech.in.