ctsm5.3.044: Introduce time-evolving LEAFCN_TARGET as function of leafcn parameter#1654
Merged
Merged
Task list completed / task-list-completed
succeeded
May 15, 2025 in 0s
20 / 20 tasks completed
All tasks have been completed
Details
Required Tasks
| Task | Status |
|---|---|
| Add more comments that say what's going on, that leafcn can be time-evolving, but the default is fixed by the parameter file | Completed |
| Add a test that exercises the code for a simple time-evolving case just to show that it continues to work | Completed |
| Submit aux_clm again (results OK now as shown in post above) | Completed |
| Read leafcn_co2_base and leafcn_co2_slope from the paramfiles introduced in #2365 | |
PASS ERS_D_Ld7_Mmpi-serial.1x1_smallvilleIA.IHistClm50BgcCropRs.derecho_intel.clm-decStart1851_noinitial |
|
OK ./run_sys_tests -s aux_clm -c ctsm5.3.033 --skip-generate on derecho (didn't try izumi this time) |
|
PASS ERS_D_Ld7_Mmpi-serial.1x1_smallvilleIA.IHistClm50BgcCropRs.derecho_intel.clm-decStart1851_noinitial |
|
| In the last one I replaced subname with sourcefile. | Completed |
| Commit latest mods on derecho (got stuck when file handles went stale) | Completed |
| Erik's code review parts 1 and 2: | |
| OK ./run_sys_tests -s aux_clm -c ctsm5.3.033 --skip-generate (derecho only, for now) | Completed |
| Erik's code review part 3: add new test pointing to a new custom paramfile | Completed |
| ./create_test SMS_Lm1.f10_f10_mg37.IHistClm60Bgc.derecho_intel.clm-leafcn_t_evolving | Completed |
| ./rimport the new paramfile | Completed |
| Bill's code review | |
OK ./run_sys_tests -s aux_clm -c ctsm5.3.033 --skip-generate ON IZUMI before addressing Bill's last checkbox. |
Completed |
| Open new issue for better handling of co2_ppmv | Completed |
| Unnecessary because latest aux_clm (2025/5/14) did not have this failure: Open new issue about this inconsequential test failure: #1654 (comment) | Completed |
| Update ChangeLog/Sum | Completed |
| ./rimport the new paramfile if not done | Completed |
I think some more thought needs to be given to the placement of the call to time_evolv_leafcn in the driver loop: I think you need to do one of two things (pick one): |
Completed |
| This should probably use a soil filter: currently you are setting leafcn_patch for all patches, even special landunits, which is a minor performance issue as well as giving non-sensical values for special landunits. (However, if you keep this call before the call to dyn subgrid driver, then you will need to use the version of the filters that includes inactive as well as active points.) | Completed |
| If this is kept here, it should be put in a loop over clumps to activate threading. | Completed |
| Updating to ctsm5.3.030 raised numerous conflicts | Completed |
| ./run_sys_tests -s aux_clm -c ctsm5.3.032 --skip-generate | |
| Resolved FAILs such as | |
tests_0318-152518de/SMS_D_Mmpi-serial_Ld5.5x5_amazon.I2000Clm60Bgc.derecho_gnu.clm-HillslopeC |
|
| by allocating leafcn_patch if_first_step in time_evolv_leafcn (allocates again in InitAllocate). | Completed |
| ./run_sys_tests -s aux_clm -c ctsm5.3.033 --skip-generate | |
| OK derecho: resolved long list of restart, initialization, and non-cn failures | |
| OK izumi | Completed |
Loading