multi-element litter, mass checks and diagnostics #524
Conversation
…, element chemical naming convention, and better organ looping.
…lling of phosphorus
…itter flux boundary condition
…updates to intialization as well
… shorter than this commit message
… mass pools. changing create_cohort()
…ruitment, shifted things so that the PRT object initialization is not a cohort object argument.
|
@jkshuman , I think I addressed all your comments, and accepted your suggestions. Let me know if I missed anything. I went into the EDPhysiology code and double checked how the resource_management diagnostics are filled, and I think the refactors maintain what is in master. |
|
I ran some comparison simulations with an updated base branch that is based off of tag 1.29. Things are looking good. In this example, both the test and base are started with inventory at BCI, simulate for 8 years, and have logging, fire, disturbance and recruitment all active. I've also done similar tests with prescribed physiology mode as well. https://drive.google.com/file/d/1QIROC4-NDLlX4veNpMvsPYAXVk49Qhtj/view?usp=sharing Will continue testing, next is restart tests (again), hydro and then a long smoke test with the F10 grid. If anyone wants to get reviews in on this, please do so, hopefully we can get this integrated very soon. |
… leaves and roots. Fixed error tracker on patch resizing.
rgknox
removed
the
PR status: Not Ready
|
f10 simulations are progressing, in year 80: /glade/scratch/rgknox/fates-clm-tests/me-litt-v3-f10_f10_mg37/run Standard regression tests (no baseline comparison) all expected PASS: |
Co-Authored-By: Gregory Lemieux <[email protected]>
|
|
||
| seed_mass = num_dead * store_m * EDPftvarcon_inst%allom_frbstor_repro(pft) | ||
|
|
||
| ! SEED DISTRIBUTION IS BREAKING MASS CONSERVATION RIGHT NOW... |
There was a problem hiding this comment.
is this still the case and if so do we need to keep track of it with an issue?
There was a problem hiding this comment.
yes, I was not able to get tight mass conservation on this, I will make an issue of this once we pull.
Note that we do allow storage mass to convert to seed in certain grasses, it just doesn't disperse and mix at the site level.
|
re-running regression tests, after applying updates from @glemieux |
|
All expected PASS on the re-submission: |
|
110 year f10 simulation completed without issue: /glade/scratch/rgknox/fates-clm-tests/me-litt-v3-f10_f10_mg37 |
Description:
This changeset refactors the core FATES code to cycle multiple elements beyond carbon12. This mostly is represented in three pieces of memory, the litter, diagnostics used in writing history diagnostics and mass balance checking and accounting. These changes have significant impacts on how we process input and output boundary conditions to the model.
This changeset also adds rooting profile to the cohort structure (although I think it may be useful to have a converstation about this and how to reduce memory needs here).
Specifics:
FatesLitterMod.F90
global: element_list()
site_fluxdiags_type
site_massbal_type
This PR should be synchronized with CTSM PR: .
Collaborators:
@ckoven, Bill Riley, @jkshuman, @walkeranthonyp, Fates Modeling Team
Expectation of Answer Changes:
The results in this PR should give reasonably similar results to previous simulations. However, so much code has changed, that I do expect round-off errors to be large enough to prevent B4B comparisons from passing (however, they may).
Checklist:
Test Results:
TBD
CTSM (or) E3SM (specify which) test hash-tag:
CTSM (or) E3SM (specify which) baseline hash-tag:
FATES baseline hash-tag:
Test Output: