clean/refactor phenology #204
Comments
|
I just realized that this is just a rediscovery of #65. |
bandre-ucar
added a commit
that referenced
this issue
Apr 7, 2017
Merge branch 'newpftparams' Pulled a bunch of PFT parameters, including the ones from @xuchongang and @elias-massoud's sensitivity study, as well as others, into the new FATES parameter file. Parameters pulled out include both PFT-level parameters and scalar parameters. PFT parameters include: * allometric: for DBH:Bleaf, DBH:height, DBH:biomass, SAI * background mortality rates * soil moisture threshold for drought mortality in non-hydrodynamic model * kinetic parameters for vcmax, jmax, and tpu * seed turnover times for germination and decay (and thus, by ratio, seed-recruit survival fraction) Scalar parameters include: * phenology. in the long run these should all be PFT-level, but punting on that for now (#204) * patch/chort fusion tolerances * CWD cellulose/lignin fractions * bb_opt for C3 and C4 plants * base respiration rates for Ryan respiration scheme I also cleaned up the new FATES parameter file, to get rid of all the CLM parameters that were still on it. So at this point only FATES variables should be left there. Fixes: #56, #75, #115, partial #65 User interface changes?: Yes, requires a new fates parameter file with additional parameters. Code review: requesting @xuchongang, @rosiealice, and @rgknox take a look, and also the @rgknox run the rapid science check, since non-b4b. Reasons for this being non b4b are because: 1. moving the parameters from fortran to netcdf isn't expected to be b4b. 2. There's a minor inconsistency, the old code had a number (1.4976) hard-coded in there, which wasn't quite correct for (10.0_r8**c)*m given the default parameters, which resolves to 1.50030644180475. this code here was copied from @xuchongang and @elias-massoud's code, and seems correct, but wouldn't mind someone else checking to make sure this is as its meant to be. 3. The high-temp scaling factors were specified once, but are now calculated inline to make consistent with kinetic parameters, this is the same to like 7 or 8 digits but not b4b. Testing: ckoven: Test suite: ed suite, lawrencium, intel Test baseline: bc571f9 Test namelist changes: new default parameter file Test answer changes: should be roundoff or close to it Test summary: functionality passes, not bit for bit as expected andre: Test suite: ed - yellowstone gnu, intel, pgi hobart nag Test baseline: 2c82568 Test namelist changes: new default parameter file Test answer changes: not bit for bit Test summary: all functionality tests pass. Not bit for bit as expected Test suite: clm_short - yellowstone gnu, intel, pgi Test baseline: clm4_5_12_r195 Test namelist changes: none Test answer changes: bit for bit Test summary: all tests pass, bit for bit.
|
this will be addressed in #862 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
There are a couple issues with respect to phenology that need fixing. I'm just noting this now as a thing to do when we have time.
(1) most of the phenology parameters (effectively everything other than leaf_long, and the flags to use a given routine themselves) are not at the PFT level; thus we can't really have competition by PFT due to differing phenologic traits (other than overall strategy). perhaps not a big deal for NGT, but this ought to be fixed for all other biomes. simplest way is to just put the existing code within a PFT loop and calculate each of these site-level phenology switches separately for each PFT using possibly different parameters for each PFT
(2) there's at least one hard-coded PFT parameter that becomes a bug in the new PFT paradigm: https://github.com/NGEET/ed-clm/blob/master/components/clm/src/ED/biogeochem/EDPhysiologyMod.F90#L485
in which it assumes that the drought-deciduous PFT is number 7. This can (and ought to, at some point soon) be fixed by the above strategy...
The text was updated successfully, but these errors were encountered: