CTSM icon indicating copy to clipboard operation
CTSM copied to clipboard

Add some testmods for FATES to test some different FATES parameter file options

Open ekluzek opened this issue 2 years ago • 7 comments

As a first step in moving toward #2126 we want to create some testmod directories with shell_commands that do ncgen on the FATES cdl file and then modify it using the FATES tools in order to test some different switches on the FATES parameter file.

Definition of Done:

  • [x] @glemieux updates #2077
  • [ ] @adrifoster tries running with all the different options on all at once ("kitchen sink" test) and make sure it doesn't blow up
  • [ ] @adrifoster PR for new testmods for all the options
  • [x] @rgknox add testmod to forthcoming two-stream #2265

ekluzek avatar Sep 11 '23 02:09 ekluzek

It would be good to show proof of concept for this by doing a simple PR that just adds one simple testmod with the simplest modification possible.

ekluzek avatar Sep 11 '23 02:09 ekluzek

The upcoming seed dispersal PR https://github.com/ESCOMP/CTSM/pull/2077 will be a good use of this.

Also size and age dependent mortality have undefined values in the default parameter file and don't engage until they've been given reasonable values.

@rgknox would the different model switches in the parameter file (e.g. fates_maintresp_leaf_model, fates_regeneration_model, etc) be good candidates to test as well?

glemieux avatar Sep 13 '23 17:09 glemieux

We have a bunch of module switches that would be good candidates. I'm ordering these in order of complexity, I think its best to get tests going on those that have the most complexity. The regeneration model, especially with the light mortality dynamics is the most complicated.

fates_regeneration_model fates_leaf_photo_tempsens_model fates_allom_fnrt_prof_mode fates_hydro_htftype_node (UPDATE: maybe avoid this one in the first set) fates_rad_model (Note: not on fates main branch yet)

fates_maintresp_leaf_model fates_leaf_stomatal_model

fates_leaf_stomatal_assim_model

It would be nice to test these in the future for longer tests fates_maxcohort fates_maxpatch_primary

rgknox avatar Sep 13 '23 17:09 rgknox

During discussion today @rgknox noted that if we address #1277 we could reduce the time to stress test some of the run modes that need more structure.

Also see https://github.com/NGEET/fates/issues/846 and #1685

glemieux avatar Oct 13 '23 16:10 glemieux

Per discussion today, we agreed that we should try and create a PR to develop testmods of all of the above with the exception of the long term tests (maxpatch) and hydro. @rgknox to add the relevant testmod to the API update that will come along with two-stream.

glemieux avatar Oct 13 '23 16:10 glemieux

@glemieux thanks for the list of tasks from the discussion. To make it easier to see, I'm going to move it to the top and add it as "definition of done".

ekluzek avatar Oct 13 '23 18:10 ekluzek

Item 1 in the definition of done will be complete via https://github.com/ESCOMP/CTSM/pull/2077/commits/2896889184aeb5cca6a7220f116c49af693920b0

glemieux avatar Oct 17 '23 21:10 glemieux

Two of the four tasks were completed here. The last two will be changed because FATES is going to move logical switches to the namelist. So I'll create an issue for those final two from this one and mark this as done.

ekluzek avatar Dec 05 '24 18:12 ekluzek