Limits the definition of cohort dim in ELM restart file

[bishtgautam]

In ELM restart file, the cohort dimension is only defined when FATES is turned on.
This code modification allows ELM to write elm.rh0 when hist_empty_htapes=.true.
is set in user_nl_elm.

Fixes #5694
[BFB]

[bishtgautam]

This PR removes the dimension (cohort) from ELM restart file, except when FATES is turned on. Thus, cprnc will report DIFFs for many files.

[bishtgautam]

@glemieux @rgknox For elm.h0, all the dimension names for FATES have a prefix fates_. For elm.r*, do you want to stick with cohort as the dim name or should it be changed to fates_corhort or something else?

[rljacob]

But will this change the elm.h0 files? The tests ignore the restart files.

[rgknox]

That dimension in the restart file is maintained to handle information at the "cohort" level, but it also holds other data structures that have information density higher than the column level. So.. its not exclusive to cohort stuff. I suppose you could argue that its more of a fates utility dimension.

[peterdschwartz]

@bishtgautam running the tests with this merged to next and didn't encounter any DIFFs so this can be changed to BFB I think.

[bishtgautam]

@peterdschwartz I have updated the description to be BFB.

[glemieux]

@bishtgautam running the tests with this merged to next and didn't encounter any DIFFs so this can be changed to BFB I think.

@bishtgautam @peterdschwartz do you want me to create a PR to this pointing to the latest fates tag and updating the paramfile pointer after NGEET/fates#1041 is integrated? Or should I simply create a new, separate PR since this PR has been tested already? As a reminder, the fates PR fixes the DIFFs that cropped up with #5599 (comment).

[rljacob]

status: @glemieux and @bishtgautam will talk.

[glemieux]

Per our conversation, this PR can go forward. I'll be creating a seperate PR to update to the new fates tag and reinstate some fates debug tests to the land developer tests.