…ased cocco gQfe_0

…ref for diat

… C lim

…work.

… slightly more competitive

…occos

… for cocco

…ccos

…Kfe and with the blue Tfunc for coccos

…O3 for diat

…itive

… for sp

…arms (too much cocco)

…dal grazing

Conflicts:
	src/Makefile
	src/marbl_config_mod.F90
	src/marbl_diagnostics_mod.F90
	src/marbl_interface_private_types.F90
	src/marbl_mod.F90
	src/marbl_parms.F90
	src/marbl_sizes.F90
	tests/driver_src/Makefile

…CaCO2 dissolution in zooplankton guts increased to 0.7

Modify tfunc

…bug concerning temperature in marbl_mod; changed default settings.yaml to accomodate coccos

Conflicts:
	src/marbl_diagnostics_mod.F90
	src/marbl_mod.F90
	src/marbl_pft_mod.F90
	src/marbl_settings_mod.F90

Includes SPECTRA support

Also, only include C_lim diagnostics for carbon limited diagnostics.

Also, introduce the temp_func_form_opt = 'power' option (which will be
used by coccolithophores). Note that this commit changes the default
Tref behavior -- rather than depending on temp_func_form_opt (Tref = 30
for q_10 and Tref = 25 for arrhenius), the default is currently based on
the PFT shortname and set to 30 for all current PFTs. I have some ideas
on how to change the way we apply dependencies to allow Tref to depend
on the per-autotroph temp_func_form_opt but haven't implemented that
yet.

Tref is still in the PFT settings_type data structures, but values are
derived based on temp_func_form_opt (30 for q_10, 25 for arrhenius, and
0 for power -- note that the power function doesn't use the Tref
parameter)

MARBL_generate_settings_file was still putting Tref in the input files
because it was in the settings file.

cesm2.1 will match the final configuration of MARBL in the CESM 2.1 release
(while future tuning changings will go into settings_latest). The cesm2.1+cocco
file will contain settings for Kristen's coccolithophore runs, but in this
commit is identical to the latest settings.

Something to look at in code review: I added a workaround to
marbl_settings_mod.F90 to set PFT_defaults = 'user-specified' if the namelist
uses 'CESM2.1+cocco'; I can't use 'CESM2' because that will get the autotroph
count wrong, and we intentionally did not hard-code the cocco settings in the
fortran code so this seemed like a necessary step.

Now that we have cesm2.1+cocco settings file, we can set QCaCO3_max to
different values when using coccolithophores instead of needing a
separate settings variable.

intent(out) variable should be last (also added comments defining
dimensions of each argument)

Some default variable settings were missing the comment to see the CESM
NOTE in marbl_settings_mod, others had an incorrect message (claiming
the CESM NOTE was in marbl_pft_mod rather than settings_mod).

Also, changed the order in which the autotroph settings are set so that
it matches the order variables appear in the settings YAML file.

Several small issues were raised in code review on May 22nd, and this
merge fixes most of them

…dency_mod.F90 and switched CaCO3_PROD output from medium to low in diagnostics_lastest.yaml

Following the refactor to move Tref and temp_func_form_iopt onto the autotroph
and zooplankton settings types, the variable Tfunc_ind was no longer an
appropriate name (it wasn't just indexing Tfunc). So Tfunc_ind is now PFT_ind
(matching the "PFT_cnt" dimension in the comments). I also cleaned up the
comments about Tref in the function header to note that the setting is not used
when Tfunc is computed with the power function.

116007e changed diagnostics_latest.yaml but did not change the JSON file

Missing a space between variable name and '==' operator

This will make it easier to add additional valid values for PFT_defaults
in different settings files

Fixed poor spacing around operator in MARBL_share.py and also improved
the in determining proper values of append_to_keys and PFT_keys in
MARBL_settings_file_class.py