NSST initialization changes

[grantfirl]

Following discussion with @hertneky, this changes the following:

• Change the initial value of the ocean mixed layer depth (xz) to 20.0 m to match UFS
• Change ifd or the flag to turn on NSST to 1.0 if the surface is ocean; this only has an effect if prescribed surface fluxes are off and NSST is in the SDF and nstf_name(1) > 0.
  -- if ifd starts as zero, NSST code can eventually set it to 1 if the model time is after 6 AM local time and ifd starts as 0, but during some cases, it was noticed that ifd wouldn't change to 1 (and therefore activate NSST) until the first "sunrise" after the model initializes (e.g. if a case started at noon, ifd wouldn't set itself to 1 until the next morning, even though NSST should be active immediately)
• Set the nstf_name(2) value to 1 (NSST spin-up) in every SCM physics namelist to be consistent; the only chance for the SCM to not need spin-up is for a UFS-based case that has all of the NSST variables ready to be read in; NOTE: for UFS cases needing to use NSST and when NSST variables are included in the case data file, the nstf_name(2) should be set to 0 (I don't think that this is documented anywhere. We can also probably make sure this happens programatically in the UFS_case_gen scripts. Added Set nstf_name(2) to 0 for UFS_case_gen cases #577 to address in the future).

New functionality needed for @bluefinweiwei:

• Add do_sst_initialize_only namelist variable (default value of F) to the case configuration namelist; if set to T, scm_state%T_surf (which gets pointed to by the sea_surface_temperature variable) only gets initialized from the case data file SST value. The default/original behavior is that when the surface_forcing_temp variable in the DEPHY-formatted case data file is set to ts (or any of the other options too, for that matter), the scm_state%T_surf gets updated from the case forcing file every timestep and physics has very little "authority" over this variable -- it may get changed during the physics call, but it will get reset when forcing occurs. The new behavior allows physics (including NSST) to change sea_surface_temperature how it wants to.

[scrasmussen]

Changes look good to me! Probably want another reviewer with a better understanding of the physics but from a coding stand point it seems good to me.

[hertneky]

These changes look good to me.