htop_patch is being incorrectly used in FATES compsets

[adrifoster]

Brief summary of bug

htop_patch is being incorrectly used in FATES compsets, especially (only?) when FATES SP-mode is on. This is because the htop_patch data structure is indexed by CLM PFTS whereas in FATES runs we should be using a data structure that is indexed by FATES PFTS.

see here for an example.

General bug information

CTSM version you are using: ctsm5.3.020

Does this bug cause significantly incorrect results in the model's science? Yes
Configurations affected: FATES, especially FATES-SP

[adrifoster]

This same logic applies to tlai_patch and tsai_patch

[adrifoster]

So the effect of this bug is seemingly far-reaching, when we look at tlai_patch and tsai_patch.

I am not sure we want a quick fix for this. I think perhaps we should try to implement the long-term solution first, in that we modify the SatellitePhenology module to use a separate data structure and then have FATES just use that for read-in and still update the htop_patch, tlai_patch, and tsai_patch as we normally do for any other kind of run.