Mapping between FATES PFTs and CLM PFTs

[robbert02]

Hi all! I have a question regarding the 'mapping' or 'translation' between FATES PFTs from the fates_params_default.nc file and the CLM PFTs from the surface data input file. The PFT indices of FATES' parameter file is different than CLM's PFT indices in the surfdata. For example, in the parameter file pft=0 corresponds to tropical evergreen tree, whereas pft=0 in the surfdata corresponds to not vegetated. Where in the model does this mapping happen? Is there a specific file that translates / matches the FATES pfts with the corresponding CLM pft? And is this hard-coded or based on long name?
Thanks in advance!

 

[slevis]

@afoster or @samrabin
could you answer this one? I realize that I don't know the answer off the top of my head...

 

[samrabin]

My sense from a look at the code is that these are just completely separate things with no mapping at all. CLM has a bunch of places where plant behavior is determined based on the name of the PFT, which we should avoid doing in favor of what FATES does, which is to have everything about a PFT defined as a parameter. A search through the code shows that the fates_pftname variable from the parameter file isn't even read anywhere.

I'm not sure what happens if you try to save a CLM history variable with the pft dimension during a FATES run, but I'd assume any such output to be nonsense. Instead, you'd want to save the equivalent FATES variable with the PF suffix.

 

[robbert02]

Okay, thanks for the answer. I would however like to add a new pft - which I have now tried to do by appending a new pft to both fates_params.nc and the fsurdat file. Do you have any suggestions how to go about this if there is no way for both of these files to communicate with each other?

 

[samrabin]

Have you encountered problems when you tried that? If so, what was your case setup (compset, user_nl_clm, and code version of CTSM and FATES)?

 

[SadieMing]

Hello! I would like to know which parts I should modify if I only want to use CLM-FATES to simulate the growth changes of two PFTs. Can I just modify fates_params.nc? Do I need to modify any other files? (For example, fsurdat. nc) If I want to modify fates_params. nc, do I only need to keep 2 values of fates_cruit_initd and set all others to 0, or do I still have to modify other parts of the parameter file? Thank you all!

 

[slevis]

Pinging @afoster, but given the holiday:
You could start trying the various ideas that you already came up with to see which one works.

I will lock this thread and move my response to the new thread:

How to keep only 1-3 PFTs when running CLM-FATES?

Hello everyone, I would like to know which parts I should modify if I only want to use CLM-FATES to simulate the growth changes of two PFTs. Can I just modify fates_params.nc? Do I need to modify any other files? (For example, fsurdat. nc) If I want to modify fates_params. nc, do I only need to...

bb.cgd.ucar.edu