p3_lookup_table_1.dat-v4.1.2 doesn't actually get downloaded

[mahf708]

Testing on a new machine, p3_lookup_table_1.dat-v4.1.2 doesn't actually get downloaded automatically. I suspect it is not part of the input_data procedure. The absence of this lookup table leads to very confusing segfaults.

[ndkeen]

Yes, this is another example of a file I have to manually copy. Note this also happens in scream repo.

[rljacob]

Why isn't it in the inputdata/atm dir?

[rljacob]

Oh it is. It also needs to be in $CASEDIR/Buildconf/eam.input_data_list

[mahf708]

Any pointers to how to fix this or is there a chance you could assign it to someone?

[bartgol]

This is b/c the file is not in our xml defaults, so our buildnml does not put it in the list of input files for CIME to download. For P3, the listed tables are

      <tables type="array(file)">
        ${DIN_LOC_ROOT}/atm/scream/tables/p3_lookup_table_1.dat-v4.1.1,
        ${DIN_LOC_ROOT}/atm/scream/tables/mu_r_table_vals.dat8,
        ${DIN_LOC_ROOT}/atm/scream/tables/revap_table_vals.dat8,
        ${DIN_LOC_ROOT}/atm/scream/tables/vn_table_vals.dat8,
        ${DIN_LOC_ROOT}/atm/scream/tables/vm_table_vals.dat8
      </tables>

We need to add the v4.1.2 table (or maybe replace the v4.1.1 with v4.1.2 altogether).

[bartgol]

I see here that v4.1.1 is what EAMxx is actually using. I am not sure what brings in the need for v4.1.2...

Edit: ah, this is needed by the f90 version of p3. Then this should be taken care of by EAM, and what I said above about EAMxx's namelist xml file should be disregarded.

Edit 2: notice that 4.1.2 is not present in components/eam/src/physics/p3/scream/micro_p3_interface.F90, so this does not apply to SCREAMv0 compsets either (assuming I understand eam's configuration enough to understand which one between p3/eam and p3/scream is built). I am not sure who uses the file in p3/eam; is p3 used in some "regular" F compsets?

[mahf708]

v4.1.2 is the current default in EAM:

[ac.ngmahfouz@chrlogin1 ~]$ grep micro_p3_tableversion /lcrc/group/e3sm2/ac.wlin/E3SMv3/20231209.v3.LR.piControl-spinup.chrysalis/run/atm.log.469639.240211-234819 
      micro_p3_tableversion:     4.1.2  

[mahf708]

is p3 used in some "regular" F compsets?

P3 should be used in essentially all "science" v3 compsets (F and W)

[bartgol]

Ah, I see. Well, then this is something that needs to be fixed in EAM. Not trying to dodge the bulled, but I'm not really familiar with EAM's build system to understand where to put logic to download this file. I think someone from v3 atmosphere should take care of this.

[rljacob]

Sorry thought this was EAMxx.

[bartgol]

No problem. I also thought it was EAMxx when I saw "p3_lookup_table" in the title ... :)

[PeterCaldwell]

I feel like EAMxx (SCREAMv0 and v1) should also use table version 4.1.2. @hassanbeydoun - do you understand why we don't? I have some vague recollection that this fixes a bug in snow sedimentation or something like that.