HillslopeFromFile test failing

[ekluzek]

Brief summary of bug

The Hillslope test that tests the "FromFile" options is failing at run-time.

General bug information

CTSM version you are using: ctsm5.4.050-116-g5bc7a0d82 (what will be ctsm5.4.051) undoubtly was before this though

Does this bug cause significantly incorrect results in the model's science? No

Configurations affected: Hillslope tests with FromFile option for hillslope_pft_distribution_method and hillslope_soil_profile_method

Details of bug

This test fails at runtime:

SMS_D_Ld3.f10_f10_mg37.I1850Clm60Bgc.derecho_intel.clm-HillslopeFromFile

Important details of your setup / configuration so we can reproduce the bug

The user_nl_clm changes are:

use_hillslope = .true.
use_hillslope_routing = .true.
downscale_hillslope_meteorology = .false.
hillslope_head_gradient_method = 'Darcy'
hillslope_transmissivity_method = 'LayerSum'
hillslope_pft_distribution_method = 'PftLowlandUpland'
hillslope_soil_profile_method = 'Uniform'

use_ssre = .false.
hillslope_pft_distribution_method = 'FromFile'
hillslope_soil_profile_method = 'FromFile'
hillslope_file = '$DIN_LOC_ROOT/lnd/clm2/testdata/surfdata_10x15_hist_1850_78pfts_c240216.synth_hillslopes_241001.no_nan_fill.nc'

Important output or errors that show the problem

cesm.log:

dec2394.hsn.de.hpc.ucar.edu 126:  mosart decomp info proc =       126 begr =    255151 endr =    257175 numr =      2025
dec2394.hsn.de.hpc.ucar.edu 127:  mosart decomp info proc =       127 begr =    257176 endr =    259200 numr =      2025
dec2394.hsn.de.hpc.ucar.edu 0: PIO2 pio_file.c retry NETCDF
dec2394.hsn.de.hpc.ucar.edu 1: PIO2 pio_file.c retry NETCDF
dec2394.hsn.de.hpc.ucar.edu 109: forrtl: error (65): floating invalid
dec2394.hsn.de.hpc.ucar.edu 109: Image              PC                Routine            Line        Source             
dec2394.hsn.de.hpc.ucar.edu 109: libc.so.6          000015230A51A900  Unknown               Unknown  Unknown
dec2394.hsn.de.hpc.ucar.edu 109: cesm.exe           0000000001F65274  hillslopestreamou         993  HillslopeHydrologyMod.F90
dec2394.hsn.de.hpc.ucar.edu 109: cesm.exe           0000000001F85504  hydrologydrainage         151  HydrologyDrainageMod.F90
dec2394.hsn.de.hpc.ucar.edu 109: cesm.exe           0000000000920B9D  clm_drv                  1148  clm_driver.F90

[ekluzek]

I verified the same fail in ctsm5.4.050 in:

SMS_D_Ld3.f10_f10_mg37.I1850Clm60Bgc.derecho_intel.clm-HillslopeFromFile

[samsrabin]

SE meeting: @ekluzek will ask @swensosc: How important is this?

[swensosc]

can you try this file: /glade/work/swensosc/HillslopeHydrology/InputData/Global/surfdata_10x15_hist_1850_78pfts_c240216.synth_hillslopes_241001.mod_pftndx.nc
?

[slevis-lmwg]

Adding back next. From what I see in our derecho baselines:

• the test fails until ctsm5.4.051
• at ctsm5.4.052 the test stops appearing in baselines because we moved it from aux_clm to the hillslope test suite

I'm happy to confirm that the test still fails and then try Sean's alternate file suggestion:

• The default test FAILs in SMS_D_Ld3.f10_f10_mg37.I1850Clm60Bgc.derecho_intel.clm-HillslopeFromFile.20260825_182109_1nb7ns
• Same directory, test PENDing with updated
  hillslope_file = '/glade/work/swensosc/HillslopeHydrology/InputData/Global/surfdata_10x15_hist_1850_78pfts_c240216.synth_hillslopes_241001.mod_pftndx.nc'

[slevis-lmwg]

@swensosc same error with the file that you recommended. See cesm.log files here:
/glade/derecho/scratch/slevis/SMS_D_Ld3.f10_f10_mg37.I1850Clm60Bgc.derecho_intel.clm-HillslopeFromFile.20260825_182109_1nb7ns/run

Removing "next" as needing follow-up from Sean.