Segmentation fault in wrfgc_convert_state_mod.F90 - wetlscnvfrc_n0

[qiujx20]

Hello,

I am encountering a segmentation fault when running WRF-GC3.0, which appears to indicate an error in line 2468 of the wrfgc_vert_state_mod.f90 program. Could you please provide a solution to this issue?

Thank you very much for your assistance.

[jimmielin]

Hello @qiujx20, could you provide the &chem section of your namelist.input file?

[qiujx20]

Okay, this is &chem , followed by the namelist. input

namelist.input.txt

[jimmielin]

Could you try uncommenting these lines in HISTORY.rc in the run directory?

             #'WetLossConv',
             #'WetLossLS',

[qiujx20]

Thank you very much for your assistance. After modifying History.rc, the previous error has been resolved.

However, a new error has arisen, and I would appreciate your help in resolving it.

[qiujx20]

I tried changing the e_vert value in the namelist.input file from 46 to 40, but I still encountered the same error related to ucx_mod.F90

This issue seems to be related to ucx_mod. Could it be possible that the setting on line 1236 in the F90 program is causing this problem?

[CADiMaria]

Were you able to figure out what was causing this error? I am consistently encountering the exact same error in the full-chemistry simulations:

forrtl: severe (408): fort: (2): Subscript #3 of the array PEDGE has value 43 which is greater than the upper bound of 42

Image              PC                Routine            Line        Source
wrf.exe            000000000557F316  Unknown               Unknown  Unknown
wrf.exe            0000000004C5B3B5  ucx_mod_mp_settle        1236  ucx_mod.F90
wrf.exe            0000000004A28429  sulfate_mod_mp_ch         543  sulfate_mod.F90
wrf.exe            0000000004AC4FCC  chemistry_mod_mp_         538  chemistry_mod.F90
wrf.exe            00000000049720C7  gigc_chunk_mod_mp        1279  gigc_chunk_mod.F90
wrf.exe            0000000001B63684  chem_driver_              824  chem_driver.f90
wrf.exe            000000000198D6EF  Unknown               Unknown  Unknown
wrf.exe            000000000060B05B  Unknown               Unknown  Unknown
wrf.exe            00000000004198D1  Unknown               Unknown  Unknown
wrf.exe            0000000000419884  Unknown               Unknown  Unknown
wrf.exe            0000000000419812  Unknown               Unknown  Unknown
libc-2.17.so	   00002B43C857A555  __libc_start_main     Unknown  Unknown
wrf.exe            0000000000419729  Unknown               Unknown  Unknown

I originally suspected this might be due to a problem in the wet deposition routines because I was finding small negative trace gas concentrations, but even after I fixed that issue this error keeps happening. I also thought it might be due to my nested domain setup, but the error also occurs on single-domain runs. The CO2 and CH4 simulations are working well, but I keep running into this problem in the full-chemistry model.

It seems like in all of my runs where this error occurs, I also get this error in the rsl.out.**** files:

 MET_CLDTOPS                    | Maximum cloud top he | xy    | level
 MET_CLDFRC                     | Column cloud fractio | xy    | 1
 MET_CHEMGRIDLEV                | Highest level of the | xy    | 1
 MET_AREAM2                     | Surface area of grid | xy    | m2
 MET_ALBD                       | Visible surface albe | xy    | 1
===============================================================================
GEOS-Chem ERROR: Could not allocate fullchem_mod.F90:PL_Kpp_Id!
 -> at GC_CheckVar (in Headers/errcode_mod.F90)
===============================================================================

===============================================================================
GEOS-Chem ERROR: Error encountered in "Init_FullChem"!
 -> at GC_Init_Extra (in module GeosCore/gc_environment_mod.F90)
===============================================================================

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%               HEMCO: Harmonized Emissions Component                 %%%%%
%%%%%               You are using HEMCO version 3.6.2                     %%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Reading settings & switches of HEMCO configuration file: HEMCO_Config.rc

Did you also see a similar error? The model usually keeps running for a long time after this before crashing with the PEDGE error, but I wonder if this is the root cause and if so how to fix it.