Problems incorporating water-level forcing in a multigrid model

[jmirianco]

Hello everyone,

In my multi-grid model configuration, I want to include water levels as forcing for my two finest grids: "costa" and "local".
When I run WW3 using water levels only on the "costa" grid, it runs without any problems and produces good results. However, when I also try to include water levels on the "local" grid, the run stops after the first simulation hour.

Below I’m sharing the ww3_multi output where the error appears. If you need any additional input to help diagnose the problem, or if anyone has any idea where the issue might come from, I would really appreciate your help.

Thank you

MWW3 calculating for 2025/10/20 00:00:00 UTC at 14:24:37  status [ 0- 0]

  Updating input for grid  1 at 2025/10/20 00:00:00 UTC
     Updating winds        
        First updating wind         

 *** WAVEWATCH III WARNING IN W3IOBC : 
     INPUT FILE WITH BOUNDARY CONDITIONS NOT FOUND
     BOUNDARY CONDITIONS WILL NOT BE UPDATED     1


  Updating input for grid  2 at 2025/10/20 00:00:00 UTC
     Updating winds        

  Updating input for grid  3 at 2025/10/20 00:00:00 UTC
     Updating water levels 
        First updating level        
     Updating winds        

  Updating input for grid  4 at 2025/10/20 00:00:00 UTC
     Updating water levels 
     Updating winds        

  Updating input for grid  3 at 2025/10/20 00:00:00 UTC
     Updating water levels 
        First updating level        

  Updating input for grid  4 at 2025/10/20 00:00:00 UTC
     Updating water levels 

  MWW3 calculating for 2025/10/20 00:30:00 UTC at 14:24:41  status [ 1- 5]
  MWW3 calculating for 2025/10/20 01:00:00 UTC at 14:25:03  status [ 1- 5]

  Updating input for grid  1 at 2025/10/20 01:00:00 UTC
     Updating winds        
        First updating wind         

  Updating input for grid  2 at 2025/10/20 01:00:00 UTC
     Updating winds        

  Updating input for grid  3 at 2025/10/20 01:00:00 UTC
     Updating water levels 
        First updating level        
     Updating winds        

  Updating input for grid  4 at 2025/10/20 01:00:00 UTC
     Updating water levels 
     Updating winds        
--------------------------------------------------------------------------
mpirun noticed that process rank 15 with PID 70365 on node node47 exited on signal 11 (Segmentation fault).
--------------------------------------------------------------------------

[mickaelaccensi]

Hi Joaquim,

It is possibly a memory issue, try to allocate more memory for your run.
you can also recompile WW3 with debug option, for example :
cmake .. -DCMAKE_BUILD_TYPE=Debug -DSWITCH=Ifremer2
to have more details about the error

Mickael

[jmirianco]

Mickael, thanks for the help.
After increasing the memory for the run, it was able to complete successfully. However, I’m still having an issue: in the local grid the WLV field is coming out entirely zero, while in the coastal grid it looks correct.

Regarding the suggestion to use -DCMAKE_BUILD_TYPE=Debug, I wasn’t able to compile properly with that option (I assume it was due to too much output or something similar). But I did manage to get it working using this option instead:
-DCMAKE_BUILD_TYPE=RelWithDebInfo.

What do you think? Do you have any idea what might be going on?

Thank you very much.

[mickaelaccensi]

Hi Joaquim,

it is probably due to the activation of the current in the grid, it must be activated in ww3_multi.nml

It is weird to not be able to compile with Debug option, if you can provide the exact error message when running cmake.

Best,

Mickael

[jmirianco]

Hi Mickael,
Just to be sure.. when you told me that currents needed to be activated, was that a mix-up with the water levels, right? here it is the multi configuration I used:

! -------------------------------------------------------------------- !
! WAVEWATCH III ww3_multi.nml - multi-grid model                       !
! -------------------------------------------------------------------- !

&DOMAIN_NML
  DOMAIN%NRINP  = 2
  DOMAIN%NRGRD  = 4
  DOMAIN%UNIPTS = T
  DOMAIN%IOSTYP = 3
  DOMAIN%PSHARE = T
  DOMAIN%START  = '20251020 000000'
  DOMAIN%STOP   = '20251021 000000'
/

&INPUT_GRID_NML
  INPUT(1)%NAME                   = 'wind'
  INPUT(1)%FORCING%WINDS          = T

  INPUT(2)%NAME                   = 'level'
  INPUT(2)%FORCING%WATER_LEVELS   = T
/


&MODEL_GRID_NML
  MODEL(1)%NAME                     = 'AtlSur'
  MODEL(1)%FORCING%WINDS            = 'wind'

  MODEL(2)%NAME                     = 'Region'
  MODEL(2)%FORCING%WINDS            = 'wind'

  MODEL(3)%NAME                     = 'local'
  MODEL(3)%FORCING%WINDS            = 'wind'
  MODEL(3)%FORCING%WATER_LEVELS     = 'level'

  MODEL(4)%NAME                     = 'costa'
  MODEL(4)%FORCING%WINDS            = 'wind'
  MODEL(4)%FORCING%WATER_LEVELS     = 'level'
/

&OUTPUT_TYPE_NML
  ALLTYPE%FIELD%LIST       = 'DPT CUR WND HS DIR SPR FP DP WLV MSS TUS USS'
  ALLTYPE%POINT%FILE       = 'points.list'
  ALLTYPE%POINT%NAME       = 'POINTS'
/

&OUTPUT_DATE_NML

  ALLDATE%FIELD          = '20251020 000000' '3600' '20251021 000000'
  ALLDATE%POINT          = '20251020 000000' '3600' '20251021 000000'
  ALLDATE%RESTART        = '20251020 000000' '86400' '20251021 000000'
/

! -------------------------------------------------------------------- !
! WAVEWATCH III - end of namelist                                      !
! -------------------------------------------------------------------- !

About the debug option: the error is not in the compilation, but in the WW3 run.
Once I compile with the debug option, the multi run returns these errors:

       Type 4 : Restart files                 
      -----------------------------------------
            From     : 2025/10/20 00:00:00 UTC
            To       : 2025/10/21 00:00:00 UTC
            Interval :       1/01 00:00:00

-------------------------------------------------------
Primary job  terminated normally, but 1 process returned
a non-zero exit code.. Per user-direction, the job has been aborted.
-------------------------------------------------------
--------------------------------------------------------------------------
mpirun detected that one or more processes exited with non-zero status, thus causing
the job to be terminated. The first process to do so was:

  Process name: [[4329,1],58]
  Exit code:    2
--------------------------------------------------------------------------

At line 4729 of file /clusteruy/home03/flyma/WW3_7.14/WW3/model/src/wminitmd.F90
Fortran runtime error: Index '8' of dimension 1 of array 'outpts' above upper bound of 7
At line 4729 of file /clusteruy/home03/flyma/WW3_7.14/WW3/model/src/wminitmd.F90
Fortran runtime error: Index '8' of dimension 1 of array 'outpts' above upper bound of 7

(this message is repited like 20 times)

[node47.datos.cluster.uy:28444] 59 more processes have sent help message help-mpi-btl-base.txt / btl:no-nics
[node47.datos.cluster.uy:28444] Set MCA parameter "orte_base_help_aggregate" to 0 to see all help / error messages

Thanks again,
Joaquín

[mickaelaccensi]

I've corrected this bug last week, so if you update your git repo it should be fine.
About ww3_multi.nml, your configuration seems ok. I hope that with debug option, you'll have more information

Mickael

[jmirianco]

Hi Mickael,

I’m running the model on a cluster that doesn’t have the latest MPI version required to compile this new branch. I’ve been trying to compile it over the past few days, but I haven’t been able to get it working yet.

Would it be possible for me to send you a zip file with the necessary input files so you can run it in your environment and take a look at the error? I know it’s a lot to ask, but it would really help me.

Thank you very much,
Best regards,
Joaquin

[jmirianco]

Hi Mickael,

I was finally able to solve the issue with the water levels on the larger grid!

The fine grid resolution (costa) was 0.67/60°, while the larger grid (local) had 2/60°, and the water levels were defined every 1.2/60°.
By decreasing the resolution of the water-level grid to 3/60°, I was able to run the model without any problems on both grids.

Thanks a lot for your help!
Best regards,
Joaquin

[mickaelaccensi]

Great !