Stops - "NetCDF: Variable not found" for ubar_east / vbar_east (interannual run)

run / config issues
,
1

Hi,
I am running an interannual CROCO simulation for October 2019 and the model stops after some time with the following error (croco_interY2019M10.out):

21598 7228.49861 2.327056888E-03 5.0606013E+01 5.0608340E+01 3.5545640E+15 0
21599 7228.49931 2.327127077E-03 5.0606004E+01 5.0608332E+01 3.5545640E+15 0

ERROR in nf_read_bry: nf_get_vara netCDF error code = -49
Cause of error: NetCDF: Variable not found mynode = 30

ERROR in get_bry: cannot read variable ‘ubar_east’

ERROR in nf_read_bry: nf_get_vara netCDF error code = -49
Cause of error: NetCDF: Variable not found mynode = 30

ERROR in get_bry: cannot read variable ‘vbar_east’

I have already checked:

  • The .bry file does contain ubar_east and vbar_east variables
  • No NaNs in these variables.

I cannot figure out why this happens for October 2019.
Has anyone experienced something similar?

Thank you in advance.

2

I had a similar error, it could be because CROCO is trying to read boundary data beyond the available time steps in your boundary file “.bry”. * Your simulation requires boundary conditions for every time step in October, including the transition to November. If your boundary file ends exactly at 00:00 on 1 November (or earlier), CROCO will fail when trying to read data for the next time step (e.g., 06:00 on 1 Nov, depending on your time resolution). I resolved this adding at least an extra time step in the boundary conditions.

3

In my case, the issue arose after I added additional rivers to the simulation; prior to that, it was running smoothly. I verified the boundary condition time steps, and I have three bry_time entries corresponding to the first, middle, and last day of October. This setup is consistent for every month. To avoid further errors, I temporarily removed some rivers and re-ran the simulation.

1 Like