Strange loss of tidal signal in SSH (zeta) at end of each month

[jordiiglesias]

Hello everyone,

I’m running a ROMS simulation forced with the Medsea hourly boundary data (which already include tides, so I’m not adding tidal forcing separately). The model generally behaves well, but at the end of every month my sea surface height (zeta) loses its tidal variability, the signal flattens and the daily tidal cycle disappears.
This issue repeats every month, always during the last day(s) before day 1 of the next month.
I checked the boundary forcing:
The Medsea boundary SSH looks correct (plot attached).
I also computed FFTs of Medsea U and V (comparing 10-day averages vs. daily windows) to see if tides were missing at the end of the month, but I don’t see any significant difference.


Has anyone seen this behaviour before?
Is there something in ROMS) that could cause tidal energy to vanish only at the end of each month?
Additionally i have this problem with CROCO.
I attach the 3 different points signal, averaged over all domain signal (with other parameters that doesn't have this problem), and bry from forcing conditions
Any suggestions or ideas would be greatly appreciated.

Best regards,
Jordi

signal_3_pints.png (34.3 KiB) Viewed 638 times

bry.png (80.22 KiB) Viewed 638 times

[MBendoni]

Do you have your OBC data in a single netcdf or multiple monthly files?
Maybe you can try to put all data in a single bry file and check if you still have the issue.

[jordiiglesias]

Hello, and thanks for the reply.

All the OBCs are stored in a single NetCDF file, and I'm only running 10-day simulations. This makes the behavior a bit strange, because there is only one time record for all the boundary conditions. Despite that, only the sea surface height (SSH), as shown in the plot, is being affected, while the other variables have the natural behaviour following the parent model.

[rtoste]

It is probably related to your input files. Check your log file and see what is happening at the end/beginning of the month