open TELEMAC-MASCARET The mathematically superior suite of solvers

I am running telemac3d v7p2r2, and get errors in output when coupling with WAQTEL. If anyone can suggest alternatives to try in the telemac or waqtel steering files to help solve or diagnose this problem, this would be much appreciated.

I am running with initial conditions and boundary forcing provided from a regional ocean model, as shown in fortran and steering files attached. Everything is stable until I try to introduce atmospheric forcing through WAQTEL. After ~ 6h the model output becomes NaN. I have attached a sample output file for reference to show these errors. Aside from exceeding iterations, there is no obvious cause of errors (the model in fact keeps running, but with useless output!). Coupling is required to provide the surface heat fluxes and therefore maintain the stratification during the simulation, so we need to resolve this problem.

I have tried running for a different month, and with a different surface forcing source, but still get the same error. So the problem doesn't seem to be caused by an error in the met forcing file itself (e.g. no spike in wind speed or temperature).
I have tried changing the turbulence scheme. This reduces the "exceeding iterations" errors, but still get NaN after ~ 6h.
The problems I've experienced sounded similar to those described here. Following the suggestions here I then changed the advection schemes (from 1 to 14 for velocity and k-epsilon), and also the option for each scheme to 4. However, this showed no improvement.
As the model includes tidal flats, we use(since option 2 is said to be unstable for versions < v8p0). Following the note of a bug in calcs3d_thermics here I updated this routine in the fortran file. Again this showed no real improvement.

The latest version of the model steering files are attached. As mentioned earlier, the model is stable until we introduce WAQTEL, so I suspect the problem is somehow down to the treatment of this coupling.

Any advice you can give would be appreciated. Thank you.

Dear jgraham,

Sorry for this issue. We are working on it...

Best regards,
Jacques

Hi Jacques,

Thank you for responding.

Do you know when you'll be able to provide further suggestions or solutions?
Let me know if there's anything I can do to help, as we've been stuck on this for a while and need to try and make some progress soon.

In the meantime, I am running further tests, but no luck so far. I'll let you know if anything changes.

Thanks again,
Jenny

Did you try the latest version v8p0?

I haven't yet. We made a lot of modifications in the fortran file for dealing with the initial conditions and boundaries, so have been hesitant to move to the recent version since we would need to reproduce the fortran files for the configuration (I heard there were a lot of changes in the code structure moving to v8p0?).

However, if there is a known problem with v7p2r2 that would be causing these stability issues, and you know has been fixed in v8p0, please let me know.

I don't know if this has been fixed but as you could imagine, spending time to try to solve a problem which could be maybe already solved with the latest development, this is not really efficient.
V7p2 is 2 years old...
The changes exists but are not so huge. In fact, some structure changed to split the different process and have a clearer code.
You probably have to spend some time to transfer your modifications but copy/paste could be use and you will not have to rewrite your entire modifications...

regards

I have now tried a simulation at v8p0r2, but still have the same problem - NaN values throughout after 6h when coupling with WAQTEL.

To simplify the problem, I have run a simulation using uniform T/S for initial conditions, and just tidal forcing at the boundary.
The atmospheric forcing is also uniform, to rule out atmospheric variability as cause for timing of instability.

New fortran, steering and forcing files are attached. I have also attached the mesh and boundary files so you can reproduce the problem if needed.

Any advice would be greatly appreciated.