open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi everyone,

Scenario 1:
I couple telemac2d-tomawac online. no problem.







in the output of tomawac, water levels and current velocities are correct (equal to the ones from output of telemac). but i see that they are not taken into account in the calculation (no tidal modulation). i realize the switch "CONSIDERATION OF TIDE " is set to "No" which has to be in case of online coupling i think. right?

Scenario 2:
I do online coupling again. i switch "CONSIDERATION OF TIDE " into "YES" and i get this error,

***************************************
ATTENTION : COUPLING TELEMAC-TOMAWAC :
CURRENT/WATER LEVEL FILE CANNOT BE
USED AS INPUT FILE.
END OF THE COMPUTATION
***************************************
I check the code (wac.f line 134), and it is normal to get this error as "MAREE" is set to True.

Scenario 3:
I do offline coupling. i run telemac. afterwards i run tomawac, and use output of telemac as binary current file. the results are fine. i see that tidal modulation is happening.

Question: what am i doing wrong in Scenario 1? how can i do this online coupling correctly?

The steering files i use in scenario 1 are attached.

thank you in advance for reading my long question and your help.

Hello

Could you please send the full case so that we try to run your case

thanks in advance

Thanks tfouquet for showing interest

I modified the steering files to make to the point. so please use the last ones.
and here you have all other files needed to run the case. I am using V7p0.

I also want to add, that the results i get from tomawac are spherical when i run it alone whereas they are UTM when i run it together with Telemac. although i keep the switch for spherical run on the whole time.




thanks

Hello

- i ran your case and now i am trying to analyse the results. Seeing the result of tomawac (Result2D_WesternScheldt_WAC.slf), I see that water levels and velocities vary but i don't know how to see tidal modulation.

- Could you also send scenario 3 (as it is the one who works) so that i can check that you use the same frequency of record between online and offline.

- I am also surprised that you simulate on only 1/12 of tide...

thanks for your answer.

Hi Thierry

this is the scenario 3.








It is simply the same as previous ones. the only difference is that, i just run Telemac alone. afterwards i run tomawac alone, using telemac results file as input for current and water level.

about tidal modulation, as you said, the values are currents and water levels which appear in the output of tomawac are correct. the problem is they do not cause tidal modulation of wave height. Thus in shallow water where i expect to get tidal modulation due to tide, i do not get that. you would see that if you compare the results of scenario 3 with scenario 1.

one more strange things that you will notice if you run Scenario 1 and sceanrio 3 is that , the results of telemac and tomawac are both Spherical in sc 3, but in sc 1, telemac results is spherical whereas tomawac is UTM.

- I "simulate only 1/12 of tide..."? what does that mean? i do not understand it. i am complete beginner with telemac.

thanks again for investigating this case

Hello

I agree about the problem with spherical coordinates we will investigate that point.

Concerning the results between scenario1 and scenario3 i am really sorry but i do not understand what you see on scenario3 that wouldn't be on scenario1. On the general view I can't explain the difference but i do not see anything more on Scenario3.

As for shallow water i supposed that you were interested in Scheldt so i zoomed on the river i still don't know what to look at, to see where they would be more modulation in scenario3. I would say there are difference but i can't see why Scenario3 is more influenced by the tide

As for 1/12 i just mean that you simulated one hour over 12 hours of tidal period.

Hi Thiery

sorry for very late reply, this issue got more and more complicated. now i have few clues. I did investigate this case further and found out some issue concerning shperical/mercator and coupling.

As the grid file i use both for telemac as well as tomawac is presented in spherical coordinates,

at the beginning of the run, telemac switches the grid to mercator (in subroutine readgeo3) because of the switch "SPATIAL PROJECTION TYPE : 3". The results of telemac however are written in spherical because of user defined subroutine write_mesh. what i get out of this makes sense.

when it comes to tomawac, i use the same grid which is in spherical coord. but i can make the choice to run (more precise to solve the equation) in mercator or spherical.
If i run tomawac in mercator, with the switch "SPHERICAL COORDINATES = NO", Tomawac uses user_defined subroutine corrxy to switch the grid to mercator, and uses write_mesh to write results in spherical coord.

Here appears the first problem. as sub readgeo3 calls corrxy, if corrxy is defined (to be used by tomawac) inside telemac_user_fortran_file, the grid will be converted to mercator two times which results in nonsense . I think i bypassed this issue in readgeo3.

I can get the same result from tomawac running in mercator in online and offline coupled case. that makes sense.

but what stays a problem is that for the same case,

for running tomawac in spherical, i get offline coupled results which are different from Mercator run. I do not manage to get online coupling working if tomawac is running in spherical. I get NaN.

does this problem sound like familiar a bug? anyone has had that? or is this what i can hope will be different in next version or do you think i have some errors in my settings?

I add the latest files as attachment. this run results in NaN which i can not trace if it starts in telemac or tomawac. dt and other settings can not be influential i guess as they have worked in offline case.















thank you again
Homayoon

Hello

Some improvements have been made in version 7.1 around spherical coordinates, so may be that your problems will disapear with the new version that is coming very soon.

Meanwhile, i saw that your Nan comes in the error in convergence in telemac calculation, so may be that you can take some smaller time steps to see if it is better.

Hope it helps

Hello Thiery

I tried this with the v7p1 and it did not change much.

I tried to explain the issue more clear here:
#19814

thanks
Homayoon