open TELEMAC-MASCARET The mathematically superior suite of solvers

Elodie,

Thanks for your answer.
I try to modify my wind file. I change the latitude and longitude system in a UTM system but the error is always here.
May be can i create a run with telemac2d in using the routine conv_mercator_to_degdec in order to transform my domains in lat/lon?

I work again on my fortran file which is composed by semip.f, noudon.f and lecdoi.f.

I try to make a new mesh in changing my coordinate system.

Regards
Florian

If you modified your wind file coordinates, did you then look at your values of USNEW : are there still nonphysical values, as NAN?

Elodie

With the debugger, i don't see NaN. For USNEW, i have a constant value for all my points (59478.1166279736). I try with an other period of simulation but i have the same problem with a wind file in lat lon or in UTM...

So you see that your values of USNEW are, as I said, really nonphysical (USNEW represents the friction velocity!).It shows that you did not solve your wind interpolation problem. You can print the values of your wind before and after the interpolation in FASP to see what it gives but I am pretty sure your value after interpolation is -1.D6...

Elodie

The order of the routines in my fortran file can affect the run?
Is it preferable to put semimp after noudon or noudon after semimp?

It is Noudon which calls fasp.f but i am not sure that semimp goes up "call noudon".

Regards

The order of the routines in my fortran file can affect the run?
=> no, that's the problem here

It is Noudon which calls fasp.f but i am not sure that semimp goes up "call noudon".
=> lecdoi.f does the interpolation at the first time step and then noudou.f at the next ones. If the interpolation does not work in lecdoi.f at the fist time step, it won't work at the other ones... Please try to print you variables in lecdoi.f and correct your interpolation problem!

I try to resolve my error.
My run reads lecdon and not lecdoi... first error

In fact, the values for U1, V1, XRELV,YRELV... are false. I convert my bathymetry file in lat/lon to create a new mesh but Matisse doesn't take the lat/lon. In the script to extract my wind file, i will try to modify it in order to work with a UTM system.

If this solution doesn't give some results, i will not have more ideas to correct this error

Elodie,

My different tests to correct my interpolation error are not successful.

If i make a coupling of telemac 2d and tomawac, i can use the routine convert_Mercator_to_degdec.f to transform my domains in latitude and longitude. After, in the routine corxy.f, i can make a call of the routine convert_Mercator_to_degdec in order to consider a modification of my coordinates.

Can i realize this operation?
What do you think about this idea?

Thanks in advance
Regards

Florian

Hello,

I think you should rather try to correct your wind or your bathymetry file to use the same coordinates before runing your Tomawac simulation (and not inside the simulation).

But you can indeed use the routine convert_Mercator_to_degdec as a basis to make your own fortran program in order to modify the coordinates.

Regards,

Elodie