open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello Francois,

You have these warnings because some nodes of your open boundary are located outside the domain of the geofin mesh. This is due to the shoreline of the geofin mesh whose boundary nodes have an elevation of 0 m CD, not like you I think.

The parameters XSHIFT and YSHIFT in TIDAL_MODEL_T2D subroutine enable to apply a translation for the open boundary nodes so that the interpolation of the harmonic constituents are done with elements of the geofin file. In your case, I think that you only have to change YSHIFT (if you delete the comment symbol ! before the 3 lines containing 57 in BORD_TIDAL_BC, an extra ASCII file is written, that contains the coordinates of the open boundary of your model in the same geographic system as geofin; then, you can visualize those nodes with geofin to see the translation to apply).
Nevertheless, if you do nothing, the interpolation of constituents is done with the nearest element of geofin.

For your strong velocities in SW corner, can you check if the node(s) is(are) always or not? If not always, it is advised to change its bathymetry, eg in CORFON as it is done in the test case, but you have to adapt it.

Regards,

Chi-Tuan

PS: the only thing I have changed in the last Fortran file compared to yours with standard suboutines, is a test for GEOSYST.EQ.1 where LAMBDA and PHI arrays are initialized with XBTIDE and YBTIDE arrays. Are you sure that you have not kept old XSHIFT and YSHIFT values?

Hello Chi-Tuan,
ok for the warnings
for the high velocities, it's not all the time, I'll look CORFON





Actually that's normal that I have not warnings error with the file princi1modif.f (princi1.f i've modified) because I removed the subroutine BORD_TIDAL_BC as you advise me in the first message.
Thank you

I've changed the bathymetry values of some nodes and it seems to work well. I've now to calibrate velocities and levels.
Many thanks for your help Chi-Tuan