open TELEMAC-MASCARET The mathematically superior suite of solvers

I'm trying to run a tidal simulation in Telemac v7p0r1 between 0000 UTC 19 June 2016 - 0000 UTC 23 June 2016. I have attached the boundary conditions (
, ), mesh geometry () and steering () files. I am using the TPXO European Shelf model with both H and H+UV forcing on the open boundaries (Thompson boundary conditions).

However I am getting very high velocities and maximum water levels, see
and
.

I expect this is because of tidal flats/ very shallow areas of bathymetry/ my boundary location, see . I have therefore tried smoothing the bathymetry and including treatment of tidal flats, but the problem just changes so that different areas now have unreasonably large velocities/ maximum water levels (not shown, but always near boundary nodes).

Is there any way to obtain more realistic results without changing the free surface gradient compatibility? I am aiming to run this with atmospheric wind and pressure forcing, and the free surface is unacceptably smoothed for my purposes by having a free surface gradient compatibility of < 0.9.

I have an idea to slowly ramp up the forcing and see if I still get these very high velocities and maximum water level. I would like to force the open boundaries with just elevation, and ramp up over time such that the elevations start at 0 and goes to 'real' elevations 24 hours later. (I would therefore not be interpolating to TPXO for my initial conditions, but having 0 m elevation initial conditions.) Which module should I therefore edit in Telemac2D? Would it instead be easier to do this by changing the binary TPXO files? If so, how do I get into these files and change them manually? Or should I change my mesh to have boundaries further outside of the domain I am interested in, and therefore also in deeper water?

Many thanks,
David

For anyone interested, I have found that you can edit the subroutine BORD_TIDE_TPXO such that there is a RAMP function dependent on time (TEMPS). Editing Z such that Z = RAMP*CTIDE*TPXO_PTIDE + MSL seems to do the trick when wanting to do a cold start in Telemac2D using TPXO. I have found that more stable solutions have been found when ramping up from a cold start over 2 days for the English Channel.