My question is not specifically related to TELEMAC, but I think some of the users could advise me on the matter. That is why I post it on this forum.
I am simulating tides and surges over a very large domain, including the Northwestern European continental shelf and parts of the Atlantic and Arctic oceans (see figure in attachment).
The bathymetry is defined by interpolating the EMODnet database, which covers entirely my domain with a resolution of 1/8 arc-minute:
portal.emodnet-bathymetry.eu/
I am using TPXO 7.2 to impose the tide at the open boundary.
The problem is that EMODnet is defined with respect to LAT (lowest astronomical tide, i.e. the height of the water at the lowest possible theoretical tide) as vertical datum, while TPXO is defined with respect to MSL (mean sea level).
Therefore, I need to convert EMODnet bathymetry from LAT to MSL. For sure, it won't make a significant difference in most of my domain, but in very shallow coastal areas, it will.
My idea is to make use of the TPXO database to evaluate the lowest possible theoretical tide as the sum of the amplitude of all available harmonic constituants, and consider that it is the difference between LAT and MSL. This has to be done at every grid cell.
However, there is a discussion in our team whether we can do that or not:
1) Is the sum of the amplitudes of all harmonic constituents representative of the lowest theoretical tide, or do we have to take care of nodal/satellite modulation or other features as well?
2) Are TPXO and local equivalents (e.g. OTIS Atlantic Ocean and European Shelf) are accurate enough for such calculation?
Those who are not in favor of this TPXO solution claim that we need to run a model over a very large period (more than 18.6 years to account for the lunar orbit tilt) and get the minimum water levels everywhere.
Every contribution on this discussion would be very welcome. Thank you!