open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear Users,

Can anyone give me an advice on how to secure the minimum water depth in domain as 900m. So basically I have a tank of water with constant depth everywhere as 900m. Also I have tides. So at the moment when tide is coming my water depth is changing from 900m to 100m and velocities are very high as a result. Is there is a way to secure the water depth level? I modified Condim subroutine but it did not fix the problem. Also I have 900m depth in HBOR in boundary condition file. Maybe because I have a positive water depth, as I set up a hypothetical tank, in correct geographical location, but my z is positive, but if I used the real bathymetry then z would be negative? I am using BlueKenue for creating mesh. If I need to have the negative depth , say -900, then should I go back to my mesh and fix it with BlueKenue(if it is possible) or there is another way to do it in Steering file or some Subroutine? But maybe my problem has other roots that I am not aware yet due to my lack of experience with modelling. Please advise. See figure on water depth error. Thanks in advance.

Kind Regards!

Violeta

Hello,

It is not totally clear. How the tide is imposed: by prescribing the depth and/or the velocity, with or without Thompson boundary conditions ? As you observe high velocities, it seems that you do not impose them, but as you mention that the depth goes down to 100 m, so you do not impose it either ?

Regards,

JMH

Hi Violeta
I'm not sure to well understand what you're doing and the figure is missing.
Nevertheless, I would suggest you to be very clear with the variable you use.

First, the "bottom level" which is the only "fixed" value of your system.
Secondly, the "free surface" which move during the simulation. In the tide simulation, you impose the free surface elevation!
Third, the "water depth" which is the difference between the free surface and the bottom.
The bottom and the free surface have absolute elevation as the water depth should be always positive (sometimes a little bit negative for numerical reason).

Hope this helps you to clarify your situation

Dear Jean-Michele/Christophe,

Thank you very much for your reply. I am not sure myself if this can work. As when I run another simulation I have for North Sea region with real bathymetry everything works fine. In one case I used TPXO file with u_tpxo and h_tpxo, and another case with JMJ harmonic files plus tid_tide and bdd_tide and necessary changes in Cas file depends on tidal database. When I moved to hypothetical tank, without real bathymetry, by using similar Cas files, one for TPXO and antoher JMJ data, the simulation works but water depth and velocities are wrong.

When I created Selafin file with BlueKenue, I specified the values at Free Surface when I used Add Variable. So I have Free Surface minus Bottom to give me 900m. Also when I created boundary conditions file I entered 900m as HBOR on all points. It all worked fine for other tank experiments, but not for tidal one. I probably quoted it wrong saying about negative water depth, I meant the depth of domain should be -900? This is a main difference between my real case North Sea and the tank one. I downloaded tidal data for Greenland fjord area and placed my tank on that location so geographically it is specified, so it has the Cartesian coordinates of that location. I think because I did not use real bathymetry, so my z coordinates is positive, as if I used real one as for North Sea then my z would vary, say from -0.01 to -900.

Since I have u_tpxo and h_tpxo, so I presume tidal velocities and depth are used from those files. As Option for Liquid Boundaries I have Thompson method -2. I have attached the figure now as format was wrong on my first post. Also I have attached the steering file, just in case. As boundary conditions on open boundary I tried both 566 and 544 but got same results. Please let me know if you need more details. Thanks in advance.

PS: I used 123_tide case as model.

Kind Regards!

Violeta

Dear Jean-Michele/Christophe,

I created a new bathymetry on this case. I placed points on the depth -900 on column z in my xyz file, interpolated it and created a new mesh with BlueKenue. Then I used it for my new selafin file. So looks like water depth is ok now, and velocities as well. I am not sure if I did it right but if you could take a look at attached, then could you please let me know if they are ok. I still have to work on other parameters to make it as close to "reality" as possible before I introduce the real bathymetry, but at least water depth looks ok to me.
I used both JMJ and TPXO data, and JMJ data gave me more realistic results. I am not sure if JMJ database contains tidal data from location 65N/38W (North Atlantic)but I got something there generated as harmonic constant. I also have a TPXO data but constituencies values are very low, so I suspect it might not be accurate, so my simulation results with TPXO are not as good, the tide is very low, hence velocities.
My question is on by introducing -900 in my interpolated mesh I did not violated any hydrology calculations and it really shows the 900m depth. Also it runs extremely fast, what is suspicious, even if I did it on cluster( 1min 40sec for simulation of 2days?). So I am not sure if it worked fine in spite of positive results. Thanks in advance.

Kind Regards!

Violet