open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello everyone,

I am using Telemac to simulate tides using TPXO but also other atlases like FES2014.
When using these atlases, I have to convert the coordinates of the nodes (in my case Lambert93) into the same coordinate system as the atlas (in my case WGS84).
I follow the official guide 'Méthodologie pour la simulation de la marée avec la version 6.2 de TELEMAC-2D et TELEMAC-3D' and I see on p26 that I should set the angle BETA between these two projections (see attached picture).
But I cannot find how to calculate this guy, I thought it would be easy but then I searched more about Geodesy and …


I made a try using Matlab (see attached picture), my method was choosing two nodes and creating a vector going between these two nodes ; then create a second vector with the converted coordinates of the same nodes. Finally measuring the angle between these two vectors.
Problem : I think it is wrong as the value of the angle BETA is different if I choose different nodes. I know it can vary a little bit but I observe a variation from 3° to 10° on a mesh which measures 'only' 93km.

Thanks in advance to anyone helping me.

Hi,

As explained in the document, this angle is only used for the JMJ database (to create tidal boundary conditions).
With TPXO and FES-2014 databases, you only need to provide a mesh in projected coordinates (Lambert 93 for instance) or real lat/lon coordinates (WGS84).
You can have a look to the "tide" example.

Hope it helps.

Regards,
Laurent

Thank you Laurent for your reply.

I understood that it is used to create boundary conditions, it is clearly visible with BlueKenue when I force with tide currents. The vectors describe nice circles.

With TPXO no need as it is already two binary files that I can feed directly to Telemac.
But I thought that except for TPXO I had to give this angle.
Indeed, in the methodology it is described that BETA is needed for the subroutine TIDAL_MODEL_T2D.f where the JMJ database is used to create the harmonic consituents file.
With NEA, the file is already done, called har_tide-NEA_prior.txt in the example.
On the contrary with FES or Previmer, I have to create this harmonic constituent file myself (I use Matlab to go from the multiple netCDF files to a unique file to feed to Telemac). Doing basically the same work as TIDAL_MODEL_T2D.f if I understood well. And hence, I think I should use this angle to project my velocities amplitudes.
Am I wrong ?


I hope I am clear, if not please ask and I'll reformulate (or directly write in French if necessary)

No, You don't need to set BETA for other miscellaneous databases (as NEA, FES, PREVIMER). BETA is only used to create an harmonic constants file from JMJ database on a different grid (TIDALDB.EQ.1 in tidal_model_t2d.f).
And you're in the case TIDALDB=3.

Be aware that the format of the harmonic constants file isn't the same for JMJ database and other databases.

Please, have a look to the "tide" example with NEA database.

Regards,
Laurent

Thanks again for your fast replies.

I understand that I don't need to set BETA in Telemac.

Here is what I understand from the example 'tide' :
With JMJ, you feed bdd_tide-jmj.txt (ASCII file) to Telemac and it generates itself the harmonic constituents file har_tide-jmj.txt.

With NEA, you already have the har_tide-NEA_prior.txt harmonic constants file that you feed to Telemac.

With TPXO, you feed only the two binary files (h_tpxo9.v1 and u_tpxo9.v1) and Telemac handles the rest.

But
With FES and Previmer I don't have the harmonic constituents file. I only have numerous netCDF files.
So with Matlab I create a file with exactly the same formatting as har_tide-NEA_prior.txt using the netCDF files of the databases.
Hence I wonder if, during the creation of the harmonic constants file (with Matlab), I should use this angle BETA to project the velocities amplitudes.
Because, for all models not defined with WGS84 coordinates I have to convert these coordinates into WGS84 myself in order to interpolate the database grid and create the harmonic constants file.
So the positions of the nodes will be correct, but the direction of the north on my original projection is not necessarily the same as in WGS84.

Note that perhaps there is no problem at all and it's just me misunderstanding geodesy..

With JMJ database, you have 2 steps. A first one (TIDALBCGEN=.TRUE.), to generate the harmonic constants file from the ASCII DATABASE FOR TIDE and from TIDAL MODEL FILE and according to your mesh (in this case, you have to provide BETA). And a second one (TIDALBCGEN=.FALSE.), to run the model from the harmonic constant file previously created.

BETA is the angle between the projection used with JMJ database (a specific Mercator-JMJ) and Lambert projection (geo_tide.slf). Have a look to the tid_tide-jmj.slf (Mercator-JMJ projection)



With NEA and other miscellaneous databases (Previmer, FES...), you just have to create by yourself an harmonic constants file from each netcdf files you have, and with the same format as provided in appendix 3.
The feature to create the harmonic constants files from these databases is not available in current release. So no need to provide BETA and other input parameters from your input database (ASCII DATABASE FOR TIDE and TIDAL MODEL FILE) as you already built your harmonic constants file by yourself.


Regards,
Laurent

Thanks to Laurent for clarifying how Telemac handles the databases.
After some private messages, it is now clear that I have no Telemac issue but more a geodesy question.

Little recap :
* BETA is only needed with JMJ database when Telemac creates the harmonic constants file (step 1).
* With TPXO everything is done by Telemac.
* With other atlases the user has to create the harmonic constants file on its own.

I already have a Matlab script that does that.
So actually my question is more about how to calculate this angle, in order to use it into my own calculation when I create the harmonic constants file (and also because I'm curious).
For now it's not urgent, I only do a minor error with the projections I use, but this won't be the case with a non-conformal projection.