open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear all

I have a problem when applying Spherical Coordinates and would be grateful for any suggestions.

When SPERICAL COORDINATES are invoked (requiring also CORIOLIS), the forcing terms FU and FV are calculated at all points using a Coriolis Coefficient calcilated from the latitude of the point ( SINLAT%R(I) ) (from PROSOU).

I am able to see that the Latitude is varying (from an output of SINLAT), LATITU has calculated the latitudes from ordinates of the mesh and the Latitude of the Origin Point and FU%R(I) and FV%R(I) are calculated.

But my results however are not right when SPHERI = TRUE. I can only get a correct result when applying uniform Coriolis Coefficient.
My mesh has a UTM mercator coordinate base.

Am I missing another step?

John

Hello John,

I can not guess if you are missing any step while you did not attach your steering file. However, I recall hereafter the keywords you need to activate to use a Coriolis force which varies in space:
CORIOLIS = YES // to activate the force
SPHERICAL COORDINATES = YES // to activate spherical coordinates
LATITUDE OF ORIGIN POINT = ... // used to compute the spherical coordinates from cartesien coordinates of the geometry file
NORTH = ... // gives the angle (degrees) between the geographic north and the Y-axis of the model (positive in the trigonometric direction)

Verify if you have all these keywords. If it is the case, you have an other problem somewhere. Send us your folder and I will try to see more in deep the problem.

With my best regards.

Riadh Ata

Hi Riadh

Thanks for your reply.

Yes - I have set all those keywords correctly, originally I hadn't set NORTH but it defaults to 0 which is correct.

I've attached the steering file and fortran file which includes a local copy of PROSOU.f which I was using to debug.

Best Regards
John

Hi John,

I think that you have attached the fortran file twice instead if the steering file.

Moreover, I see that you print out SINLAT (which is varying in your case) and FU%R and FV%R which are not varying (I guess), isn't it?. So the problem comes from VN%R and UN%R which are not varying (or at least not correctly). Try to check what is the problem with VN and UN.

I will wait for your steering file, and if you dont mind, send me your model directly on my email This email address is being protected from spambots. You need JavaScript enabled to view it., I will run it on my computer.

Regards,

Riadh

Error in attaching the file - steering file attached

Hello,

I am not sure that Universal Transverse Mercator is what is expected by the program. The (simple) Mercator projection is :

x = (Radius of Earth) * (longitude-longitude of origin)
y = (Radius of Earth) * (Log[tan(latitude/2+PI/4)]-Log[tan(latitude of origin/2+PI/4)])

and this formula is inverted to find back the longitude and latitude. It seems to me that UTM Mercator is much more complicated, with a definition that changes locally. We could treat it if someone is able to explain what it really is.

With best regards,

Jean-Michel Hervouet

Hi Jean-Michel

Ok - I can transform the mesh with a simple mercator projection. Or this could be done directly in CORRXY, is that correct?

I think the UTM mercator was used because the basic data were prepared using ARCMAP. Nothing about ARCMAP is simple!

Thanks for your help.
John

Hello,

Yes you can do it in CORRXY. The projection is required not only to get the latitude, but also to get the local scale in every element.

Regards,

JMH