open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I have created a mesh, with nodes referenced to lat/long positions.
Is this OK?

I have added SPHERICAL COORDINATES = YES, but still get problems.

(I think it says in the manual that LATITU is called which converts Cartesial coordinates of geometry file to lat/long, but what happens if the nodes are already in lat/long?)

Thanks
Peter

Hello,

In your case, it is not exactly OK, you should have a Mercator projection instead of lat/long coordinates. A solution is to change your coordinates in subroutine CORRXY (bief library), with an origin point and you then give in the keywords LATITUDE OF ORIGIN POINT and LONGITUDE OF ORIGIN POINT.

We try to simplify this in the next version... in the while if someone has an example at hand to give...

With best regards,

Jean-Michel Hervouet

Thanks,

So, to clarify:

1. I convert my lat/long node-coordinates into x/y coordinates in CORRXY (this is quite a big task!?) - does this mean I don't need to convert back to lat/long if I don't want?
2. Shall I use my bottom left node as orinin in cas file?
3. Origin points in decimal degrees? (e.g. lat=50, long=-8)

In the past, I've converted into cartesian coordinates when making the grid, but I'd rather work in lat/long (days of work re-making the grid in BlueKenue, unless theres a way to convert coordinates there?)

Hello,

>1. I convert my lat/long node-coordinates into x/y coordinates in CORRXY (this >is quite a big task!?) - does this mean I don't need to convert back to >lat/long if I don't want?

Don't understand what is converting back to lat/long. Once you give the X and Y of the Mercator projection, the program will get the latitude from them.

2. It is better to have an origin at the centre of the domain.

3. Yes, degrees, from 0 to 90 for latitude in North hemisphere.

Regards,

JMH