Closed ml14je closed 2 years ago
hm, for the wavelength sizing function, it has to calculate the wavelength of the M2 tidal constituent. It does this in SI units naturally because gravity and frequency use meters and seconds.
If you're in WSG84 (which you can check by looking at the CRS attribute, I guess you could multiply by the 111e3*cos(mean(y))
to get to meters where y
represents the latitude coordinates in your domain.
I understand that, but since individuals then feed that into their edge_size
function, surely it needs to be converted? Sorry, I am just coming across this issue now. I understand I can quite simply perform the above scaling factor, but thought perhaps there was a more robust approach taking into account the user-prescribed crs.
That’s what I’m suggesting. If you’re meshing in degrees, the above logic I wrote in the previous post will convert the edge lengths to degrees then since everything else is in degrees, it’s all fine. If you want to mesh in meters and you’re in a map projection nothing needs to be done there since it’s already in meters.
The
grid.values
are given in SI units of metres inwavelength_sizing_function
ofoceanmesh/edgefx.py
. These need to be translated to the grid coordinate system. Should be an easy fix?