Hi,
I try to produce a fan-shaped grid with seagrid, but I doubt that the orthogonality of the grid affects the calculation. I'm not sure whether the curvilinear orthogonal grid is proper.
curvilinear orthogonal grid
Re: curvilinear orthogonal grid
Are you asking how orthogonal your grid is? Does SeaGrid give an "orthogonality error"? Last I checked, SeaGrid only called "rect" once. In the ancient Fortran code, we would call it iteratively until it crashed, then try again with one less iteration. "rect" improves the orthogonality logarithmically with each iteration. It helps if your corners are close to 90 degrees.
Re: curvilinear orthogonal grid
Thanks kate for your reply,
I checked the orthogonality of my grid, the results is not good.
Should I use curvilinear grid carefully, and just use the land masking?
I checked the orthogonality of my grid, the results is not good.
Should I use curvilinear grid carefully, and just use the land masking?
Re: curvilinear orthogonal grid
You have several choices:
1. Edit the code to call rect more times.
2. Make a rectangular grid and use the land mask.
3. Explore other options like gridgen or the ancient Fortran code.
1. Edit the code to call rect more times.
2. Make a rectangular grid and use the land mask.
3. Explore other options like gridgen or the ancient Fortran code.