[MITgcm-support] Questions Setting Up Bathymetry File

[Martin Losch]

Hi Keerthijan,

did you get a reply?

I think the python code looks good to me, except that I would expect to have your “bandarray” with dimensions (ny,nx) following the python convention. I then usually don’t need the ravel method. In the end you’ll find out by trial and error, which ordering is correct.

- Nx = sNx*nSx*nPx, where Nx are the number of points in x-direction in your bathy file, and similar for Ny. nSx/nSy should normally be 1 unless you want a hybrid MPI/OpenMP executable.
- you can set (constant) resolution by using dxSpacing, dySpacing in “data”, the units depend on the grid you choose. For a cartesian grid it’s meters. For a spherical polar grid degrees.
- Nr is the number of vertical grid point.

Martin

> On 24. Jul 2018, at 17:31, Keerthijan Radhakrishnan <kradhakrishnan at edu.uwaterloo.ca> wrote:
> 
> Hello,
> 
> I am working on modelling water flow through a strait using MITgcm. I am new to this and I have some questions setting up the bathymetry file for upload to MITgcm. I would appreciate any help.
> 
> To create my bathymetry file, I downloaded the bedrock in a geotiff format from NOAA. The coordinates were in latitude/longitude and I warped it to metres. Then I cropped the geotiff to remove all areas without data and obtained a rectangle. This I converted to a Numpy array on python. My array is of shape: (points in x, points in y) and the elevation at each point is the value in the matrix. I converted the array to a bin using the following code on Python:     
> dt='>f8'
> fid = open("bathymetry.bin, "wb")
> bandarray.astype(dt).ravel(order='F').tofile(fid)
> fid.close()
> 
> My question is, when I provide info in the SIZE.h file, are the following correct:
> 	• sNx*nPx should equal the number of X points on the bathymetry file.
> 	• sNy*nPy should equal the number of Y points on the bathymetry file. 
> 	• The resolution should equal the spacing between each point in the X/Y and in metres? 
> 	• Then what is Nr? We are not providing a grid for the Z coordinates only values of elevation. 
> 	• Also, is my process above correct?
> 
> Regards,
> Keerthijan
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support