How to create a curvilinear coordinate grid from scrach with python?

Question

I have a NetCDF obtained from a WRF simulation with a curvilinear grid with the following attributes:

I want to create new coordinates for the same domain, but with a horizontal resolution of 5km. I would like to do it with xarray if it's possible.

Is your goal to create the dimension coordinates in xarray that are absent from your WRF dataset, or is it to regrid your existing data to a grid of reduced grid spacing? If it is the first, [this gist](https://gist.github.com/jthielen/8881d32d08d625c75c906a7e8ad7583f) may be helpful, otherwise, I suggest looking into [xesmf](https://xesmf.readthedocs.io/en/latest/) — Jon Thielen, Aug 13 '19 at 18:44

score 0 · Answer 1 · answered Aug 08 '19 at 11:31

At present it seems difficult to deal with curvilinear grids with xarray (https://github.com/pydata/xarray/issues/2281)

But you can use cdo to transform your grid to a regular grid using cdo (http://gradsusr.org/pipermail/gradsusr/2014-February/036493.html)

And then you can actually regrid to any resolution you may want. You can follow the example at http://xarray.pydata.org/en/stable/interpolation.html#example to interpolate to a higher resolution.

1 Answers1