Non-conservatively interpolate from a latitude-longitude grid to a curvilinear (tri-polar) ocean grid with a given land/sea mask and fill in where needed.
./interp_and_fill.py -h
./interp_and_fill.py ocean_hgrid.nc ocean_mask.nc source_data.nc variable_name new_file.nc
- Super-sample ocean grid by binary division until the number of degrees of freedom in the i-direction are finer than those of the source data.
- Linearly interpolate source data to super-sampled grid.
- No extrapolation or nearest neighbor interpolation.
- Average super-sampled data to model cells.
- Objectively fill-in missing data using an elliptic solver.