With xarray-regrid it is possible to regrid between two rectilinear grids. The following methods are supported:
- Linear
- Nearest-neighbor
- Conservative
- Cubic
- "Most common value" (zonal statistics)
Note that "Most common value" is designed to regrid categorical data to a coarse resolution. For regridding categorical data to a finer resolution, please use "nearest-neighbor" regridder.
pip install xarray-regrid
The xarray-regrid routines are accessed using the "regrid" accessor on an xarray Dataset:
import xarray_regrid
ds = xr.open_dataset("input_data.nc")
ds_grid = xr.open_dataset("target_grid.nc")
ds.regrid.linear(ds_grid)
For examples, see the benchmark notebooks and the demo notebooks.
The benchmark notebooks contain comparisons to more standard methods (CDO, xESMF).
To be able to run the notebooks, a conda environment is required (due to ESMF and CDO).
You can install this environment using the environment.yml
file in this repository.
Micromamba is a lightweight version of the much faster "mamba" conda alternative.
micromamba create -n environment_name -f environment.yml
This package was developed under Netherlands eScience Center grant NLESC.OEC.2022.017.