xarray_regridder

A flexible regridding utility API and CLI based on xarray, pyproj and xesmf. Maps from input data grid(s) onto a regular target grid.

Note - the target grid must have equally spaced, one-dimensional coordinates.

Environment

Installation into a miniforge enviromnent is suggested. See https://github.com/conda-forge/miniforge for installing miniforge.

mamba create -n xarray_regridder_env python=3.10
mamba activate xarray_regridder_env
mamba install netcdf4 xarray pyproj xesmf scipy
pip install git+https://github.com/surftemp/xarray_regridder.git

supported regridding modes

This tool will analyse the set of source pixels that map to each destination pixel in the regular target grid.

mode name	function
min	retain the min value of all valid source pixels that map to a destination pixel
max	retain the max value of all valid source pixels that map to a destination pixel
mean	retain the mean value of all valid source pixels that map to a destination pixel
nearest	retain the value of the nearest valid source pixel that maps to a destination pixel
sum	retain the sum of all valid source pixels that maps to a destination pixel
count	retain the count of all valid source pixels that maps to a destination pixel
distance	retain the distance of the closest valid source pixels that maps to a destination pixel

limitations

• All input variables must have only 2 non-unit dimensions, with the Y dimension placed before the X dimension
• The mapping is performed by binning each source pixel onto the target grid
• Coordinates on the target grid MUST be evenly spaced
• Nearest source to destination regridding is not supported. If the target grid is coarser than the source grid, the output data will have missing values.

input data

input data can be specified as an input netcdf4 file or a folder containing multiple netcdf4 files

Examples

Convert from data imported from landsat (see see https://github.com/surftemp/landsat_importer) to a regular lat-lon grid (target_grid.nc)

Use the nearest neighbour mode to map variables B2, B3, B4 to the target grid:

xarray_regridder landsat_imported.nc target_grid.nc regridded_output.nc --source-crs 4326 --source-x lon --source-y lat --target-crs 4326 --target-x lon --target-y lat --variables B2:nearest:B2 B3:nearest:B3 B4:nearest:B4

Here, target_grid.nc has 1-dimensional lat and lon variables:

ncdump -h target_grid.nc 
netcdf grid {
dimensions:
	nj = 2000 ;
	ni = 2000 ;
variables:
	double lat(nj) ;
		lat:_FillValue = NaN ;
		lat:units = "degrees_north" ;
		lat:standard_name = "latitude" ;
	double lon(ni) ;
		lon:_FillValue = NaN ;
		lon:units = "degrees_east" ;
		lon:standard_name = "longitude" ;
}