We want to concatenate multiples of these files along the dimension sounding_dim but xarray complains, sounding_dim is not a coordinate. The time variable coud be used in this case as coordinate as it is monotonically increasing. The steps are (xarray v0.17):
files = glob.glob("*.nc")
datasets = [xr.open_dataset(f, decode_cf=False) for f in files]
datasets = [ds.swap_dims({"sounding_dim" : "time"}) for ds in datasets] # turns time into a coordinate variable
combined_dataset = xr.combine_by_coords(datasets, combine_attrs='override')
# adjust combined_dataset.attrs["time_coverage_end"] to combined_dataset.time[-1]
Ideally, we could implement such exotic rules by a custom pre-processor.
A CCI Greenhouse Gases daily product is like
We want to concatenate multiples of these files along the dimension
sounding_dimbut xarray complains,sounding_dimis not a coordinate. Thetimevariable coud be used in this case as coordinate as it is monotonically increasing. The steps are (xarray v0.17):Ideally, we could implement such exotic rules by a custom pre-processor.