PANGEO coord bounds fix - do we need it?

[agstephens]

From my testing withour Zarr output, we retain the bounds in the standard Xarray Dataset structure:

"http://cmip6-zarr-o.s3.jc.rl.ac.uk/CMIP6.AerChemMIP.NIMS-KMA.UKESM1-0-LL/"
"hist-piNTCF.r3i1p1f2.Amon.evspsbl.gn.v20200224.zarr"

>>> ds.data_vars
Data variables:
    evspsbl    (time, lat, lon) float32 dask.array<chunksize=(1980, 144, 192), meta=np.ndarray>
    lat_bnds   (time, lat, bnds) float64 dask.array<chunksize=(1980, 144, 2), meta=np.ndarray>
    lon_bnds   (time, lon, bnds) float64 dask.array<chunksize=(1980, 192, 2), meta=np.ndarray>
    time_bnds  (time, bnds) object dask.array<chunksize=(1980, 2), meta=np.ndarray>
>>> ds.time.bounds
'time_bnds'
>>> ds.time.bounds[:]
'time_bnds'
>>> type(ds.time.bounds)
<class 'str'>

Which matches:

>>> ds3 = xr.open_mfdataset('/badc/cmip6/data/CMIP6/AerChemMIP/NIMS-KMA/UKESM1-0-LL/hist-piNTCF/r3i1p1f2/Amon/evspsbl/gn/v20200224/*.nc', use_cftime=True, combine="by_coords")
>>> ds3.data_varsData variables:
    time_bnds  (time, bnds) object dask.array<chunksize=(1200, 2), meta=np.ndarray>
    lat_bnds   (time, lat, bnds) float64 dask.array<chunksize=(1200, 144, 2), meta=np.ndarray>
    lon_bnds   (time, lon, bnds) float64 dask.array<chunksize=(1200, 192, 2), meta=np.ndarray>
    evspsbl    (time, lat, lon) float32 dask.array<chunksize=(1200, 144, 192), meta=np.ndarray>
>>> ds3.time.bounds'time_bnds'
>>> type(ds3.time.bounds)<class 'str'>

Note that this relates to:

pydata/xarray#3689

naomi-henderson/cmip6collect#1

[agstephens]

I think this is a historical requirement and we don't need it, so closing.