Zarr Encoding Specification

In implementing support for the Zarr storage format, Xarray developers made some ad hoc choices about how to store NetCDF data in Zarr. Future versions of the Zarr spec will likely include a more formal convention for the storage of the NetCDF data model in Zarr; see Zarr spec repo for ongoing discussion.

First, Xarray can only read and write Zarr groups. There is currently no support for reading / writing individual Zarr arrays. Zarr groups are mapped to Xarray Dataset objects.

Second, from Xarray’s point of view, the key difference between NetCDF and Zarr is that all NetCDF arrays have dimension names while Zarr arrays do not. Therefore, in order to store NetCDF data in Zarr, Xarray must somehow encode and decode the name of each array’s dimensions.

To accomplish this, Xarray developers decided to define a special Zarr array attribute: _ARRAY_DIMENSIONS. The value of this attribute is a list of dimension names (strings), for example ["time", "lon", "lat"]. When writing data to Zarr, Xarray sets this attribute on all variables based on the variable dimensions. When reading a Zarr group, Xarray looks for this attribute on all arrays, raising an error if it can’t be found. The attribute is used to define the variable dimension names and then removed from the attributes dictionary returned to the user.

Because of these choices, Xarray cannot read arbitrary array data, but only Zarr data with valid _ARRAY_DIMENSIONS or NCZarr attributes on each array (NCZarr dimension names are defined in the .zarray file).

After decoding the _ARRAY_DIMENSIONS or NCZarr attribute and assigning the variable dimensions, Xarray proceeds to [optionally] decode each variable using its standard CF decoding machinery used for NetCDF data (see decode_cf()).

Finally, it’s worth noting that Xarray writes (and attempts to read) “consolidated metadata” by default (the .zmetadata file), which is another non-standard Zarr extension, albeit one implemented upstream in Zarr-Python. You do not need to write consolidated metadata to make Zarr stores readable in Xarray, but because Xarray can open these stores much faster, users will see a warning about poor performance when reading non-consolidated stores unless they explicitly set consolidated=False. See Consolidated Metadata for more details.

As a concrete example, here we write a tutorial dataset to Zarr and then re-open it directly with Zarr:

In [1]: import os

In [2]: import xarray as xr

In [3]: import zarr
---------------------------------------------------------------------------
ModuleNotFoundError                       Traceback (most recent call last)
Cell In [3], line 1
----> 1 import zarr

ModuleNotFoundError: No module named 'zarr'

In [4]: ds = xr.tutorial.load_dataset("rasm")
---------------------------------------------------------------------------
ModuleNotFoundError                       Traceback (most recent call last)
File /build/python-xarray-sfx7CS/python-xarray-2023.01.0/xarray/tutorial.py:132, in open_dataset(name, cache, cache_dir, engine, **kws)
    131 try:
--> 132     import pooch
    133 except ImportError as e:

ModuleNotFoundError: No module named 'pooch'

The above exception was the direct cause of the following exception:

ImportError                               Traceback (most recent call last)
Cell In [4], line 1
----> 1 ds = xr.tutorial.load_dataset("rasm")

File /build/python-xarray-sfx7CS/python-xarray-2023.01.0/xarray/tutorial.py:284, in load_dataset(*args, **kwargs)
    247 def load_dataset(*args, **kwargs) -> Dataset:
    248     """
    249     Open, load into memory, and close a dataset from the online repository
    250     (requires internet).
   (...)
    282     load_dataset
    283     """
--> 284     with open_dataset(*args, **kwargs) as ds:
    285         return ds.load()

File /build/python-xarray-sfx7CS/python-xarray-2023.01.0/xarray/tutorial.py:134, in open_dataset(name, cache, cache_dir, engine, **kws)
    132     import pooch
    133 except ImportError as e:
--> 134     raise ImportError(
    135         "tutorial.open_dataset depends on pooch to download and manage datasets."
    136         " To proceed please install pooch."
    137     ) from e
    139 logger = pooch.get_logger()
    140 logger.setLevel("WARNING")

ImportError: tutorial.open_dataset depends on pooch to download and manage datasets. To proceed please install pooch.

In [5]: ds.to_zarr("rasm.zarr", mode="w")
---------------------------------------------------------------------------
ModuleNotFoundError                       Traceback (most recent call last)
Cell In [5], line 1
----> 1 ds.to_zarr("rasm.zarr", mode="w")

File /build/python-xarray-sfx7CS/python-xarray-2023.01.0/xarray/core/dataset.py:2099, in Dataset.to_zarr(self, store, chunk_store, mode, synchronizer, group, encoding, compute, consolidated, append_dim, region, safe_chunks, storage_options, zarr_version)
   1982 """Write dataset contents to a zarr group.
   1983 
   1984 Zarr chunks are determined in the following way:
   (...)
   2095     The I/O user guide, with more details and examples.
   2096 """
   2097 from xarray.backends.api import to_zarr
-> 2099 return to_zarr(  # type: ignore
   2100     self,
   2101     store=store,
   2102     chunk_store=chunk_store,
   2103     storage_options=storage_options,
   2104     mode=mode,
   2105     synchronizer=synchronizer,
   2106     group=group,
   2107     encoding=encoding,
   2108     compute=compute,
   2109     consolidated=consolidated,
   2110     append_dim=append_dim,
   2111     region=region,
   2112     safe_chunks=safe_chunks,
   2113     zarr_version=zarr_version,
   2114 )

File /build/python-xarray-sfx7CS/python-xarray-2023.01.0/xarray/backends/api.py:1629, in to_zarr(dataset, store, chunk_store, mode, synchronizer, group, encoding, compute, consolidated, append_dim, region, safe_chunks, storage_options, zarr_version)
   1627     already_consolidated = False
   1628     consolidate_on_close = consolidated or consolidated is None
-> 1629 zstore = backends.ZarrStore.open_group(
   1630     store=mapper,
   1631     mode=mode,
   1632     synchronizer=synchronizer,
   1633     group=group,
   1634     consolidated=already_consolidated,
   1635     consolidate_on_close=consolidate_on_close,
   1636     chunk_store=chunk_mapper,
   1637     append_dim=append_dim,
   1638     write_region=region,
   1639     safe_chunks=safe_chunks,
   1640     stacklevel=4,  # for Dataset.to_zarr()
   1641     zarr_version=zarr_version,
   1642 )
   1644 if mode in ["a", "r+"]:
   1645     _validate_datatypes_for_zarr_append(zstore, dataset)

File /build/python-xarray-sfx7CS/python-xarray-2023.01.0/xarray/backends/zarr.py:363, in ZarrStore.open_group(cls, store, mode, synchronizer, group, consolidated, consolidate_on_close, chunk_store, storage_options, append_dim, write_region, safe_chunks, stacklevel, zarr_version)
    346 @classmethod
    347 def open_group(
    348     cls,
   (...)
    361     zarr_version=None,
    362 ):
--> 363     import zarr
    365     # zarr doesn't support pathlib.Path objects yet. zarr-python#601
    366     if isinstance(store, os.PathLike):

ModuleNotFoundError: No module named 'zarr'

In [6]: zgroup = zarr.open("rasm.zarr")
---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
Cell In [6], line 1
----> 1 zgroup = zarr.open("rasm.zarr")

NameError: name 'zarr' is not defined

In [7]: print(os.listdir("rasm.zarr"))
---------------------------------------------------------------------------
FileNotFoundError                         Traceback (most recent call last)
Cell In [7], line 1
----> 1 print(os.listdir("rasm.zarr"))

FileNotFoundError: [Errno 2] No such file or directory: 'rasm.zarr'

In [8]: print(zgroup.tree())
---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
Cell In [8], line 1
----> 1 print(zgroup.tree())

NameError: name 'zgroup' is not defined

In [9]: dict(zgroup["Tair"].attrs)
---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
Cell In [9], line 1
----> 1 dict(zgroup["Tair"].attrs)

NameError: name 'zgroup' is not defined