Kerchunk

Kerchunk.jl is a Julia package that enables loading Kerchunk reference catalogs as Zarr.jl arrays.

Installation

This still relies on an experimental branch of Zarr.jl to read most Kerchunk files.

] 
add Kerchunk

(clean up by running ]free Zarr ZarrDatasets Rasters)

Quick start

Kerchunk.jl is simply a storage backend to Zarr.jl. Zarr does integrate with the more fully featured packages Rasters.jl and YAXArrays.jl, which are the packages you will want to use to interact with Kerchunk data.

using Kerchunk, Zarr

za = Zarr.zopen("reference://path/to/kerchunk/catalog.json")
# and treat it like any other Zarr array!
# You can even wrap it in YAXArrays.jl to get DimensionalData.jl accessors:
using YAXArrays
YAXArrays.open_dataset(za)
# or open it as a Rasters.RasterStack:
using Rasters
Rasters.RasterStack(
    "reference://catalog.json", 
    source = Rasters.Zarrsource(),
    lazy = true, # need to include this
) # source must be explicit

It's most useful to open Kerchunk datasets as either RasterStacks or YAXArrays datasets, since both of those packages have great dimensionality support.

Background

kerchunk is a Python package that generates the reference catalogs.

Limitations

• No support for gen references with templates.
• No support for complex Jinja2 templates in refs. (Although Kerchunk hardly supports this either...)

Acknowledgements

This effort was funded by the NASA MEaSUREs program in contribution to the Inter-mission Time Series of Land Ice Velocity and Elevation (ITS_LIVE) project (https://its-live.jpl.nasa.gov/).

Alternatives and related packages

• You can always use Python's xarray directly via PythonCall.jl
• FSSpec.jl is an alternative storage backends for Zarr.jl that wraps the same fsspec that xarray uses under the hood.

This package is of course built on top of Zarr.jl, which is a pure-Julia Zarr array library. YAXArrays.jl is a Julia package that can wrap Zarr arrays in a DimensionalData-compatible interface.