xpystac

xpystac provides the glue that allows xarray.open_dataset to accept pystac objects.

The goal is that as long as this library is in your env, you should never need to think about it.

Example

Search collection of COGs:

import pystac_client
import xarray as xr


catalog = pystac_client.Client.open(
    "https://earth-search.aws.element84.com/v1",
)

search = catalog.search(
    intersects=dict(type="Point", coordinates=[-105.78, 35.79]),
    collections=['sentinel-2-l2a'],
    datetime="2022-04-01/2022-05-01",
)

xr.open_dataset(search, engine="stac")

Here are a few examples from the Planetary Computer Docs

import planetary_computer
import pystac_client
import xarray as xr


catalog = pystac_client.Client.open(
    "https://planetarycomputer.microsoft.com/api/stac/v1",
    modifier=planetary_computer.sign_inplace,
)

Read from a reference file:

collection = catalog.get_collection("nasa-nex-gddp-cmip6")
asset = collection.assets["ACCESS-CM2.historical"]

xr.open_dataset(asset)

ref: https://planetarycomputer.microsoft.com/dataset/nasa-nex-gddp-cmip6#Example-Notebook

Read from a zarr file:

collection = catalog.get_collection("daymet-daily-hi")
asset = collection.assets["zarr-abfs"]

xr.open_dataset(asset)

ref: https://planetarycomputer.microsoft.com/docs/quickstarts/reading-zarr-data/

Install

pip install git+https://github.com/stac-utils/xpystac

How it works

When you call xarray.open_dataset(object, engine="stac") this library maps that open call to the correct library. Depending on the type of object that might be a stacking library (either odc-stac or stackstac) or back to xarray.open_dataset itself but with the engine and other options pulled from the pystac object.

Prior Art

This work is inspired by https://github.com/TomAugspurger/staccontainers and the discussion in stac-utils/pystac#846