New openEO backend for openEO Platform
We are pleased to announce that the capabilities of openEO Platform are growing. We are in the process of onboarding the Sentinel Hub openEO backend within the federation and invite you to test it already and give us feedback.
How to test?
At this stage, the SH openEO backend is integrated into the development backend of the platform. You can connect to it via https://openeocloud-dev.vito.be. To test processes explicitly on the SH openEO backend you either need to select one of the 21 collections that are only supported there (e.g. CNES Land Cover Mao, ERA5 Wind, Sentinel-5 14-days average…) or, if you are using the Python client, specify provider:backend in the properties of your load_collection process (see code snippet below).
# Load the data cube from the S2 L2A collection for a 3-week period in July 2021 (5 acquisitions)
cube = connection.load_collection(
"SENTINEL2_L2A_SENTINELHUB",
spatial_extent={"west": 14.5830, "south": 46.1034, "east": 14.5999,
"north": 46.1122},
temporal_extent=["2021-07-09", "2022-07-30"],
bands=["B04", "B08"],
properties={"provider:backend":lambda v: v =="sentinelhub"}
)
*If you want to connect directly to the SH openEO backend you can do this via: https://openeo.sentinel-hub.com/production/
Where can I give feedback?
You can use this forum post for general feedback or create a new entry in the forum for specific problems (tag @daniel.thiex)
Supported processes
Currently not all of the platform defined core processes are supported by the Sentinel Hub openEO backend*. If any of these processes would be useful for your use case, let us know and we will see if it can be supported.
*not supported: aggregate_spatial, apply_kernel, load_result, mask_polygon, rename_dimension, resample_cube_spatial, array_modify, nan, is_infinite