Spatial resamping using Resample Operators with OpenEO

Resample Operators

Sometimes we need to align the spatial or temporal dimension of two datacubes, so that they can be merged together.

resample_cube_spatial: spatial resampling Sentinel-2 to match Landsat

Start importing the necessary libraries and initialize a local connection for Client-Side Processing.

import openeo
from openeo.local import LocalConnection
local_conn = LocalConnection('')

Create two datacubes, one for Sentinel-2 and one for Landsat

url = "https://earth-search.aws.element84.com/v1/collections/sentinel-2-l2a"

spatial_extent = {"west": 11.4, "east": 11.42, "south": 45.5, "north": 45.52}
temporal_extent = ["2023-06-01", "2023-06-30"]
bands = ["red","nir"]

s2_cube = local_conn.load_stac(url=url,
   spatial_extent=spatial_extent,
   temporal_extent=temporal_extent,
   bands=bands
)
s2_cube.execute()

url = "https://planetarycomputer.microsoft.com/api/stac/v1/collections/landsat-c2-l2"
bands = ["red","nir08"]
l8_cube = local_conn.load_stac(url=url,
                    spatial_extent=spatial_extent,
                    temporal_extent=temporal_extent,
                    bands=bands)
l8_cube.execute()

From the previous outputs, notice the shape difference in the spatial dimensions x and y.

This is due to the different resolution of the two collections: 10m for Sentinel-2, 30m for Landsat.

Now we use the resample_cube_spatial process to resample the Sentinel-2 data to match Landsat.

s2_cube_30m = s2_cube.resample_cube_spatial(target=l8_cube,method="average")
s2_cube_30m

Check what happens to the datacube inspecting the resulting xArray object. Now the x and y shape is the same as Landsat:

s2_cube_30m.execute()