`aggregate_spatial` improvements

[clausmichele]

Tasks:

• check aggregate_spatial definition https://processes.openeo.org/#aggregate_spatial: does it accept vector cubes with a different CRS than the raster cube? -> https://github.com/Open-EO/openeo-processes/issues/516 -> https://github.com/Open-EO/openeo-processes/issues/499
• aggregate_spatial implementation: benchmark current xvec implementation versus exactextract:
  • check if results are matching
  • If not : why or how much do they differ
• Does it work with vector cubes with a different projection? Currently xvec works despite the vector cubes has a different projection compared to the raster cube (reprojects it internally)
• For exactextract: what’s the vector cube data format that it accepts as input? Check the docs
• Reference Pangeo forum discussion: https://discourse.pangeo.io/t/advice-for-scalable-raster-vector-extraction/4129/19

Roadmap issue for reference: https://github.com/EOEPCA/roadmap/issues/56

[clausmichele]

Sample code:

import json
import geopandas as gpd
from openeo.local import LocalConnection
local_conn = LocalConnection("./")

url = "https://earth-search.aws.element84.com/v1/collections/sentinel-2-l2a"
spatial_extent =  {"east": 11.40, "north": 46.52, "south": 46.46, "west": 11.25}
temporal_extent = ["2022-06-01", "2022-06-30"]
bands = ["red"]
properties = {"eo:cloud_cover": dict(lt=80)}
s2_datacube = local_conn.load_stac(
    url=url,
    spatial_extent=spatial_extent,
    temporal_extent=temporal_extent,
    bands=bands,
    properties=properties,
)

s2_datacube = s2_datacube.resample_spatial(projection="EPSG:4326",resolution=0.0001).drop_dimension("band")
data = s2_datacube.execute().compute() # Try also without compute

from exactextract import exact_extract
import fiona
import rioxarray
import xarray

polys = fiona.open('./sample_polygons.geojson')

results = exact_extract(data, polys, 'mean')
results

sample_polygons.zip