ESA WorldCover — 10 m land cover¶

ESA WorldCover is a global 10 m land-cover map derived from Sentinel-1 and Sentinel-2. This notebook fetches the 2021 (v200) tile for a small AOI over the Gulf of Guinea via the amazon-s3 backend, then reads the cropped GeoTIFF back with pyramids' Dataset.

Setup¶

The imports: tempfile / Path for a scratch download directory, earthlens for the unified EarthLens entry point, and pyramids' Dataset for reading the resulting raster.

In [ ]:

import tempfile
from pathlib import Path

from earthlens import EarthLens
from pyramids.dataset import Dataset

import tempfile from pathlib import Path from earthlens import EarthLens from pyramids.dataset import Dataset

Fetch the land-cover tile¶

WorldCover is a file-writing backend, so we give it a scratch directory to write into. A fresh tempfile.mkdtemp() keeps the download out of the working tree.

In [ ]:

out = Path(tempfile.mkdtemp(prefix='wc_'))

out = Path(tempfile.mkdtemp(prefix='wc_'))

Build the request¶

Construct the EarthLens request first — the amazon-s3 data source, the esa-worldcover dataset, the 2021 date window, and the AOI bounding box. Keeping the construction on its own line makes each parameter easy to read and tweak.

In [ ]:

src = EarthLens(
    data_source="amazon-s3",
    start='2021-01-01',
    end='2021-01-01',
    aoi=[6.40, 0.40, 6.45, 0.45],
    dataset='esa-worldcover',
    path=str(out),
)

src = EarthLens( data_source="amazon-s3", start='2021-01-01', end='2021-01-01', aoi=[6.40, 0.40, 6.45, 0.45], dataset='esa-worldcover', path=str(out), )

Download¶

download() fetches the tile, crops it to the AOI, and returns the list of written output paths.

In [ ]:

paths = src.download(progress_bar=False)
paths

paths = src.download(progress_bar=False) paths

Read the result¶

Read the first written GeoTIFF back with pyramids' Dataset and inspect its bounding box and shape to confirm the crop landed on the requested AOI.

In [ ]:

raster = paths[0]
ds = Dataset.read_file(str(raster))
print('bbox:', [round(x, 3) for x in ds.bbox], 'shape:', ds.shape)

raster = paths[0] ds = Dataset.read_file(str(raster)) print('bbox:', [round(x, 3) for x in ds.bbox], 'shape:', ds.shape)