Skip to content

Getting Sentinel-2 data

fetch_sentinel2

pip install "synapse-sr[stac]"

import synapse_sr

from synapse_sr import fetch_sentinel2

path = fetch_sentinel2(lat=30.935, lon=75.800, start="2024-11-01", end="2025-01-31", size_m=3000)

fetch_sentinel2 searches the public Earth Search STAC catalogue. It picks the least-cloudy L2A scene over the point and writes:

  • a 10-band GeoTIFF in model order, with named bands, on the 10 m grid (20 m bands nearest-resampled);
  • the offset in the BOA_ADD_OFFSET tag, plus the STAC item id and cloud cover;
  • <name>_scl.tif, the scene classification layer.

super_resolve picks up the offset and the SCL file automatically. Pass api= to use another STAC API that serves the sentinel-2-l2a collection with Earth Search asset names.

Other sources

Source Notes
Copernicus Data Space (SAFE) stack B02-B12 onto the 10 m grid with nearest resampling; pass offset=-1000 for baseline 04.00 or later
Google Earth Engine COPERNICUS/S2_SR_HARMONIZED export at scale=10; the offset is already removed
cubo or stackstac cubes pass cube.isel(time=k) directly; band names are read from the band coordinate
odc-stac datasets pass ds.isel(time=k).to_array("band")