s21·mission
Sentinel-2
Europe's 10-metre true-colour land camera
active DAAC: none (ESA mission — not distributed by a NASA DAAC) landwateragriculturebiosphere
Sentinel-2
ESA’s optical workhorse for the Copernicus program — a two-satellite constellation that images all of Earth’s land (and coastal/inland water) in true colour at 10 m, every five days. Not a NASA mission and not in NASA’s DAACs, but the data are free and open, and the atlas uses them anywhere you need finer detail than Landsat or daily MODIS/VIIRS can give.
What it sees
- MSI (MultiSpectral Instrument) — 13 spectral bands spanning visible, near-infrared, and shortwave-infrared.
- Three native resolutions: 10 m (blue, green, red, NIR), 20 m (red-edge, SWIR — vegetation + moisture), 60 m (atmospheric correction bands).
- Wide 290 km swath — the secret to the fast 5-day global revisit.
Why it matters
- 10 m, free, every 5 days — the highest-resolution global optical record anyone can download without paying.
- The red-edge bands (missing on Landsat) make it exceptional for crop type, plant health, and stress detection.
- Pairs naturally with Landsat (the Harmonized Landsat Sentinel-2, HLS, product fuses both into one 30 m every-2-3-day series).
Where to get the data
- Copernicus Data Space Ecosystem:
dataspace.copernicus.eu— the official ESA source, full archive + processing APIs. - AWS Open Data:
s3://sentinel-2-l2a-cogs— cloud-optimized GeoTIFFs, no egress fees. - Element84 Earth Search: STAC API at
earth-search.aws.element84.com— best for programmatic search. - Microsoft Planetary Computer: STAC + hosted compute.
What it enables
- Land-cover and land-use mapping at field scale
- Crop monitoring, yield estimation, irrigation tracking (agriculture)
- Deforestation and forest-disturbance detection (biosphere)
- Inland and coastal water-quality / algal-bloom monitoring (water)
- Disaster mapping (floods, fire scars) at higher detail than daily sensors
Gotchas
- Optical means cloud-limited. Unlike Sentinel-1 radar, clouds block the view; cloudy regions get sparse usable scenes.
- L1C vs L2A. L1C is top-of-atmosphere; almost everyone wants L2A (surface reflectance, atmospherically corrected). Older archive segments may need processing to L2A yourself.
- Tiling system. Data come in 100×100 km MGRS tiles; an area of interest can straddle several tiles.
- Band resolutions differ. You must resample to combine 10 m, 20 m, and 60 m bands.
Related missions
- Landsat 9 (NASA+USGS): coarser (30 m) but with thermal bands and a 50-year archive; harmonized with Sentinel-2 via HLS.
- Sentinel-3 (ESA): coarser, wider-swath ocean/land colour companion in the Copernicus family.
- MODIS / VIIRS (NASA): daily global coverage, but much coarser pixels.
Related datasets
HLSL30
Sharp land snapshots (Landsat)
Harmonized Landsat 30m
LP DAAC · 30 m
HLSS30
Sharp land snapshots (Sentinel-2)
Harmonized Sentinel-2 30m
LP DAAC · 30 m (resampled from 10/20/60 m Sentinel-2 native)
MOD13Q1
How green the land is
MODIS Terra Vegetation Indices 16-Day 250m
LP DAAC · 250 m
MOD11A1
How hot the ground gets each day
MODIS Terra Land Surface Temperature/Emissivity Daily 1km
LP DAAC · 1 km
MCD12Q1
What covers the land (forest, city, crops)
MODIS Combined Land Cover Yearly 500m
LP DAAC · 500 m
MCD64A1
Where land has burned
MODIS Combined Burned Area Monthly 500m
LP DAAC · 500 m
📚 Problem Finder KB
Not yet tracked in the KB.