Written and maintained by CASRAI Editorial Board
Last updated
Sentinel-2 is a European wide-swath, high-resolution, multispectral Earth observation mission developed and operated under the European Union Copernicus programme. The mission delivers systematic optical imagery across global land surfaces and coastal waters to support vegetation monitoring, crop assessment, water quality tracking, and disaster response. Sentinel-2 data is accessible worldwide at no charge under the European Union full, free, and open data policy.
In modern remote sensing and geological research, Sentinel-2 serves as a cornerstone optical imaging system alongside the USGS/NASA Landsat constellation. With its high revisit frequency and specialized red-edge spectral bands, researchers frequently ingest Sentinel-2 scenes into cloud computing platforms like Google Earth Engine, climate repositories such as the Copernicus Climate Data Store, and data hubs like the PANGAEA Data Publisher to track biophysical dynamics at high spatial and temporal fidelity.
What Sentinel-2 Is
The Sentinel-2 mission consists of a constellation of identical, sun-synchronous, polar-orbiting satellites flying in the same orbital plane at an altitude of approximately 786 kilometers, phased at 180° to each other. This orbital configuration produces a high revisit frequency of five days at the equator and two to three days across mid-latitudes under cloud-free conditions.
Each spacecraft carries a single optical payload: the Multi-Spectral Instrument (MSI). The MSI captures incoming radiance using a push-broom sensor architecture across a wide swath width of 290 kilometers. The instrument records electromagnetic radiation across thirteen discrete spectral channels spanning the visible, near-infrared (NIR), and shortwave-infrared (SWIR) spectrum, resolving features on the ground at spatial resolutions of 10 meters, 20 meters, and 60 meters.
Who Runs Sentinel-2 and What It Covers
The mission is managed and financed by the European Commission as part of the Copernicus programme, with space segment development, satellite manufacturing, launch operations, and orbital management directed by the European Space Agency (ESA). Spacecraft operations are conducted from ESA’s European Space Operations Centre (ESOC) in Darmstadt, Germany, while payload ground segment data processing and distribution are managed via the European Space Research Institute (ESRIN) in Frascati, Italy.
Sentinel-2 systematically images:
- All continental land surfaces between latitudes 56° South and 84° North.
- All coastal waters up to 20 kilometers from the shoreline.
- All islands greater than 100 square kilometers in area, along with all European islands and small Mediterranean islands.
- Closed seas, inland water bodies, and major river basins.
The constellation launched with Sentinel-2A on June 23, 2015, followed by Sentinel-2B on March 7, 2017. To guarantee mission continuity, Sentinel-2C was successfully launched on September 5, 2024, maintaining uninterrupted operational data collection across the global scientific community.
Sentinel-2 Spectral Bands and Spatial Resolution
The MSI instrument splits incoming optical radiation into thirteen spectral bands. The bands are categorized into three distinct ground sampling distances (spatial resolutions):
| Band | Central Wavelength | Bandwidth | Spatial Resolution | Primary Scientific Application |
|---|---|---|---|---|
| B1 | 443 nm | 20 nm | 60 m | Coastal aerosol detection and atmospheric correction |
| B2 | 490 nm | 65 nm | 10 m | Blue visible spectrum; soil and vegetation discrimination, bathymetry |
| B3 | 560 nm | 35 nm | 10 m | Green visible spectrum; green vegetation reflectance and vigor |
| B4 | 665 nm | 30 nm | 10 m | Red visible spectrum; chlorophyll absorption peak in vegetation |
| B5 | 705 nm | 15 nm | 20 m | Vegetation red edge 1; monitoring plant stress and leaf chlorophyll content |
| B6 | 740 nm | 15 nm | 20 m | Vegetation red edge 2; leaf area index (LAI) and canopy structure |
| B7 | 783 nm | 20 m | 20 m | Vegetation red edge 3; biomass estimation and vegetation boundary demarcation |
| B8 | 842 nm | 115 nm | 10 m | Broad Near-Infrared (NIR); standard vegetation indices (NDVI) and surface water mapping |
| B8A | 865 nm | 20 nm | 20 m | Narrow Near-Infrared; atmospheric water vapor correction and biophysical modeling |
| B9 | 945 nm | 20 nm | 60 m | Water vapor absorption channel for atmospheric correction |
| B10 | 1375 nm | 30 nm | 60 m | Shortwave Infrared (SWIR) cirrus detection and cloud masking |
| B11 | 1610 nm | 90 nm | 20 m | SWIR 1; vegetation moisture stress, snow/ice vs cloud discrimination |
| B12 | 2190 nm | 180 nm | 20 m | SWIR 2; soil mineralogy, burn scar mapping, and geological classification |
Data Products and Atmospheric Correction
Sentinel-2 observations are downlinked as raw data (Level-0) and processed systematically through the Copernicus Ground Segment into standardized user products distributed in SAFE (Standard Archive Format for Europe) format:
- Level-1C (L1C): Represents Top-Of-Atmosphere (TOA) reflectance in fixed cartographic geometry. L1C products are tiled into ortho-images of 110 km × 110 km projected onto the Universal Transverse Mercator (UTM) spatial reference system using the WGS84 datum. The grid uses a 100-kilometer step, providing a 10-kilometer overlap between adjacent tiles. L1C granules include per-pixel radiometric measurements and cloud masks.
- Level-2A (L2A): Represents Bottom-Of-Atmosphere (BOA) surface reflectance. L2A products are generated by applying atmospheric correction algorithms (traditionally via the Sen2Cor processor) to Level-1C scenes. This step corrects for aerosol scattering, cirrus absorption, and terrain illumination effects. Level-2A packages include a 20-meter Scene Classification Map (SCL) that classifies every pixel into distinct categories: cloud high probability, cloud medium probability, cirrus, cloud shadows, vegetation, non-vegetated bare soil, water, snow/ice, or dark features.
How to Access and Download Sentinel-2 Data
Researchers can discover, visualize, and download Sentinel-2 imagery through several official and federated routes:
- Copernicus Data Space Ecosystem (CDSE): Visit
dataspace.copernicus.euand access the Copernicus Browser. Users can define a geographical area of interest (AOI), set cloud-cover thresholds, select Level-1C or Level-2A products, and download scenes directly or generate visual composites. - Programmatic STAC and OData APIs: Query the CDSE catalog programmatically using the SpatioTemporal Asset Catalog (STAC) API or OData API to integrate automated scene ingestion into Python scripts.
- OpenEO and Cloud JupyterLab: Execute server-side band mathematics and time-series aggregations within CDSE hosted JupyterLab environments using openEO syntax without downloading multi-gigabyte zip files locally.
- Google Earth Engine Catalog: Access the harmonized Level-1C and Level-2A global image collections (
COPERNICUS/S2_HARMONIZEDandCOPERNICUS/S2_SR_HARMONIZED) directly within the Earth Engine Code Editor or Python API for cloud-scale zonal reductions. - Public Cloud Open Data Buckets: Ingest Cloud-Optimized GeoTIFF (COG) formatted Sentinel-2 data directly from public Amazon Web Services (AWS) S3 or Microsoft Planetary Computer STAC endpoints.
Strengths and Limitations
Sentinel-2 provides an exceptional combination of spatial detail and spectral depth, but users must evaluate its characteristics against project constraints:
Key Strengths
- Open Access and No Royalties: Complete, full, and open data access for scientific, educational, and commercial purposes with no licensing fees.
- High Spatial Resolution: 10-meter ground resolution in visible and near-infrared bands offers significantly finer spatial detail than traditional 30-meter Landsat channels.
- Dedicated Red-Edge Channels: Three specialized 20-meter red-edge bands provide sensitive indicators of plant chlorophyll dynamics and canopy nitrogen content.
- 5-Day Revisit Schedule: Twin-satellite constellation delivers regular temporal sampling critical for agricultural monitoring and phenology tracking.
Technical Limitations
- Optical Cloud Vulnerability: As a passive optical sensor, Sentinel-2 cannot penetrate cloud cover, requiring users to filter cloudy scenes or fuse observations with Sentinel-1 SAR radar imagery.
- No Thermal Infrared Sensor: Unlike Landsat 8/9, Sentinel-2 does not carry a thermal infrared sensor, making direct land surface temperature (LST) and evapotranspiration calculations impossible from Sentinel-2 alone.
- Storage and Bandwidth Overhead: A single uncompressed Sentinel-2 Level-2A tile is roughly 500 MB to 1 GB in size, requiring substantial bandwidth and local storage for large multi-temporal regional studies.
Alternatives to Sentinel-2
Researchers select optical Earth observation systems based on trade-offs between spatial resolution, spectral coverage, and revisit frequency:
| System | Operator | Spatial Resolution | Revisit Interval | Data License |
|---|---|---|---|---|
| Sentinel-2 (MSI) | European Space Agency (ESA) | 10 m / 20 m / 60 m (13 bands) | 5 days (2-3 days mid-latitudes) | Open and free (Copernicus) |
| Landsat 8 and 9 (OLI-2 / TIRS-2) | USGS / NASA | 30 m optical, 15 m pan, 100 m thermal | 8 days (combined constellation) | Open and free (USGS) |
| PlanetScope | Planet Labs | 3 m to 3.7 m (4 or 8 bands) | Daily global coverage | Commercial proprietary (academic grants available) |
| MODIS (Terra and Aqua) | NASA | 250 m / 500 m / 1000 m (36 bands) | 1 to 2 days | Open and free (NASA) |
Frequently Asked Questions
Is Sentinel-2 still active?
Yes. The Sentinel-2 mission is fully active and operational. The mission is maintained by the European Space Agency and the European Commission, operating with Sentinel-2A (launched 2015), Sentinel-2B (launched 2017), and the newly launched Sentinel-2C (launched September 5, 2024) to ensure long-term mission continuity.
Is Sentinel-2 imagery free?
Yes. Sentinel-2 imagery is free of charge under the European Union Copernicus full, free, and open data policy. Academic researchers, commercial developers, government agencies, and the public can search, view, and download Level-1C and Level-2A data without fees, subscriptions, or commercial licensing royalties.
What is Sentinel-2 used for?
Sentinel-2 is used for land cover classification, agricultural crop health and yield monitoring, forest canopy change detection, inland water quality assessment, urban growth tracking, and natural disaster emergency mapping. Its three red-edge bands make it especially valuable for precision agriculture and vegetation biophysical modeling.
How accurate is Sentinel-2?
Sentinel-2 delivers high radiometric and geometric accuracy. The Multispectral Instrument maintains a radiometric accuracy within 3% to 5%. In terms of geometric accuracy, absolute geolocation without ground control points is better than 12.5 meters at 95% confidence, while multi-temporal co-registration using the Global Reference Image (GRI) achieves an accuracy of better than 0.3 pixels (less than 3 meters).
Sources
- Copernicus Data Space Ecosystem Sentinel-2: https://dataspace.copernicus.eu/explore-data/data-collections/sentinel-data/sentinel-2 (November 2026)
- SentiWiki Sentinel-2 Mission Specification: https://sentiwiki.copernicus.eu/web/s2-mission (November 2026)
- SentiWiki Sentinel-2 Products Specification: https://sentiwiki.copernicus.eu/web/s2-products (November 2026)
- SentiWiki Sentinel-2 Processing Baseline: https://sentiwiki.copernicus.eu/web/s2-processing (November 2026)
