This guide helps you choose the right Sentinel satellite data for ocean, land, or climate research. You’ll learn what each Sentinel mission measures, how to match data products to your research question, and how to access them. By the end, you’ll be able to pick the correct sensor, product level, and resolution for your study.
What You’ll Need
- A clear research question (e.g., monitoring sea surface temperature, land cover change, or atmospheric aerosols).
- Basic familiarity with remote sensing concepts (spatial, temporal, spectral resolution).
- An account on the Copernicus Open Access Hub or another data portal (free registration).
- Software to view and process satellite data (e.g., QGIS, SNAP, Python with rasterio).
Step-by-Step
- Define your research variable and required accuracy. Write down exactly what you need to measure (e.g., chlorophyll-a concentration, soil moisture, ice extent). Note the required spatial resolution (how small a feature you need to see), temporal frequency (how often you need new data), and spectral bands (which wavelengths are relevant).
- Match your variable to a Sentinel mission. Each Sentinel has a specific focus:
- Sentinel-1 (radar): all-weather, day/night imaging. Use for ocean wind/waves, ship detection, land subsidence (InSAR), flood mapping, and ice motion.
- Sentinel-2 (optical high-res): 10-60 m resolution, 13 bands. Use for land cover, vegetation indices (NDVI), agriculture, coastal water quality, and glacier monitoring.
- Sentinel-3 (optical/thermal + radar altimetry): medium resolution (300 m for optical, ~1 km for thermal). Use for sea surface temperature, ocean color (chlorophyll), land surface temperature, and sea level/ice thickness (altimetry).
- Sentinel-5P (atmospheric): measures trace gases (NO2, SO2, O3, CO, CH4) and aerosols. Use for air quality and climate forcing studies.
- Sentinel-6 (radar altimetry): high-precision sea level measurement for ocean circulation and climate.
- Choose the data product level. Sentinel data comes in several processing levels:
- Level-0/1: raw or radiometrically/geometrically corrected. Use if you need to apply your own corrections.
- Level-2: geophysical products (e.g., sea surface temperature, chlorophyll, soil moisture). Use these for most research—they are ready to analyze.
- Level-3/4: gridded, merged, or reanalyzed products (e.g., daily global SST). Use for climate time series.
- Check spatial and temporal resolution against your needs. For example, Sentinel-2 has a 5-day revisit at the equator (with two satellites), but cloud cover can reduce usable images. Sentinel-3 has a daily revisit for ocean color but coarser resolution. If you need daily high-res land data, consider combining Sentinel-2 with Landsat.
- Select the specific product and band combination. For each mission, browse the product catalogue (e.g., Sentinel-2 MSI Level-2A). Choose the bands that correspond to your variable: e.g., for NDVI use bands 4 (red) and 8 (NIR). For ocean color, use Sentinel-3 OLCI bands 1-21.
- Access and download the data. Use the Copernicus Data Space Ecosystem (newer) or the legacy Open Access Hub. Draw your area of interest, set the date range, and filter by mission, product level, and cloud cover (for optical). Download the granules you need.
- Validate data quality. Check for cloud cover, missing lines, or processing anomalies. For optical data, use the scene classification layer. For radar, check for orbit and calibration accuracy.
- Process and analyze. Use SNAP (ESA’s toolbox) or Python to open, subset, and apply any needed corrections. For Level-2 products, you can directly compute indices or extract time series.
Common Mistakes
- Ignoring cloud cover in optical data—this can ruin a time series. Always filter by cloud percentage or use radar for cloudy regions.
- Mixing product levels (e.g., comparing Level-1 reflectance with Level-2 surface reflectance) without proper conversion.
- Choosing too coarse resolution for small-scale features (e.g., using Sentinel-3 for a 10 m river).
- Overlooking the need for atmospheric correction when using Level-1 data for land/ocean color.
- Assuming all Sentinel-2 bands have the same resolution—some are 20 m or 60 m; resample carefully.
Tips
- For climate studies, use Level-3 or Level-4 products from the Copernicus Climate Data Store—they are already homogenized.
- Combine Sentinel-1 (radar) with Sentinel-2 (optical) for flood mapping: radar sees through clouds, optical gives land cover.
- Use the Copernicus Data Space API for bulk downloads—saves time for large areas.
- Check the mission status pages for known issues (e.g., Sentinel-1A/1B anomalies).
- For ocean research, prefer Sentinel-3 OLCI and SLSTR over Sentinel-2 because of better spectral bands and daily revisit.
Troubleshooting
- Data not found for your area/date: Widen the date range or area. Some regions have sparse coverage due to orbit patterns or cloud cover.
- Download fails: Use the API or try a different portal (e.g., AWS open data). Check your internet connection and storage space.
- Product won’t open in your software: Ensure you have the correct plugin (e.g., SNAP for Sentinel-3). Try converting to GeoTIFF with the SNAP command line.
- Values look wrong (e.g., negative reflectance): Check the scaling factor—many products require a scale factor (e.g., 0.0001 for Sentinel-2).
- Time series has gaps: Use gap-filling methods or combine with other satellites (e.g., Landsat).
Frequently Asked Questions
- What is the difference between Sentinel-2 and Landsat? Sentinel-2 has higher spatial resolution (10 m vs 30 m) and more spectral bands, but Landsat has a longer historical record (since 1972). Both are useful for land studies.
- Can I use Sentinel-1 for ocean color? No, Sentinel-1 is a radar sensor—it doesn’t measure ocean color. Use Sentinel-3 OLCI or Sentinel-2 for that.
- How do I get daily sea surface temperature data? Use Sentinel-3 SLSTR Level-2 products, which have a daily revisit. For global coverage, use the Copernicus Climate Data Store’s SST product.
- Is Sentinel-5P data suitable for local air quality studies? Yes, but the spatial resolution is coarse (7 km). For city-scale studies, combine with ground measurements or use higher-resolution models.
- What is the best product for land cover classification? Sentinel-2 Level-2A with 10 m resolution is ideal. Use the 10 m bands (2,3,4,8) and apply a supervised classification.
- Do I need to do atmospheric correction? If you use Level-2 products, no—they are already corrected. For Level-1, yes, especially for ocean color or vegetation indices.
Related
Method: the standard formula for this calculation, rounded as each result states. See how the calculators work or report an error.
More cards in this drawer
- Data Center Security: Encryption, Backups, Access
- Best Mini PCs for Gaming: Top Picks
- Phone Buttons: Clever Ways to Use Them
- Resonance Solstice Codes: What They Are
- Blue Protocol Star Resonance: Best Graphics Settings for Smooth FPS
- Jack Daniel’s Ransomware: Data Stolen?