SAR Satellite Data for Mapping: Step-by-Step Guide
This guide walks you through accessing and interpreting Synthetic Aperture Radar (SAR) satellite data for a mapping project. You’ll learn where to find free SAR imagery, how to handle the data, and how to interpret the radar signal to create useful maps—even through clouds or at night. No prior radar experience is needed, but basic GIS knowledge helps.
What You’ll Need
- A computer with internet access
- An Earth observation data platform account (e.g., NASA Earthdata, Copernicus Open Access Hub, or Alaska Satellite Facility) – free registration required
- GIS software: QGIS (free) or ArcGIS (paid) with SAR plugins/extensions
- Optional: Python with libraries like GDAL, rasterio, or pyroSAR for advanced processing
- Basic understanding of raster data and map projections
Step-by-Step
- Define your mapping goal and area of interest (AOI). Decide what you need the SAR data for: land cover classification, flood detection, ground movement (InSAR), or something else. This determines which SAR product and polarization you need. Draw a polygon or bounding box of your AOI.
- Choose a SAR satellite and data product. Common free SAR sources include Sentinel-1 (C-band, 10 m resolution), ALOS-2 PALSAR-2 (L-band), and NASA-ISRO NISAR (launching soon). For most mapping projects, Sentinel-1 is the go-to due to its open access and frequent revisit (6–12 days). Select a product level: Level-0 (raw), Level-1 (Single Look Complex or Ground Range Detected), or Level-2 (orthorectified). For mapping, start with Level-1 GRD.
- Access the data portal. Go to the appropriate platform based on your chosen satellite. For Sentinel-1, use the Copernicus Data Space Ecosystem or the Alaska Satellite Facility (ASF) Vertex. For ALOS-2, use ASF Vertex as well. Create an account and log in.
- Search for scenes over your AOI. In the portal, draw your AOI on the map or upload a shapefile. Set the date range to cover your study period. Filter by satellite, product type, and polarization (e.g., VV+VH for Sentinel-1). Preview the footprint and choose scenes with minimal cloud cover (SAR is unaffected by clouds, but check for noise).
- Download the data. Select the scenes you need and download them. For Sentinel-1, you’ll get a .zip file containing the .tiff image and metadata. For ALOS-2, you may need to request data via a different process. Ensure you have enough storage—each scene can be 1–4 GB.
- Preprocess the SAR data. Open the data in your GIS software. SAR data requires several corrections before analysis:
- Radiometric calibration: Convert digital numbers to backscatter coefficients (sigma0 or gamma0).
- Speckle filtering: Apply a filter (e.g., Lee or Refined Lee) to reduce the salt-and-pepper noise inherent in SAR.
- Terrain correction: Use a digital elevation model (DEM) to correct geometric distortions caused by topography.
- Reprojection: Set the coordinate reference system (CRS) to match your project (e.g., UTM zone).
Many tools automate this, such as SNAP (ESA’s free software) or the Sentinel-1 Toolbox in QGIS.
- Interpret the backscatter values. SAR measures the intensity of radar signal reflected back to the satellite. In the resulting image, bright areas indicate strong reflection (e.g., buildings, rough surfaces, wet soil), while dark areas indicate weak reflection (e.g., smooth water, dry sand). For mapping, you’ll often use the backscatter coefficient (sigma0) in dB.
- Apply classification or change detection. Depending on your goal:
- For land cover: Use the backscatter values to classify pixels into water, urban, forest, or agriculture using thresholds or machine learning.
- For flood detection: Compare a pre-event image with a flood image; water appears as a dark region in the flood image that was not dark before.
- For InSAR: Use phase information from two or more images to measure ground displacement.
- Validate and export your map. Compare your results with ground truth data or higher-resolution imagery. If they match, export your final map as a GeoTIFF or shapefile for use in your project.
Common Mistakes
- Using uncalibrated data: Raw digital numbers are not comparable across scenes. Always calibrate to backscatter.
- Ignoring speckle: Without filtering, noise can be mistaken for features.
- Forgetting terrain correction: In mountainous areas, layover and shadow can distort the image.
- Mixing up polarizations: VV and VH respond differently to surfaces; keep them separate unless you have a reason.
- Choosing the wrong product type: Level-1 SLC is needed for InSAR, but GRD is simpler for backscatter mapping.
Tips
- Use the ASF Vertex search tool—it’s user-friendly and has a tutorial.
- For Sentinel-1, download the SNAP software from ESA; it has built-in processing graphs for SAR.
- Check the incidence angle: Steeper angles (near range) have different backscatter than shallow angles (far range). For consistent mapping, select scenes with similar incidence angles.
- If you need high temporal resolution, combine data from ascending and descending orbits.
- Store downloaded scenes in a structured folder with date and orbit info.
Troubleshooting
- Data won’t download: Check your internet connection, clear browser cache, or try a different browser. Some portals require a download manager for large files.
- Image looks all black or white: This is often a display issue. Stretch the histogram in your GIS software (e.g., apply a 2% linear stretch).
- Terrain correction fails: Ensure you have a DEM that covers your AOI. If using SNAP, download the SRTM DEM within the tool.
- Backscatter values are negative: That’s normal when in dB. Convert to linear scale if you need positive values.
- Scene has gaps or stripes: This may be due to scan mode (e.g., Sentinel-1 IW mode has swaths). Use a different acquisition mode or combine multiple scenes.
Frequently Asked Questions
- Do I need a paid subscription to access SAR data? No, Sentinel-1 data is freely available from Copernicus, and ALOS-2 data is free for research via ASF.
- Can SAR see through clouds? Yes, SAR uses microwave radiation, which penetrates clouds and works day and night.
- What is the best polarization for flood mapping? VV polarization is often used because it shows a strong contrast between water (dark) and land (bright).
- How do I convert SAR data to dB? After calibration, the backscatter coefficient is in linear units. Convert to dB using: dB = 10 * log10(sigma0).
- What is the difference between GRD and SLC? GRD is multi-looked, ground-range, and reduced speckle, suitable for backscatter analysis. SLC retains phase information, needed for interferometry (InSAR).
- Can I use SAR data for urban mapping? Yes, urban areas show very bright returns due to double-bounce scattering from buildings. It’s useful for detecting structures, but shadows and layover can be issues.
Related
Method: the standard formula for this calculation, rounded as each result states. See how the calculators work or report an error.