Working with Landsat & Sentinel Data
**Landsat** and **Sentinel-2** are the two most widely used free optical satellite programs in remote sensing. Knowing how to find, download, and process their data is a core skill for any GIS/RS practitioner. ---
📋 Prerequisites
- Registered Google Earth Engine account.
- Access to GEE Code Editor (code.earthengine.google.com).
Step-by-Step Instructions
Understand the Two Programs
- **Landsat (NASA/USGS):** Running continuously since 1972, offering the longest historical satellite record — ideal for long-term change studies - **Sentinel-2 (ESA/Copernicus):** Launched 2015, higher resolution (10m) and faster revisit (5 days) — ideal for detailed, frequent monitoring ---
Download Landsat Data
1. Go to **USGS EarthExplorer** (earthexplorer.usgs.gov) and create a free account 2. Under **Search Criteria**, define your area of interest (draw on map or enter coordinates) 3. Set your date range 4. Under **Data Sets**, select **Landsat → Collection 2 Level-2** (already atmospherically corrected) 5. Click **Results**, review cloud cover %, and download your chosen scene ---
Download Sentinel-2 Data
1. Go to the **Copernicus Data Space Ecosystem** (dataspace.copernicus.eu) and create a free account 2. Search your area of interest and date range 3. Filter by **Sentinel-2**, and choose **Level-2A** products (already atmospherically corrected, ready to use) 4. Download the product for your selected date ---
Load and Composite Bands in QGIS
Both Landsat and Sentinel-2 come as separate band files. To view a natural color image: 1. **Layer → Add Layer → Add Raster Layer** — load the Red, Green, and Blue bands 2. Use **Raster → Miscellaneous → Build Virtual Raster** (or **Merge**) to combine them into one multi-band file 3. In Symbology, assign: Band 1 → Red, Band 2 → Green, Band 3 → Blue 4. For a **False Color Composite** (great for vegetation), use NIR-Red-Green instead ---
Know Your Band Numbers
| Band | Landsat 8/9 | Sentinel-2 | |---|---|---| | Blue | Band 2 | Band 2 | | Green | Band 3 | Band 3 | | Red | Band 4 | Band 4 | | NIR | Band 5 | Band 8 | | SWIR 1 | Band 6 | Band 11 | | SWIR 2 | Band 7 | Band 12 | ---
Filter for Cloud-Free Imagery
Cloud cover ruins analysis. When searching: - Set a **maximum cloud cover filter** (e.g., under 10%) in EarthExplorer or Copernicus Browser - Visually inspect the preview thumbnail before downloading - For time series work, consider using Google Earth Engine to automatically select the least-cloudy image over a date range (see the GEE tutorial) ---
Choose the Right Dataset for Your Task
- **Need historical data (pre-2015)?** → Landsat - **Need high detail / frequent updates?** → Sentinel-2 - **Need all-weather / cloud-penetrating data?** → Sentinel-1 (SAR) ---
🧠 Quick Recap & Practice Questions
## Quick Recap
- Landsat: longest historical record, 30m resolution, 16-day revisit
- Sentinel-2: 10m resolution, 5-day revisit, ESA/Copernicus
- Always use Level-2/Level-2A products (already atmospherically corrected)
- Filter by cloud cover before downloading
---
## Practice Questions
1. Which program would you choose for a 20-year land change study?
2. What's the resolution difference between Landsat and Sentinel-2?
3. Why should you prefer Level-2/Level-2A products over Level-1?
---
*Next Tutorial: Map Layout & Cartographic Design in QGIS*