Baseline Satellite Data
Baseline satellite data consists of remote-sensing products that have been corrected and prepared for analysis — Level-2 (per scene) and Level-3 (temporal composites). These products underpin most analysis on the Piksel platform, including vegetation, land, water, and surface temperature monitoring, as well as change detection.
Piksel provides baseline products from six satellite missions across four different measurement types.
Product Matrix
| Measurement | Landsat 5 | Landsat 7 | Landsat 8 | Landsat 9 | Sentinel-1 | Sentinel-2 |
|---|---|---|---|---|---|---|
| Surface Reflectance | ✓ | ✓ | ✓ | ✓ | — | ✓ |
| Surface Temperature | ✓ | ✓ | ✓ | ✓ | — | — |
| SAR Backscatter | — | — | — | — | ✓ | — |
| Annual Composite | — | — | — | — | — | GeoMAD |
The matrix shows the available options and highlights the gaps — for example, Sentinel-2 has no thermal band, the Landsat family has no SAR, and annual composites are currently available only from Sentinel-2.
Baseline Satellite Sections
- Surface Reflectance — Surface reflectance from four Landsat missions and Sentinel-2. The basis for vegetation analysis (NDVI/EVI), land-cover classification, and optical change detection.
- Surface Temperature — Land surface temperature from four Landsat missions. Used to monitor urban heat islands, drought, vegetation stress, and thermal activity.
- Radar Backscatter — Sentinel-1 C-band SAR data with radiometric terrain correction. Works in all weather, day and night.
- Sentinel-2 GeoMAD — Annual composites (Geometric Median + MAD statistics) from Sentinel-2. A virtually cloud-free annual snapshot.
Choosing a Data Type
- Need to analyse vegetation, spectral indices, or land cover? → Surface Reflectance. Use Landsat for long time series (1984–present) or Sentinel-2 for 10 m resolution and red-edge bands.
- Need surface temperature data for urban heat, drought, or geothermal analysis? → Surface Temperature. All Landsat missions are available; combining L8+L9 provides an 8-day revisit.
- Need data for floods or landslides during heavy cloud cover or the tropical wet season? → Radar Backscatter. It penetrates cloud and does not depend on sunlight.
- Need a cloud-free annual snapshot to analyse conditions across years? → Sentinel-2 GeoMAD. A statistical composite from a full year of acquisitions.
- Combining optical and radar data (sensor fusion)? → Load Surface Reflectance + Radar Backscatter together. Optical data provides colour and vegetation information; radar provides structure and moisture information.
All products in this section are Analysis Ready Data (ARD) — already corrected radiometrically, geometrically, and atmospherically. They can be loaded directly with odc-stac, rioxarray, or other tools without additional preprocessing.