Geometric calibration is the determination, for an Earth observation instrument and its platform, of the geometric model that links each detector to a viewing direction, so that image positions can be placed on the Earth’s surface.12
The USGS describes it as driven by geometric characterisation of spacecraft, instrument and image product, and as involving calibration of the spacecraft and instruments.1 The instrument line of sight is derived from static focal plane characteristics combined with telemetry.1 For Sentinel-2, an automatic correlation against a reference image supplies ground control points that refine the viewing model for changes in attitude, position or thermoelastic deformation.2
Calibration is separable from the correction applied to each scene. Geometric correction uses telemetry and external resources to remove small random errors in the inputs.1 The Landsat systematic product applies only ephemeris-based corrections, whereas the precision and terrain product is orthorectified with ground control points and a digital elevation model.3 NASA lists geometric calibration coefficients among the ancillary information annotated on reconstructed instrument data.4
Landsat and Sentinel-2 practice is all the retained sources describe. They do not establish a mission-independent calibration procedure, and no accuracy figure is stated here because the accuracies in those sources belong to particular missions and processing tiers.
Footnotes
-
USGS, “Landsat Geometric and Radiometric Calibration and Characterization”. https://pubs.usgs.gov/publication/fs20243039/full ↩ ↩2 ↩3 ↩4
-
ESA, “SP-1322/2 Sentinel-2”. https://sentinels.copernicus.eu/documents/247904/349490/S2_SP-1322_2.pdf ↩ ↩2
-
USGS, “ECCOE Landsat quarterly Calibration and Validation report”. https://pubs.usgs.gov/publication/ofr20231044/full ↩
-
NASA Earthdata, “Data Processing Levels”. https://www.earthdata.nasa.gov/learn/earth-observation-data-basics/data-processing-levels ↩