Revisiting Lookup Table-Based Atmospheric Correction for Sentinel-2: Making AOT Retrieval Observable and Analyzable
Principal Investigator(s)
Eon Rehman
Aaron Gerace
Research Team Members
Fei Zhang
Michael Gartley
Amirhossein Hassanzadeh
Project Description
This project revisits lookup table–based atmospheric correction for Sentinel-2 by making aerosol optical thickness (AOT) retrieval transparent and analyzable. Unlike operational processors that primarily expose final products, the developed modular pipeline reconstructs the lookup table, forward model, and inversion process, allowing users to inspect vegetation-based pixel selection, per-pixel AOT estimates, spatial aggregation, and retrieved surface reflectance. The pipeline has been completed and delivered, with its retrieval workflow tested and validated across more than a dozen Sentinel-2 image patches.
Figures and Images
Comparison of the Sentinel-2 Level-2A RGB composite (left) and the corresponding RGB composite produced using AOT retrieved by the developed pipeline (right), illustrating the effect of atmospheric correction.