Abstract
The Operational Land Imager (OLI) instruments on Landsat 8 and Landsat 9 collect hundreds of Earth scenes per day, covering a wide mix of surface types and signal levels. This report demonstrates that detector overlap statistics can be used to directly characterize non-uniformity at the focalplane boundary zones inherent to the OLI focalplane design and correct for these non-uniformities in image data products. Detector pairs of coregistered pixels within boundary zone represent measurements of the same ground target collected at the same time but with minor view angle differences. The use of multiple pairs of detectors offers varying scene types that provide robustness to the uniformity analysis and help mitigate any surface directional reflectance effects. The scene-based banding uniformity characterization is shown to be at the level of < 1% (1sigma) over the mission lifetime. Results from the approach are shown to be comparable to those obtained by the onboard solar demonstrating that this analysis method can be applied to similarly staggered focalplane sensor chips. Overall, this approach describes a simple method for reliably monitoring detector level uniformity in pushbroom imagers and potentially leading to autocorrection for banding artifacts.
Image uniformity evaluation of Landsat OLI using detector overlap statistics of Earth scenes
The Operational Land Imager (OLI) instruments on Landsat 8 and Landsat 9 collect hundreds of Earth scenes per day, covering a wide mix of surface types and signal levels. This report demonstrates that detector overlap statistics can be used to directly characterize non-uniformity at the focalplane boundary zones inherent to the OLI focalplane design and correct for these non-uniformities in image data products. Detector pairs of coregistered pixels within boundary zone represent measurements of the same ground target collected at the same time but with minor view angle differences. The use of multiple pairs of detectors offers varying scene types that provide robustness to the uniformity analysis and help mitigate any surface directional reflectance effects. The scene-based banding uniformity characterization is shown to be at the level of < 1% (1sigma) over the mission lifetime. Results from the approach are shown to be comparable to those obtained by the onboard solar demonstrating that this analysis method can be applied to similarly staggered focalplane sensor chips. Overall, this approach describes a simple method for reliably monitoring detector level uniformity in pushbroom imagers and potentially leading to autocorrection for banding artifacts.