It has been known and documented that the Thermal Infrared Sensor (TIRS) on-board Landsat 8 suffers from a significant stray light problem (Reuter et al., 2015; Montanaro et al., 2014a). The issue appears both as a non-uniform banding artifact across Earth scenes and as a varying absolute radiometric calibration error. A correction algorithm proposed by Montanaro et al. (2015) demonstrated great potential towards removing most of the stray light effects from TIRS image data. It has since been refined and will be implemented operationally into the Landsat Product Generation System in early 2017. The algorithm is trained using near-coincident thermal data (i.e., Terra/MODIS) to develop per-detector functional relationships between incident out-of-field radiance and additional (stray light) signal on the TIRS detectors. Once trained, the functional relationships are used to estimate and remove the stray light signal on a per-detector basis from a scene of interest. The details of the operational stray light correction algorithm are presented here along with validation studies that demonstrate the effectiveness of the algorithm in removing the stray light artifacts over a stressing range of Landsat/TIRS scene conditions. Results show that the magnitude of the banding artifact is reduced by half on average over the current (uncorrected) product and that the absolute radiometric error is reduced to approximately 0.5% in both spectral bands on average (well below the 2% requirement). All studies presented here indicate that the implementation of the stray light algorithm will lead to greatly improved performance of the TIRS instrument, for both spectral bands.
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 4.0 International License.
Department, Program, or Center
Chester F. Carlson Center for Imaging Science (COS)
Gerace, A., Montanaro, M., 2017. Derivation and validation of the stray light correction al-gorithm for the thermal infrared sensor onboard Landsat 8. Remote Sens. Environ.191, 246–257. https://doi.org/10.1016/j.rse.2017.01.029.
RIT – Main Campus