Translator Disclaimer
8 September 2015 TIRS stray light correction: algorithms and performance
Author Affiliations +
The Thermal Infrared Sensor (TIRS) onboard Landsat 8 was tasked with continuing thermal band measurements of the Earth as part of the Landsat program. From first light in early 2013, there were obvious indications that stray light was contaminating the thermal image data collected from the instrument. Traditional calibration techniques did not perform adequately as non-uniform banding was evident in the corrected data and error in absolute estimates of temperature over trusted buoys sites varied seasonally and, in worst cases, exceeded 9 K error. The development of an operational technique to remove the effects of the stray light has become a high priority to enhance the utility of the TIRS data. This paper introduces the current algorithm being tested by Landsat’s calibration and validation team to remove stray light from TIRS image data. The integration of the algorithm into the EROS test system is discussed with strategies for operationalizing the method emphasized. Techniques for assessing the methodologies used are presented and potential refinements to the algorithm are suggested. Initial results indicate that the proposed algorithm significantly removes stray light artifacts from the image data. Specifically, visual and quantitative evidence suggests that the algorithm practically eliminates banding in the image data. Additionally, the seasonal variation in absolute errors is flattened and, in the worst case, errors of over 9 K are reduced to within 2 K. Future work focuses on refining the algorithm based on these findings and applying traditional calibration techniques to enhance the final image product.
© (2015) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Aaron Gerace, Matthew Montanaro, Tim Beckmann, Kaitlin Tyrrell, Alexandra Cozzo, Trevor Carney, and Vicki Ngan "TIRS stray light correction: algorithms and performance", Proc. SPIE 9607, Earth Observing Systems XX, 96070Q (8 September 2015);

Back to Top