Quantifying the Relationship Between Intersensor Images in Solar Reflective Bands: Implications for Intercalibration

Abstract

Satellite instruments have acquired large volume images at different spatial, spectral, radiometric, and temporal resolutions. Reliable detection of long-term environmental change requires critical sensor intercalibration destined to minimize inconsistency in these multisensor images. However, uncertainty in intercalibration has not yet been comprehensively quantified in most existing studies. This paper developed a quantitative relationship between multisensor images in solar reflective bands by accounting for sensor difference, atmospheric condition, and Sun-target-sensor geometry. The relationship was validated with collocated and concurrent TERRA MODIS/NOAA-17 AVHRR images over the Dunhuang calibration site. Then, it was used to investigate sensitivity of intercalibration to intersensor scale factor, total ozone concentration (TOC), total precipitable water vapor content (TPW), and aerosol optical thickness (AOT). The main conclusions include: 1) error in intersensor scale factor may induce a maximum uncertainty of 2.35% for both visible (VIS) and near-infrared (NIR) bands; 2) error in TOC can produce a maximum uncertainty of 0.30% for VIS band but very minor impact on NIR band; 3) error in TPW may generate a maximum uncertainty of 8.81% for NIR band, particularly for a dry atmosphere; 4) error in AOT can result in a maximum uncertainty of 0.25% for VIS band and 0.96% for NIR band at near-nadir, and 10.16% and 8.18% for heavy aerosol loadings at a very high solar angle. The following study quantifies uncertainties in intercalibration for solar reflective bands and thus offers guidance for intercalibration.