Calibration of Maxar Constellation Over Libya-4 Site Using MAIAC Technique

Abstract

The very-high-resolution commercial satellite constellation of Maxar offers unique opportunities for a wide range of Earth science research and applications. The key to their widespread and effective use is stable and consistent calibration. In this work, we characterized the long-term calibration trends and cross-calibration coefficients for the four Maxar satellites (GeoEye-1, QuickBird-2, WorldView-2, and WorldView-3) using the Multi-Angle Implementation of Atmospheric Correction (MAIAC) processing technique. Utilizing MAIAC MODIS atmosphere and surface products, we calculated top-of-atmosphere (TOA) reflectance for the Blue, Green, Red, and NIR bands over Libya-4 desert site. To ensure data consistency, we applied geometric normalization to account for variations in TOA reflectance arising from different view geometries. Additionally, a spatial transfer technique was employed to increase the number of samples and yield more robust statistical trend analysis. Our analysis revealed that half of the bands exhibited statistically significant calibration trends. These trends were found to be 2-3 times higher in magnitude compared to those observed in the early Collection 6 MODIS. After detrending, Maxar sensors were cross-calibrated to MODIS Aqua, considered as a calibration standard. In this process, DESIS hyperspectral measurements were used for spectral conversion required to align Maxar with MODIS bands. The cross-calibration analysis shows that GeoEye-1, WorldView-2, and WorldView-3 were systematically higher than MODIS Aqua by 2-4% in the Blue, Green, and NIR, and by 7-8% in the Red. De-trending and cross-calibration to MODIS Aqua effectively transforms the Maxar constellation into a common Manuscript