Publications

Matthews, G (2022). Assessment of Terra/Aqua MODIS and Deep Convective Cloud Albedo Solar Calibration Accuracies and Stabilities Using Lunar Calibrated MERBE Results. REMOTE SENSING, 14(11), 2517.

Abstract
Moon calibrated radiometrically stable and relatively accurate Earth reflected solar measurements from the Moon and Earth Radiation Budget Experiment (MERBE) are compared here to primary channels of coaligned Terra/Aqua MODIS instruments. A space-based climate observing system immune to untracked drifts due to varying instrument calibration is a key priority for climate science. Measuring these changes in radiometers such as MODIS and compensating for them is critical to such a system. The independent MERBE project using monthly lunar scans has made a proven factor of ten improvement in calibration stability and relative accuracy of measurements by all devices originally built for another project called 'CERES', also on the Terra and Aqua satellites. The MERBE comparison shown here uses spectrally invariant Deep Convective Cloud or DCC targets as a transfer, with the objective of detecting possible unknown MODIS calibration trends or errors. Most MODIS channel 1-3 collection 5 calibrations are shown to be correct and stable within stated accuracies of 3% relative to the Moon, much in line with changes made for MODIS collection 6. Stable lunar radiance standards are then separately compared to the sometimes used calibration metric of the coldest DCCs as standalone calibration targets, when also located by MODIS. The analysis overall for the first time finds such clouds can serve as an absolute solar target on the order of 1% accuracy and are stable to +/- 0.3% decade(-1) with two sigma confidences, based on the Moon from 2000-2015. Finally, time series analysis is applied to potential DCC albedo corrected Terra data. This shows it is capable of beginning the narrowing of cloud climate forcing uncertainty before 2015; some twenty five years sooner than previously calculated elsewhere, for missions yet to launch.

DOI:
10.3390/rs14112517

ISSN:
2072-4292