Quality Consistency Assessment among MODIS, MERSI, and VIIRS Cloud Mask Products using China’s Space-borne Lidar ACDL

Abstract. Cloud masking is the fundamental step in accurately retrieving cloud and many other properties from satellite-based imagers. Intercomparisons among imagers are frequently conducted to assess the agreement of cloud masks at both global and granule scales, ensuring accuracy while also enabling the merging of multi-satellite products for use in NWP models and climate datasets. However, common challenges arise in scenarios with reduced spectral contrast between clouds and backgrounds, leading to generally degraded performance at polar regions and making validation through intercomparison less reliable. Building on previous synergistic observations between CALIOP and MODIS in the A-Train constellation, this study employed China's Aerosol and Carbon Detection Lidar (ACDL) as a new benchmark to evaluate cloud masks from 3 passive sensors across different operational stages. Analyzing approximately 3 million collocated samples in February and August 2024, the overall hit rates for confident cloudy/clear pixels from Aqua/MODIS, FY-3D/MERSI-II, and NOAA-21/VIIRS reached 82.9 %, 79.3 %, and 86.4 %, respectively. Detailed examinations of monthly, day/night, and latitudinal subsets revealed variation in each sensor's performance across scenarios and identified the reasons behind the discrepancies. The study highlighted the importance of active-passive validation, which confirms consistency among passive imagers while also identifying algorithmic weaknesses that need improvement for future missions.