Multi-channel Imager Algorithm (MIA): A novel cloud-top phase classification algorithm

The current Geostationary Operational Environmental Satellites (GOES-16 and 17) cloud-top phase classification algorithm is based primarily on empirical thresholds at multiple wavelengths that have varying absorption capabilities for water ice. performance of GOES-16 product largely depends the accuracy selection reflectance ratios. This study aims presenting a novel (the Multi-channel Imager Algorithm, MIA) provides more judicious relationships between channels using supervised K-mean clustering method multi-channel Red-Green-Blue images. works analogously to how human eyes separate different colors in microphysical color rendering set satellite images, which differentiates water, ice unclassified thin clouds. For phase, temperature information used further distinguish supercooled water. To evaluate MIA, an extensive comparison with Cloud-Aerosol Lidar Orthogonal Polarization (CALIOP), Moderate Resolution Imaging Spectroradiometer, products conducted, CALIOP as benchmark. Compared product, MIA demonstrates substantial improvement classification, where hit rate increases from 69% 76% over Continental United States 58% 66% full disk domain.