A Multilayer Cloud Detection Algorithm For The New Generation Of Satellite Imager Radiometer

Multilayer(ML)clouds have substantial impacts on cloud radiative effects,and also affect cloud property retrievals.Because of larger spatial and temporal scales features,satellites radiometer remote sensing is one of the most important ways to detect clouds.The latest technological developments of meteorological satellites make greatly improvements to understand and determine cloud properties.A new multilayer cloud detection algorithm based on shortwave infrared(SWIR)and longwave infrared(LWIR)channels is developed and applied to the Visible Infrared Imager Radiometer Suite(VIIRS)observations.The algorithm identifies ML clouds,specially an ice cloud overlying a water cloud,from satellite multispectral images using SWIR reflectances at the 1.38,1.6,and 2.25 ?m channels and LWIR brightness temperatures(BTs)at the 8.5 and 11?m channels.Rigorous radiative transfer simulations are performed to provide theoretical understandings of the sensitivities of aforementioned channels on ML clouds.We find that the signal of a water cloud beneath an ice cloud can still be detected by the combination of the 1.6and 2.25 ? m cloud absorbing channels due to the different absorptivity of water and ice clouds.Moderate Resolution Imaging Spectroradiometer(MODIS)and Cloud-Aerosol Lidar with Orthogonal Polarization(CALIOP)observations are used to evaluate the algorithm.By comparing with collocated active CALIOP observations,the algorithm can determine ML and single-layer clouds correctly with approximately 60% and 80% success rate,respectively,and has similar performance to the current MODIS operational ML cloud detection algorithm.We also use the algorithm for analyzing data from radiometers onboard on the geostationary Himawari-8 and FengYun-4A satellites,and obtain similar results to those of VIIRS.