资源环境科技发展态势分析平台

Current satellite cloud products from passive radiometers provide effective single‐layer cloud properties by assuming a homogeneous cloud in a pixel, resulting in inevitable biases when multiple‐layer clouds are present in a vertical column. We devise a novel method to retrieve cloud vertical properties for ice‐over‐water clouds using passive radiometers. Based on the absorptivity differences of liquid and ice clouds at four shortwave‐infrared channels (centered at 0.87‐, 1.61‐, 2.13‐, and 2.25‐μm), cloud optical thicknesses (COT) and effective radii of both upper‐layer ice and lower‐layer liquid water clouds are inferred simultaneously. The algorithm works most effectively for clouds with ice COT<7 and liquid water COT>5. The simulated spectral reflectances based on our retrieved ice‐over‐water clouds become more consistent with observations than those with a single‐layer assumption. This new algorithm will improve our understanding of clouds, and we suggest that these four cloud channels should be all included in future satellite sensors.