Characterization of liquid cloud profiles using global collocated active radar and passive polarimetric cloud measurements

Abstract. Stratiform liquid cloud profiles are key to deciphering cloud life cycles, microphysical processes, and climate change impacts. Nevertheless, remote sensing of cloud vertical structure remains largely unresolved. CloudSat active measurements provide cloud microphysical profile products but are restricted to narrow orbital tracks. Multiangle passive imagers, such as Polarization and Directionality of Earth’s Reflectance (POLDER), are capable of generating a variety of cloud properties with broad area coverage; however, they lack key prior knowledge and effective methods for obtaining cloud vertical information. Focusing on single-layer cloud profile retrieval, we first reveal the structural characteristics of stratiform cloud effective radius (CER) profiles based on global CloudSat data and find that the dominant structures include triangle-shaped and monotonically decreasing profiles, which account for approximately 88.5 % of global liquid CER profiles. Furthermore, we propose a novel approach to estimate the structural characteristics of triangle-shaped profiles from POLDER observations like the properties of the profile turning point (TP). This approach integrates vertical structure morphology recognition with a combination of fitting methods and machine learning models. The cloud profiles are then accurately reconstructed using physical parameterization models. Our retrieval results exhibit good consistency with active observations, with an RMSE of 1.1 μm for TP_CER and 0.1 for the normalized cloud optical thickness at the TP. This research advances the parameterization of liquid cloud profiles and enables profile structural characteristic retrieval based on a multiangle passive imager. Our findings provide valuable insights into improving the understanding and modeling of cloud processes in weather and climate systems.