Inferring Ice Cloud Properties from Polarized Sub-Millimeter Microwave and Infrared Spaceborne Observations

Ice clouds account for significant uncertainties in our understanding of the current climate as well as our ability to predict future changes. The high spatiotemporal variability of ice clouds coupled with their unique microphysical and macrophysical properties, makes modelling efforts difficult. To reduce uncertainties associated with modeled ice cloud properties, it is essential to have long-term global measurements of ice cloud characteristics. Previous studies have shown a measurement gap exists in ice cloud sensitivity and could potentially be filled by sub-millimeter (sub-mm) wave measurements. We present a retrieval algorithm based on the optimal estimation framework designed to retrieve ice cloud properties, which combines sub-mm and infrared (IR) radiometric and polarimetric measurements. Sub-mm and IR observations of ice clouds are complementary and exploit ice particle scattering that effectively modulates the upwelling background radiation from water vapor. The primary retrieval quantities are cloud ice water path (IWP), ice particle effective diameter (Deff), including optional cloud top height. A novel part of this algorithm is that it utilizes polarized brightness temperatures (TBs) in both the high-frequency sub-mm and IR wavelength regimes, of which there is a lack of studies exploring the benefits. A state-of-the-art database of cloud ice optical properties is incorporated into the Atmospheric Radiative Transfer Model (ARTS) to stochastically simulate TBs from CloudSat observations over the tropics, which we use to conduct retrieval experiments. Retrieved cloud properties are compared to the true values and statistically analyzed. Information content in measured TBs is also used to evaluate retrieval performance and is expressed quantitatively in terms of degrees of freedom for signal (DOF) and Shannon Information Content (SIC). Although retrieval precision varies with cloud scene, the algorithm is demonstrated to effectively infer IWP and Deff over a wide range of cloud and atmospheric conditions and shows the best performance for clouds with moderate to low IWP and Deff