An Assessment of the Impacts of Cloud Vertical Heterogeneity on Global Ice Cloud Data Records From Passive Satellite Retrievals

Spaceborne passive instruments are widely used to infer long-term ice cloud properties due to their large temporal and spatial coverage. Although observations from active instruments demonstrate ice particle variability in the vertical dimension, a pragmatic assumption made in passive cloud retrieval algorithms is that the observed scene consists of a plane-parallel cloud. In this study, a theoretical exploration on how ice cloud vertical heterogeneity (ICVH) influences passive retrievals (i.e., cloud optical thickness, cloud effective radius, and ice water path, IWP) is implemented at the pixel scale. Specifically, with an established ice cloud profile database inferred from 1-year Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation/CloudSat, we quantitatively estimate ICVH-induced biases on monthly averaged cloud macrophysical and radiative properties. Results show an average underestimation (-35%) of Moderate Resolution Imaging Spectroradiometer (MODIS) monthly IWP due to the ICVH for global ice clouds over ocean. The ICVH impacts are enhanced at large IWPs (e.g., > 500 g/m²) and solar zenith angles, resulting in a profound underestimation of MODIS IWP (up to -50%) in deep convective regions and middle to high-latitude regions in the winter hemisphere. The global-averaged ice cloudy-sky reflected solar radiation and outgoing longwave radiation derived from MODIS retrievals are slightly overestimated, suggesting that the ICVH has little impact on cloud radiative properties. Relatively large reflected solar radiation (0.3 W/m²) and outgoing longwave radiation (0.1 W/m²) flux differences occur at high and low IWPs, respectively. The largest total flux difference ( 2 W/m²), mainly contributed by shortwave part, is associated with deep convection where the typical IWP is greater than 2,000 g/m².