Sensitivity of Radiative Flux Simulations to Ice Cloud Parameterization over the Equatorial Western Pacific Ocean Region

Previous studies suggest explanations of the observed cancellation of shortwave (SW) and longwave (LW) cloud radiative effects (CREs) at the top of the atmosphere over tropical oceans where deep convection prevails, such as interactions among cloud microphysics, radiation, and dynamics. However, simulations based on general circulation models (GCMs) show disagreement in terms of the net (SW + LW) CREs over tropical deep convective ocean regions. One of the GCM uncertainty sources is the parameterization of ice cloud bulk optical properties. In this study, a combination of active and passive satellite daytime cloud retrievals is used to study the sensitivity of radiation flux calculations to ice cloud parameterization over the equatorial western Pacific Ocean region. Three ice cloud schemes are tested. The first is a widely used scheme that assumes hexagonal column ice particles. The second scheme treats ice particles as aggregates of surface-roughened hexagonal columns. The third scheme best matches the cloud ice mass-dimension relation in the cloud microphysics scheme by assuming a mixture of two ice particle habits. The results show that the hexagonal-column-based scheme has the weakest SW CRE but strongest LW CRE among the three. In addition, cloud optical thickness and effective radius are used to cluster cold-top single-layer ice clouds into three types, which resemble thin cirrus, detrained anvil clouds, and deep convective cores, respectively. In agreement with previous studies, cloud SW heating overwhelms LW cooling in the upper portion of anvil-like clouds.

Automated summary

This study investigates the sensitivity of radiation flux calculations to ice cloud parameterization over the equatorial western Pacific Ocean region using a combination of active and passive satellite cloud retrievals. Three ice cloud schemes are tested, showing differences in shortwave and longwave cloud radiative effects. Cloud optical properties are used to classify ice clouds into three types, with varying dominance of shortwave heating in different cloud types.