Effects of ice crystal surface roughness and air bubble inclusions on cirrus cloud radiative properties from remote sensing perspective

•Surface roughness and air bubble inclusions affect optical properties of ice crystals significantly.•Including both factors improves simulations of ice cloud.• Cirrus cloud particle habit model of the MODIS collection 6 achieves better self-consistency and consistency with polarimetric observations when both factors are considered.•Cirrus cloud particle habit model of the MODIS collection 6 achieves better self[HYPHEN]consistency and consistency with polarimetric observations when both factors are considered.

We study the combined effects of surface roughness and inhomogeneity on the optical scattering properties of ice crystals and explore the consequent implications to remote sensing of cirrus cloud properties. Specifically, surface roughness and inhomogeneity are added to the Moderate Resolution Imaging Spectroradiometer (MODIS) collection 6 (MC6) cirrus cloud particle habit model. Light scattering properties of the new habit model are simulated using a modified version of the Improved Geometric Optics Method (IGOM). Both inhomogeneity and surface roughness affect the single scattering properties significantly. In visible bands, inhomogeneity and surface roughness both tend to smooth the phase function and eliminate halos and the backscattering peak. The asymmetry parameter varies with the degree of surface roughness following a U shape - decreases and then increases - with a minimum at around 0.15, whereas it decreases monotonically with the air bubble volume fraction. Air bubble inclusions significantly increase phase matrix element -P12 for scattering angles between 20°–120°, whereas surface roughness has a much weaker effect, increasing -P12 slightly from 60°–120°. Radiative transfer simulations and cirrus cloud property retrievals are conducted by including both the factors. In terms of surface roughness and air bubble volume fraction, retrievals of cirrus cloud optical thickness or the asymmetry parameter using solar bands show similar patterns of variation. Polarimetric simulations using the MC6 cirrus cloud particle habit model are shown to be more consistent with observations when both surface roughness and inhomogeneity are simultaneously considered.