Quantification of the Consistency of the Choice of Ice Cloud Models in Forward Retrieval and Radiative Forcing Assessment

Abstract

For retrievals of ice cloud microphysical and radiative properties (specifically, effective particle size and optical thickness), different science teams assume different ice particle habit models in the look‐up tables that involve forward light scattering and radiative transfer model. For example, the CERES Edition 4 ice cloud property retrieval assumes a single habit model (specifically, hexagonal ice crystals with surface roughness), the MODIS Collection 6 ice cloud property retrieval assumes roughened aggregates, and the AIRS science team adopted the MODIS Collection 5 ice model (a mixture of various ice particle habits) in the infrared regime. Recently, it is found (Leob et al. 2017) that the effect of an ice particle habit on cloud radiative forcing assessment is minimized if the same ice habit model is used for the retrieval and the radiative transfer model. This finding promotes a hypothesis (hereafter, the consistency hypothesis) that, to minimize errors, a consistent ice cloud model must be used for satellite based cloud property retrievals and the assessment of ice cloud radiative forcing. The proposed research is to comprehensively validate the consistency hypothesis using MODIS Collection 6, CERES‐MODIS Edition 4, AIRS, and NPP/VIIRS cloud retrieval products. To allow robust simulation, we will use the level‐2 retrieval products. Only the pixels collocated with CALIPSO data will be used. The CALIPSO measurements will be used to constrain cloud height.

Furthermore, the MERRA reanalysis data will be used to provide atmospheric vertical profiles. The aforesaid cloud property retrieval products will be used as input for radiative transfer simulations in comparison with the CERES flux products. Furthermore, we will develop ice cloud parameterizations consistent with various satellite retrieval products for two radiative transfer models used in climate models. The purpose of developing various ice cloud parameterizations is to allow the users of cloud property retrieval products to have an appropriate parameterization in the assessment of ice cloud radiative forcing. Potential impact: the outcomes of this project will directly benefit 1) improving the accuracy of the assessment of ice cloud radiative forcing; and 2) choosing an appropriate parameterization scheme that is consistent with a specific cloud property retrieval product (e.g., MODIS Collection 6 or AIRS cloud property product) to validate climate models.