Evaluation of Cloud Ice Microphysical Models with Habit Prediction Using Polarimetric Radar Observables

Radar observables are widely used in cloud property retrievals and model evaluations as they provide continuous observations of clouds in time. Accurate model evaluations require that the models provide sufficient information to compute realistic radar observables with which to compare to observations. Lu et al. (2013) and others show that the scattering properties of ice crystals at radar wavelengths not only depend on the mass and size of the ice crystals but are also sensitive to their aspect ratios and detailed shapes. However, traditional cloud microphysical models treat ice crystals as spheres or as canonical types with aspect ratios that do not change over time. These treatments of ice crystals may lead to large errors when estimating radar observables from them with forward scattering models. The adaptive habit prediction model presented in Harrington and Sulia (2011) predicts both the mass, maximum dimension, density and aspect ratio of ice crystals, properties which are necessary for accurate ice crystal scattering calculations in forward models.

Based on Lu et al. (2013) a method that estimates ice crystal scattering properties from model outputs of ice crystal mass, density, aspect ratio and projected areas is developed. The accuracy and computational efficiency of this scattering method is evaluated by comparing its results with Discrete Dipole Approximations (DDA) calculations for ice crystals sampled across the model output space of ice crystal properties. Using ice crystal size distributions predicted by the model, radar observables such as radar reflectivity, differential reflectivity and differential phase are estimated. This methodology will be applied to the evaluation of the adaptive habit prediction model using polarimetric measurements from the new scanning W- and Ka-band cloud radars to be deployed by the Department of Energy Atmospheric Radiation Measurement program in 2014.