Atmospheric aerosols working as cloud condensation nuclei (CCN) and ice nuclei (IN) affect cloud shape, lifetime, microphysical and radiative features, further influence precipitation timing, types and amount, and eventually impact the climate system through the interactions between clouds and precipitation. Many recent studies focus on the problem of how changes in cloud microphysics, the development of precipitation, and dynamics are attributed to perturbations to aerosols. However, to what extent do changes in clouds, precipitation and dynamics regulate the distribution of aerosols has not yet been well studied. Observations showed that a fair amount of aerosols in the boundary layer can be transported into the mid-troposphere through the evaporation of hydrometeors associated with strong convective clouds. Aerosols close to the tropopause resulting from sublimated and evaporated cirrus crystals were observed as well. These recycled aerosols, from both cloud droplets and crystals, in high altitudes were thought to be long-lived and critical in the formation of secondary clouds. This aerosol recycling mechanism should be incorporated into numerical models to better represent the aerosol-cloud-precipitation-climate interactions.

In this study, a mixed phase bin (spectral) microphysics scheme (Rasmussen et al., 2002) incorporating the aerosol recycling mechanism has been coupled into the weather research and forecast (WRFV3) model. This bin scheme uses the multi-moment conservation method (Tzivion et al. 1987) to insure the conservation of mass concentration (mixing ratio) and number concentration for the following species: cloud water, rain water, cloud ice, snow, and graupel. A simple aerosol recycling treatment based on droplet-aerosol one-to-one relation is implemented into the scheme. Two ideal 2D hill cases and two ideal 2D squall line cases are run with and without the aerosol recycling mechanism. We examine how the mixed phase clouds regulate the aerosol redistribution and how the aerosol recycling affects the secondary clouds. At the same time, the simulated thermodynamic, dynamic fields, and microphysical properties, such as mixing ratios, number concentrations, and size spectral of hydrometeors from different species, are compared with the available observed data. We also plan to present two ideal 2D or 3D LES simulations to investigate the problem with shallow convective warm rain clouds.