Numerical modelling of the cloud-aerosol interactions for VOCALS

Understanding the aerosol-cloud droplet relation is important not only for climate models, but also for weather prediction models. Due to insufficient resolution these models can not capture variability in clouds and as a result require parametrization of the relation between aerosol concentration and cloud droplet concentration. Observations can be used to develop such a parametrization but results show a non-linear dependence between aerosol concentration and cloud droplet concentration and significant variability for different observational campaigns. One of the approaches used to understand cloud-aerosol interactions are high resolution, Large Eddy Simulation (LES) models. However since the majority of the spectrum predicting (bin) models represent droplet spectrum on a 1D grid, information about aerosol properties is lost after activation. This leads to problems with aerosol treatment after droplet evaporation, and together with numerical issues associated with condensation and coalescence, may lead to significant uncertainty in model solution.

Recently, alternative approaches to microphysics have been developed for LES models with the microphysics represented in a Lagrangian framework. Millions of Lagrangian parcels are tracked in this scheme with the two way interaction between Lagrangian parcels and LES dynamics and thermodynamics. Since effect of Lagrangian parcel collision is mapped on 2D Eulerian bin grid (spanning both aerosol size and cloud droplet size) full information about aerosol is available during growth/activation/deactivation. Overview of this type of the model will be presented together with the results from the model. Three cases with 250, 120 and 65 cloud droplet/cm^3 will be discussed with initial conditions based on observations from VOCALS field campaign. Model results will be compared with observations and examples of the information from the model, relating aerosol spectrum and cloud droplet spectrum will be shown.