Investigation of spatial and temporal variations in aerosol mixing state using a particle-resolved regional aerosol model

Poster PDF

Description

The aerosol mixing state is critical for determining the microphysical interactions of particles with the large-scale atmospheric system, yet accurately capturing mixing state continues to be a major challenge for regional and global models. One of the reasons contributing to this challenge are computational constraints of numerical models, which resort to simplified aerosol representations (modal or sectional) to reduce computational cost. These types of models greatly simplify the aerosol mixing state, and the assumptions made can result in errors in computed cloud and optical properties. To quantify the importance of representing aerosol mixing state, the particle-resolved model PartMC-MOSAIC was coupled with the WRF model. The resulting model not only explicitly resolves and tracks the size and composition of individual particles as they undergo transformations by coagulation and condensation in the atmosphere, but also resolves the 3D spatial distribution of aerosols and trace gases of the atmosphere, based on meteorological fields predicted by the WRF model. The novel computational methods developed for this purpose include a particle-resolved emission inventory and stochastic transport algorithms. Particle-resolved emissions are constructed using source apportionment in SMOKE. This source-oriented aerosol emission approach allows for different emission sectors to have varying aerosol composition rather than assuming all emissions of a given grid cell are internally mixed at the point of emission. Stochastic transport algorithms sample a small fraction of particles in each grid cell to be transported at each time step, allowing efficient advection turbulent diffusion while maintaining numerical accuracy. With its fully-resolved mixing state representation, WRF-PartMC-MOSAIC allows for direct intermodel comparisons with aerosol schemes used in regional models (e.g., the sectional 4 or 8 bin MOSAIC), and climate models (e.g., MAM3/MAM4/MAM7). We used WRF-PartMC-MOSAIC to simulate, for the first time, the spatial and temporal variations in aerosol composition over northern California during the CARES campaign in June 2010. Using the unique 3D particle-resolved aerosol data from the simulation, we are able to directly compute 3D distributions of mixing state metrics and the impact of mixing state on CCN and optical properties of the aerosol.