10th International Aerosol Conference
September 2 - September 7, 2018
America's Center Convention Complex
St. Louis, Missouri, USA

Abstract View


Sparse-particle Methods for Simulation of Atmospheric Aerosol

LAURA FIERCE, Robert McGraw, Brookhaven National Laboratory

     Abstract Number: 1431
     Working Group: Aerosol Modeling

Abstract
Aerosol effects on clouds and radiation have remained a large source of uncertainty in quantifying anthropogenic changes to the Earth’s energy balance, despite increases in model complexity. Climate-relevant aerosol properties depend on characteristics of individual particles, but particle-level characteristics are not easily represented in global models. Instead, global and regional aerosol schemes approximate the representation of particle size and composition, leading to errors in prediction of direct and indirect aerosol effects that have not been well quantified. At the other extreme, the particle-resolved PartMC-MOSAIC simulates the evolution of the full particle mixing state using a Monte Carlo approach but is computationally too expensive for large-scale simulation. Here a new framework is introduced for simulation complex aerosol distributions, based on the quadrature method of moments. Key features of aerosol populations are captured using quadrature approximations of multivariate moments. Each quadrature consists of a collection of abscissas, or representative particles, and associated weights. Distributions with respect to key aerosol quantities, such as the critical supersaturation for CCN activation or optical cross sections, can be approximately reconstructed using constrained maximum entropy distributions. Unlike other reduced aerosol schemes, such as modal and sectional models, the sparse-particle approach is not constrained to pre-determined size bins or assumed distribution shapes. CCN activity and optical properties from the sparse aerosol simulations are benchmarked against the particle-resolved model PartMC-MOSAIC. This study is a first step toward a new aerosol simulation scheme for large-scale simulations that will track multivariate aerosol distributions with sufficient computational efficiency for large-scale simulations.