Scaling a 3D particle-resolved aerosol model to address uncertainties in aerosol-atmosphere interactions

Matthew West, University of Illinois at Urbana-Champaign

Usage Details

This research aims to address key uncertainties associated with aerosol-climate impacts. Aerosol particles influence the large-scale dynamics of the atmosphere and climate because they interact with solar radiation, both directly by scattering and absorbing light and indirectly by acting as cloud condensation nuclei. Their sizes range from nanometers of micrometers, and a major source of difficulty in understanding the climate impact of aerosols is due to scale interactions. To address this difficulty, the objective is to produce the rst-ever particle-resolved simulation of atmospheric aerosols over a regional-scale domain of California, by scaling up the next-generation aerosol model WRF-PartMC-MOSAIC. The WRF-PartMC-MOSAIC model is unique in its ability to track size and composition information on a per-particle level. Such a simulation capability on the large-scale does not currently exist, but will provide a key advance in predicting the aerosol impact on the Earth's climate. A model of this capability is both compute-intensive and memory-intensive. How-ever, the unique capabilities of the Blue Waters system allow for these computational difficulties to be overcome. Blue Waters is an appropriate architecture for this problem because of the need for tens of thousands of cores, a fast interconnect for inter-process particle transport, and large memory per process. Blue Waters' compute power and large scale will enable the development of an ultra-high-resolution model at the large-scale production level. This capability of conducting large regional-scale simulations will allow for quantification of errors in climate-relevant quantities in current-generation climate models.