Toward Fine Horizontal Resolution Global Simulations of Aerosol Sectional Microphysics: Advances Enabled by GCHP-TOMAS

Abstract

Global modeling of aerosol-particle number and size is important for understanding aerosol effects on Earth's climate and air quality. Fine-resolution global models are desirable for representing nonlinear aerosol-microphysical processes, their nonlinear interactions with dynamics and chemistry, and spatial heterogeneity. However, aerosol-microphysical simulations are computationally demanding, which can limit the achievable global horizontal resolution. Here, we present the first coupling of the TwO-Moment Aerosol Sectional (TOMAS) microphysics scheme with the High-Performance configuration of the GEOS-Chem model of atmospheric composition (GCHP), a coupling termed GCHP-TOMAS. GCHP's architecture allows massively parallel GCHP-TOMAS simulations including on the cloud, using hundreds of computing cores, faster runtimes, more memory, and finer global horizontal resolution (e.g., 25 km × 25 km, 7.8 × 10⁵ model columns) versus the previous single-node capability of GEOS-Chem-TOMAS (tens of cores, 200 km × 250 km, 1.3 × 10⁴ model columns). GCHP-TOMAS runtimes have near-ideal scalability with computing-core number. Simulated global-mean number concentrations increase (dominated by free-tropospheric over-ocean sub-10-nm-diameter particles) toward finer GCHP-TOMAS horizontal resolution. Increasing the horizontal resolution from 200 km × 200–50 km × 50 km increases the global monthly mean free-tropospheric total particle number by 18.5%, and over-ocean sub-10-nm-diameter particles by 39.8% at 4-km altitude. With a cascade of contributing factors, free-tropospheric particle-precursor concentrations increase (32.6% at 4-km altitude) with resolution, promoting new-particle formation and growth that outweigh coagulation changes. These nonlinear effects have the potential to revise current understanding of processes controlling global aerosol number and aerosol impacts on Earth's climate and air quality.