Scale-dependent ambient air quality uncertainties within MUSICA, a fully-coupled chemistry-climate model

Abstract:

The Multi-Scale Infrastructure for Chemistry and Aerosols (MUSICA) is an ongoing initiative to develop a next-generation suite of models aimed at enabling easier community access, increasing model efficiency, and answering new questions at the intersection of climate and air quality. The first iteration of this model, MUSICAv0, is based on the Community Earth System Model (CESM), in particular the Community Atmosphere Model with full chemistry (CAM-chem) and regional refinement of the contiguous United States. This novel model approach allows for the estimation of air quality at exposure relevant scales within a fully-coupled global chemistry-climate model. Initial experiments with this model have isolated the impacts of changing model resolution, chemical complexity, and emissions assumptions on ambient air quality and exposure to health-relevant pollutants (PM_2.5 and O₃) over CONUS. Model output is evaluated using comparisons to surface observations from the EPA Air Quality System network and we perform further analysis of regional effects via aggregation to the ten EPA regions. The best agreement with observations occurs using regional refinement, high-resolution emissions, and increased chemical complexity, although regional refinement alone, when coupled with coarse resolution emissions (~1 degree grids), performs worse that the coarse non-refined resolution model. The experiments run here also highlight some specific events (e.g. wintertime ozone in the Mountain West, regional PM_2.5), where all model iterations consistently disagree with observations, demonstrating missing or incorrectly parameterized process within the models. This work aims to inform stakeholders and researchers about the uncertainties tied to model assumptions and identify specific areas of research which have an increased impact on estimation of the burden of disease related to ambient air quality.