Integrating ASCENT, Satellite and CMAQ Data over Atlanta, GA: More Questions than Answers

MERVE EKE, Ruizhe Liu, Weixin Zhang, Katherine Paredero, Yifeng Wang, Yongtao Hu, Sohyeon Jeon, Ann M. Dillner, Roya Bahreini, Nga Lee Ng, Armistead G. Russell, Georgia Institute of Technology

     Abstract Number: 111
     Working Group: Remote and Regional Atmospheric Aerosol

Abstract
2023 saw the initiation of the Atmospheric Science and Chemistry mEasurement NeTwork (ASCENT) and the beginning of data from the Tropospheric Emissions: Monitoring Pollution (TEMPO) satellite-based instrument that is gathering gas-phase pollutant data in conjunction with the Geostationary Operational Environmental Satellites (GOES) instruments capturing aerosol optical depth (AOD). Both TEMPO and ASCENT data are more complete for 2024. These data are integrated with Community Multiscale Air Quality (CMAQ) model simulations to investigate the degree to which the model can reproduce these new, more detailed data. Historically, evaluating model performance has been limited as routine measurements have not had hourly PM_2.5 composition. ASCENT provides long-term high-resolution PM data with chemical speciation and particle size distributions. There are 12 ASCENT sites and one of them is in South DeKalb which is downwind of the Atlanta metropolitan area and includes routine measurements. Using ASCENT, model performance is evaluated with not only PM_2.5 mass concentration, but also concentrations of sulfate, nitrate, ammonium, organic, elemental metals and black carbon (BC). In this study CMAQ and ASCENT-South DeKalb observations are used to examine how PM2.5 concentration and speciation differ across 2024. The hourly results showed that CMAQ misses the afternoon peak in sulfate, and that organic matter (OM) and BC are the components that contribute most to the gap between CMAQ and ASCENT PM_2.5 concentration. Comparison of TEMPO vertical column densities found that CMAQ captured the spatial trends in NO₂ and HCHO, but was biased low. Collectively, these results suggest that emissions inventories for a variety of species may be biased low and that the model is missing processes leading to the higher afternoon observations and more rapid reductions in the evening.