Abstract:

Regional air quality is strongly influenced by the interactions among emissions, atmospheric transport, and boundary-layer processes, yet these factors remain among the largest sources of uncertainty in chemical transport models. At the same time, advances in airborne and geostationary observing systems now provide unprecedented information on atmospheric structure and composition. In this talk, I will present a research framework that uses multi-platform atmospheric observations to inform and improve the physical and chemical processes governing regional air quality. I will first show how machine-learning techniques applied to airborne lidar can improve estimates of planetary boundary layer height, creating a new observational product for evaluating boundary layer representation in air quality models. I will then demonstrate how geostationary satellite observations (GEMS) can be integrated with a chemical transport model (WRF-Chem) and inversion frameworks to improve the representation of urban NOx emissions and air quality during the 2024 ASIA-AQ campaign in Bangkok. Finally, I will extend these concepts to the 2025 Los Angeles fires, using satellite observations (TEMPO, TROPOMI) and high-resolution modeling to investigate uncertainties in fire emissions, plume rise, and satellite retrieval assumptions during an extreme air-quality event. Together, this work demonstrates how multi-platform observations can be used to evaluate and improve the representation of boundary layer height, anthropogenic emissions, and smoke transport in regional air quality models.