- Sponsor
- Department of Civil and Environmental Engineering
- Originating Calendar
- CEE Seminars and Conferences
Quantifying Impact of Aerosol Mixing State on Multiphase Chemistry Through Modeling and Observation
Advisor: Professor Nicole Riemer
Abstract
This dissertation quantifies how aerosol mixing state influences multiphase chemistry and evaluates whether it can be reliably constrained from routine observations. Atmospheric aerosols are mixtures of solid and liquid particles suspended in air that interact with solar radiation, clouds, and atmospheric chemistry. Individual aerosol particles contain multiple chemical components, and their composition evolves continuously through processes such as multiphase reactions, coagulation, and condensation. As a result, particles within the same population can differ substantially in composition. This particle-to-particle compositional variability is described by aerosol mixing state. Most atmospheric models simplify this variability by assuming uniform composition within each size bin or mode for computational efficiency, yet the implications of this simplification for multiphase chemistry remain poorly quantified.
The first part of this work quantifies how simplified aerosol representation affect multiphase chemistry using particle-resolved simulations, which explicitly track the composition of individual particles and enable direct evaluation of composition-dependent processes. The analysis focuses on two representative processes: heterogeneous dinitrogen pentoxide (N2O5) hydrolysis, which depends strongly on particle composition and affects atmospheric oxidant budgets, and secondary organic aerosol (SOA) condensational growth, which influences particle size and cloud droplet formation. The results show that assuming uniform particle composition can bias estimates of reaction efficiency even when bulk gas concentration appear accurate. They further show that neglecting seed-dependent SOA growth can lead to systematic biases in aerosol size distributions and in estimates of cloud droplet formation.
The second part evaluates whether aerosol mixing state can be inferred from routine observations. Using particle-resolved simulations as a benchmark, a hygroscopicity-based retrieval method is assessed across a wide range of continental aerosol conditions and found to perform reliably, with a mean absolute error of ~2%. The validated retrieval is then applied to long-term hygroscopicity measurements from multiple global sites to construct a multi-site, size-resolved climatology of aerosol mixing state. The results indicate that ambient aerosol populations are predominantly compositionally similar, with spatial, seasonal, and size-dependent variability driven primarily by changes in emission source and environmental conditions.
Overall, this work provides a quantitative understanding of how particle-level chemical variability influences multiphase chemistry and demonstrates that aerosol mixing state can be constrained from routine observations. By linking particle-scale compositional complexity to both mechanistic understanding and observationally based characterization, this dissertation provides a foundation for improved aerosol representation in atmospheric models.