Logo image
Back
Development, application, and review of sensitivity analysis techniques in chemical transport models for secondary pollutant formation in the atmosphere
Dissertation

Development, application, and review of sensitivity analysis techniques in chemical transport models for secondary pollutant formation in the atmosphere

Jiachen Liu
Doctor of Philosophy (Ph.D.), Drexel University
May 2026
DOI:
https://doi.org/10.17918/00011376
pdf
Liu_Jiachen_202664.98 MB
PDF Embargoed Access, Embargo ends: 30 Jun 2027

Abstract

Chemical transport models Community Multiscale Air Quality Health impact analyses Hyperdual numbers Nitrogen oxides Sensitivity analyses
Chemical transport models (CTMs), such as the Community Multiscale Air Quality Model (CMAQ), are essential tools for simulating pollutant transport and formation pathways in the atmosphere. Sensitivity analysis within CTMs quantifies how pollutant concentrations respond to variations in model parameters, including precursor emissions, thereby providing critical insight into emission-air quality relationships. Traditional sensitivity methods, including the finite-difference method, often fail to accurately capture the crucial nonlinear interactions that govern pollutant formation. In this work, we introduce CMAQ-hyd, a novel augmented model that can calculate first- and second-order sensitivities to machine precision. Through comparative evaluations, we demonstrate that CMAQ-hyd robustly reproduces known sensitivities and substantially improves the characterization of nonlinear atmospheric responses. We further applied CMAQ-hyd to investigate the linear and nonlinear relationships governing atmospheric pollutants, specifically fine particulate matter (PM_[2.5]) and ground-level ozone (O₃), with respect to anthropogenic NO_x (NO_x=NO+NO₂) and Volatile Chemical Products (VCP) emissions. Our analysis highlights the critical importance of incorporating higher-order sensitivities in urban centers, revealing that neglecting these nonlinear terms leads to biased policy outcomes. Specifically, we also demonstrate the potential health disbenefits associated with large-scale NO_x emission reductions. Finally, this dissertation includes a comprehensive review of existing CTM sensitivity analysis techniques. This review provides researchers and policymakers with guidance on comparative advantages and disadvantages, computational efficiencies, and application scenarios of existing methods, ensuring the informed selection of methods to better design emission control strategies.

Metrics

1 Record Views

Details

Logo image