Civil and Environmental Engineering, 2000-2026

The study aims to investigate Baltimore City's air pollution through an environmental justice (EJ) lens by referencing a historical analysis of the city and racial and economic demographic data to understand if there are EJ disparities across the city. Data from the EPA's National Emissions Inventory (NEI) was used to calculate multiple Gaussian Plume Model (GPM) concentrations from point sources all across Baltimore City. After analysis of the models, it was concluded that there are no explicit air quality violations, but certain wind directions do subject neighborhoods of marginalized racial and economic backgrounds to higher amounts of emissions. Future studies involving an environmental justice lens in environmental engineering will result in a more holistic profile of the environmental harms communities are exposed to.

This thesis presents a computational phase-field study of crack propagation in brittle composite materials containing a single embedded inclusion. The central question is how inclusion stiffness and proximity to the crack path govern crack-inclusion interaction in plane-strain Mode I fracture. A parametric sweep is conducted across stiffness ratios Eratio = Einclusion/Ematrix ∈ {0.01, 0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 1.0, 5.0} and vertical offsets yoffset ∈ {4ℓ, 8ℓ, 16ℓ} for both square and circular inclusion geometries, yielding a master phase diagram that classifies each configuration as arrested, deflected, or passed-through. The results demonstrate that compliant inclusions (Eratio < 1) positioned close to the crack path attract and arrest the crack through a fracture energy landscape mechanism: reduced local stiffness lowers the strain energy density within the inclusion, creating an energetic minimum that draws the crack tip toward the compliant region. Stiff inclusions (Eratio = 5.0) produce the opposite effect, repelling the crack slightly upward without deflecting it in any sustained sense. For the square inclusion geometry, the arrest–deflection phase boundary lies between Eratio = 0.3 and Eratio = 0.4 at yoffset = 4ℓ. For the circular geometry, this boundary shifts to between Eratio = 0.2 and Eratio = 0.3, reflecting the smaller effective cross-section presented by a curved boundary to the approaching crack front. At large offsets (yoffset = 16ℓ), only extreme compliance contrasts produce crack arrest, placing a practical bound on the range of crack-inclusion interaction. A preliminary stochastic extension models the matrix Young’s modulus as a spatially correlated lognormal random field and evaluates crack outcomes across ten independent seeds near the deterministic phase boundary. Configurations well inside the arrested or passed-through regimes show consistent outcomes across seeds, confirming that the deterministic phase boundaries are robust under moderate microstructural heterogeneity, while configurations near the boundary exhibit sensitivity to the random field realization. These results provide a foundation for future probabilistic characterization of crack-inclusion interaction in heterogeneous brittle composites.

Natural hydrogen (H2) is an emerging energy resource with significant exploration potential, and sub-circular depressions (SCDs) have been identified as a possible surface proxy for subsurface hydrogen systems. This study applies multi-temporal Interferometric Synthetic Aperture Radar (InSAR) to characterize surface deformation at the Gassola SCD, located near the Bourakébougou natural hydrogen field in Mali, using Small Baseline Subset (SBAS) and Persistent Scatterer (PS) processing of a 64-scene Sentinel-1 ascending orbit time series spanning 2019 through 2023. The SBAS mean LOS velocity field reveals a spatially heterogeneous deformation signal concentrated within and around the depression, with velocities ranging from -32 to +30 mm/year after quality control masking. A transect across the depression resolves central uplift of approximately +15 mm/year flanked by subsidence of -20 to -30 mm/year, with a topographic diameter-to-depth ratio of ~1% consistent with published SCD morphology. Displacement time series show alternating subsidence and uplift phases on timescales of less than a year, with a broadly anti-phase relationship between depression center and rim. Annualized epoch-to-epoch displacement maps reveal episodic deformation activity, with periods of spatially diffuse regional displacement giving way to progressively localized, depression-centered deformation that contracts before decaying to inactivity. This recurring pattern is interpreted as the surface expression of pulsed gas migration converging on the SCD through preferential subsurface pathways. SBAS and PS solutions agree on deformation spatial patterns during most active intervals but disagree in magnitude by a factor of approximately five, attributed to fundamental differences in scatterer sampling between the two methods. No consistent correspondence between rainfall and displacement is identified, supporting a gas-dynamic rather than hydrological interpretation. The results demonstrate the potential of InSAR as a first-order screening tool for natural hydrogen exploration but integration with geochemical and subsurface datasets remains necessary for definitive interpretation.

The ancient Norwegian stave churches are testaments to the longevity of timber construction with proper design and preservation. These eight century old structures have received little structural analysis for centuries, so this thesis works to attain a better understanding of one important structural component: the Norwegian knee brace. This knee brace is found vertically bracing columns and beams, horizontally bracing beam connections, and stiffening the aisle walls and rafters, laterally bracing the churches. This thesis analyzes two geometries that the knee brace is found in, one nearly symmetrical and the other exceptionally asymmetrical, to determine the different mechanical behaviors and thus the importance of the geometry of the knee brace. The knee braces of the stave churches are constructed from the root flare, the part of the tree where the grain naturally curves as the trunk transitions to the roots. Wood is anisotropic, best resisting load parallel to the grain, making this material selection critical. To understand the importance of this defining characteristic of these braces, samples of the same geometry are loaded in compression and tension with three different grain directions: curving grain, grain that is parallel to one leg of the brace and perpendicular to the other, and grain that is at a 45° angle to both legs. The findings of this thesis indicate that the curving-grained samples perform stronger and stiffer than the straight-grained samples, underscoring the importance of material selection and geometry for the timber knee brace.

Deep in the Heart of Texas: Modeling Compound Hurricane Wind and Heatwave Hazards in Power Grid Outages develops a modeling framework to evaluate how compound hazards, specifically hurricane-induced winds and extreme heat conditions, influence the magnitude of power outages experienced by customers across Texas counties. The analysis is conducted at the county level using climate projections under the SSP245, a moderate greenhouse gas emission scenario, and the SSP585, a high greenhouse gas emission scenario. This work focuses on heatwave occurrence and wind-related impacts as primary drivers of outages and examines how these disruptions affect populations based on socio-economic characteristics, such as median income and families below the poverty level. Monte Carlo simulations are conducted to predict possible power outage outcomes and trends across increasingly severe wind field return periods, climate scenarios, and projected air temperature increases. This work finds that in longer return periods, wind damage saturates the power grid to a threshold, so there is no residual vulnerability for heat to amplify. This work also finds that population is the strongest predictor of compound outages and outcomes, while income is a moderate but significant predictor of compound outages. A group of Texas counties are identified as hotspots for hurricane-induced winds and extreme heat conditions. In the face of potential increases in outage risks and vulnerability, it is important to create resilience measures that can account for compound hazards.

Per- and polyfluoroalkyl substances (PFAS) resist conventional degradation because of the thermodynamic stability of their carbon-fluorine bond, and current treatment relies on granular activated carbon (AC) and ion exchange resins (IER) that only sequester rather than degrade the contaminant, thus generating PFAS-laden waste that requires disposal as hazardous waste or energy-intensive thermal regeneration above 800 °C. This thesis investigates an integrated paradigm in which AC and IER serve simultaneously as PFAS sorbents and as microbial carriers for Acidimicrobium sp. strain A6 (A6), the first organism shown to defluorinate perfluorooctanoic acid (PFOA) and perfluorooctane sulfonic acid (PFOS) through the anaerobic ammonium oxidation coupled to ferric iron reduction (Feammox) pathway. The hypothesis is that these solid sorbents can both concentrate PFOA at the solid-liquid interface while sustaining A6 metabolism, co-locating adsorption and biological transformation within one engineered unit. Four materials, SIR-110-MP-HP, SIR-110-HP, ASM110 iron hybrid resin, and AGC-PFx coconut shell activated carbon, were evaluated across five phases. Phase #1 quantified abiotic PFOA sorption; phase #2 ran a 310-day anaerobic incubation under four conditions to assess Feammox activity, fluoride production, and PFOA partitioning; phase #3 compares post-incubation adsorption capacity of A6-treated and fresh sorbents to test biological regeneration; phase #4 will characterize fluoride sorption in Tris buffer to generate buffer-specific correction factors; and phase #5 will extend the system to Pseudomonas strains. All sorbents adsorbed greater than 98% of PFOA abiotically. Sustained ammonium reduction in treatments with A6 showed active Feammox over 310 days, with AGC supporting the strongest sorbent-bound microbial performance up to 19.8%. Fluoride production in experimental conditions exceeded abiotic controls and represented 5-15% of the theoretical maximum from complete defluorination. Methanol extraction recovered measurable PFOA exclusively from A6-treated sorbents, indicating microbially mediated modification of PFAS–sorbent binding. As a whole, results indicated that AC and IER are viable Feammox carriers whose biological activity alters PFOA at the sorbent surface, supporting integrated adsorption–biodegradation platforms that will help to reduce reliance on thermal regeneration and advance toward partial PFAS destruction within engineered systems.

Ageing heritage structures increasingly present safety hazards, especially in regions exposed to frequent seismic activity. Beyond their serviceability, these structures embody significant cultural, architectural, and engineering knowledge, making their preservation dependent on a clear understanding of their structural integrity. This thesis investigates the stability and seismic performance of the nearly 1,500-year-old masonry dome of the Hagia Sophia­­­—a brick and mortar structure that has endured multiple earthquakes in Istanbul, Turkey since its reconstruction in 537 AD. Despite this resilience, the dome exhibits cracking patterns that raise concern about its performance under future seismic events. Focusing specifically on the central dome, rather than the entire mosque, this project evaluates its behavior under both static and dynamic conditions. A graphic statics approach is applied to analyze the thrust distribution and assess geometric equilibrium. Additionally, a Level-Set Discrete Element Method (LS-DEM) model is developed to simulate the dome’s response to lateral loading, capturing sliding and rocking mechanisms that classical continuum methods can’t represent. Altogether, we gain deeper insight into the relationships between geometric stability and dynamic failure responses. The results contribute to a more complete understanding of hemispherical masonry domes, complementing existing finite element methods and experimental studies.

Mechanical metamaterials have emerged as a powerful design paradigm in which geometry and structure, rather than material composition alone, govern mechanical response. Origami-based mechanical structures enable tunable stiffness, bistability, and coupling between distinct deformation modes, which enables systems to respond to both their current state and also its history of loading. This paper develops a theoretical and experimental framework for history-dependent, pathway-programmable mechanics in Kresling origami columns and multi-column assemblies. Using the Preisach formalism, we model each bistable Kresling unit cell as a mechanical hysteron characterized by two switching thresholds. For an n-cell column with ordered thresholds, we prove that the resulting state-transition diagrams are equivalent across all chirality arrangements and cell counts. We further demonstrate that these diagrams exhibit non-commutative behavior and possess a self-similar, fractal structure, which have a recursively embedded transition logic. This structure was utilized to develop a shortest-pathway algorithm that deterministically predicts the unique actuation sequence required to navigate between any two configurations. Our theoretical results are validated experimentally under both translational and twist loading, confirming that pathway predictability is a robust and programmable feature of this system we developed. Broadly, this work contributes to the growing understanding that physical matter can be designed to encode sequential information, perform logical operations, and respond to loading history in prescribed ways. The principles introduced here have implications across fields ranging from soft robotics and mechanical computing to deployable structures and adaptive architecture, pointing toward a future in which the mechanical response of a structure is not merely reactive but purposefully encoded.

Urban trees play an important role in heat resilience. However, urban trees are difficult to map bottom up, especially in the Global South. Doing a tree census is expensive and time consuming and there are limited data products at a fine enough scale such that individual trees or clusters of trees can be detected. Data fusion and machine learning using SkySat satellite imagery could help detect individual trees at the intra-urban scale, creating a cheap and scalable approach for tree canopy mapping. Intra-urban heat (air, surface, and mean radiant temperature) is also difficult to model or measure at fine scale. Having these models could help inform neighborhoods on heat mitigation strategies. Combining tree canopy maps and Google building polygons with ENVI-met could be used to model neighborhoods and simulate the impact of tree canopy to inform heat resilience. This thesis develops and evaluates a machine learning framework for urban land cover mapping that can detect individual trees and distinguishes between trees, grass, and non-vegetation. Performing data fusion, this thesis will define five neighborhood archetypes based on real-world neighborhoods. Using a 24-hour microclimate simulation in ENVI-met, this thesis explores the role of trees and white painted roofs and roads in urban heat resilience. This will be done as a case study in Chennai, India. The results of this thesis show that SkySat and machine learning can detect and map individual trees in urban areas with an overall accuracy over 80%. Tree canopy coverage also demonstrates to be a more effective way for heat resilience in cities lowering the air, surface and mean radiant temperature, compared to white painted surfaces, which only lowered the surface temperature.

The rapid growth of AI is increasing the electricity demand from data centers. As climate change accelerates, how this demand is met will play an important role in power sector CO2 emissions. It is essential to understand how data centers can be powered with more renewable energy to fuel the AI Revolution while addressing climate change. This thesis investigates how additional data center demand affects the Western U.S. grid and evaluates how clean energy subsidies shape power system outcomes. Using the GenX capacity expansion model, this study simulates several scenarios in the Western U.S. grid in 2030. The first goal aims to understand how the addition of a 5 GW data center load affects the grid. The second goal is to evaluate how clean energy subsidies under the Inflation Reduction Act (IRA), now no longer in effect, would have influenced system outcomes in the Western U.S. grid. Lastly, this thesis incorporates higher natural gas turbine and fuel prices to reflect expected future market conditions and to better represent more realistic conditions in the Western U.S. grid in 2030. This GenX analysis shows how the described scenarios influence the optimal generation mix, total system cost, and CO2 emissions. The findings reveal that the additional data center load significantly increased total system cost and emissions in the Western U.S. grid. While IRA incentives increased the deployment of renewable energy in the grid without a data center, particularly solar and wind, the additional data center load was largely met by natural gas. However, under higher gas turbine and natural gas fuel prices, the grid shifted towards a greater reliance on solar and battery storage, which supplied a large portion of the data center load. The results highlight the important role of battery storage in supporting data centers to be powered by renewable energy, as cost competitive storage allows intermittent sources to meet constant demand. This case study emphasizes that the cost competitiveness of renewable technology is increasingly important amid rapidly evolving energy market conditions, and that policy incentives such as clean energy subsidies can play a key role in improving economic competitiveness.

This thesis investigates cathode design and pressure-dependent cell architecture for sulfide-based all-solid-state batteries (ASSBs), with the primary objective of optimizing composite cathode composition and understanding how stack pressure and cell geometry govern interfacial contact loss and capacity reversibility. A nickel-rich layered oxide cathode, LiNi0.8Mn0.1Co0.1O2 (NMC811), was chosen for its high theoretical specific capacity along with argyrodite-type solid electrolyte Li6PS5Cl (LPSCl). Four single-crystal (SC) NMC811:LPSCl:VGCF cathode formulations were evaluated in PEEK cells under high stack pressure, alongside a comparison between SC and polycrystalline (PC) NMC811 using the optimized composite.

Composite formulations spanning 70–90% cathode active material (CAM) fraction were systematically evaluated to probe the tradeoff between ionic and electronic percolation. Among the formulations studied, 70:30:3 gave the highest discharge capacity, indicating an optimal balance between ionic and electronic percolation, while reducing LPSCl content to 10 wt% led to severe performance loss from insufficient ionic percolation. SC-NMC811 delivered approximately 75% higher initial discharge capacity than PC-NMC811, partially attributable to shorter Li+ diffusion length in smaller SC particles. However, SC-NMC showed accelerated capacity fade and incomplete recovery after rate testing, consistent with progressive cathode active material–solid electrolyte contact loss caused by anisotropic mechanical strain of the SC particles at the interface.

To assess practical low-pressure operation, SC-NMC811 composites were tested in a silicon-anode pouch cell format, representing a more realistic ASSB cell geometry. Capacity recovery improved substantially under reduced pressure, reaching near-complete recovery in larger pouch cells compared with limited recovery in rigid plunger cells. This improvement is attributed to reduced pressure-driven edge constraints, while the small remaining irreversible loss is assigned to pressure-independent interfacial degradation between NMC811 and LPSCl.

These results show that SC-NMC811 offers a higher capacity ceiling in sulfide ASSBs, but requires careful optimization of composite design and mechanical stack pressure conditions. Recovery capacity is proposed as a practical diagnostic for surviving electrochemically active electrode area. Future work should focus on decoupling mechanical and chemical pathways, as well as extending these design principles to pouch-cell architectures and other scalable cell architectures operating under low stack pressure.

The intention of this research is to divide theme park guests into clusters based upon their movement preferences and behaviors into waders, swimmers, and divers and provide the groundwork for improved theme park pedestrian modeling through data driven parametrization of digital agents. To create this informed groundwork, data collection at three theme parks across Southern California is conducted, consisting of GPS tracking data, a survey, and a log of a participant’s theme park day using the GPS tracks. These collection outputs are combined to form parameters with associations to definitions of waders, swimmers, and divers to cluster guests into groupings. Comparing these clusters to measured data and survey responses validates algorithmic cluster groupings, demonstrates that clustering varies across parks with different audiences, and provides parameters that can be used to inform the weighted preferences of digital agents in simulations, improving the theme park experience.

Mechanical metamaterials are engineered materials that process unconventional properties derived from its geometry, topology, and structural makeup rather than material composition. Recently, art forms like origami and kirigami have emerged as prominent fabrication methods for creating mechanical metamaterials with programmable tunable properties. Researchers have found ways to create metamaterials that contain the ability to switch stiffnesses and wave propagation by combining this paper folding technique with foundational topology theory. The Kagome Maxwell pattern is a lattice class that demonstrates asymmetrical mechanical responses when topology configuration changes. The folded kirigami technique is used to transform a single sheet of cut paper into a 3-dimensional metamaterial with close to ideal hinge conditions, allowing the lattice to closely match predicted theoretical behaviors. However, existing research is primarily on defect free lattices, leaving a critical gap in understanding how real-world imperfections impact material performance. This study addresses the gap by investigating how two categories of defects, vacancies and weak joint zones, affect the mechanical properties and polarization in Kagome lattices using the folded kirigami construction method. Results from this study reveal that vacancies disrupt polarity in the polar topological configuration of the Kagome lattice, meaning that lattices with vacancies no longer exhibit extremely polarizing stiffness behavior. The shape and position of the vacancy also have their respective impacts on the mechanical behavior, suggesting meaningful interaction between vacancy geometry and loading direction. Additionally, weak joint zones surprisingly did not consistently reduce stiffness across all configurations but rather changes as deformation displacement increases. These findings demonstrate that real world defects introduced partially predictable changes to topologically tunable metamaterials, implying that defects are crucial design considerations controlling deformation, stress distribution and other mechanical behaviors.

Photovoltaic green roofs (PV-GRs) have been found to improve solar panel efficiency, reduce building energy consumption, and counteract the urban heat island effect. Previous studies have examined these benefits through single-site experiments, but PV-GR performance varies across climates due to localized meteorological conditions. Furthermore, the effects of design parameters such as PV panel coverage and irrigation strategy on PV-GR performance remain underexplored. This study presents a flexible, physics-based modeling framework that simulates surface energy balance, heat transfer, and water transport throughout a PV-GR system. The PV-GR model is validated with field data from Beijing, China. Results show that PV-GR systems provide substantial roof cooling across five climates (Phoenix, Denver, New York, Miami, and Seattle) through two interacting mechanisms — evapotranspiration and panel shading — although these cooling mechanisms are not additive. This study finds that humidity governs which mechanism dominates the roof cooling: dry climates achieve more cooling from evapotranspiration, while humid climates benefit more from panel shading. Climates with high temperature variability also face heat-trapping at night with increasing panel coverage, offsetting daytime cooling gains. Analysis of irrigation scenarios finds that maintaining soil saturation at 30-40% is sufficient for evapotranspirative cooling across all studied climates, with minimal additional benefit at higher saturation levels. This is a critical finding because, ultimately, water availability emerges as the primary limiting factor for PV-GR implementation: dry climates require up to 15 times the average daily summer precipitation rate to maintain minimum irrigation thresholds, while humid climates can theoretically sustain irrigation from precipitation alone. These findings provide a framework for climate-informed PV-GR design and highlight the importance of water availability assessments prior to implementation.

Both coffee and matcha have an extensive history in Japan, but their presence in today’s markets differs greatly; while Japan’s coffee industry is valued at $35.43 billion, the matcha industry is valued at only $391 million, 100 times less than the coffee industry. This thesis answers the following questions: 1) What behavioral and cultural factors explain the dominance of coffee over matcha in contemporary Japan, and how do consumers perceive both beverages amid matcha's global surge in popularity? 2) How do environmental and supply chain constraints impact the cost structures and future of coffee and matcha in Japan? 3) What operational strategies can cafés implement to optimize profitability and reduce environmental impacts under emerging cost and sustainability constraints?

The project combines consumer survey responses and procurement data from a partner coffee shop with optimization modeling and life cycle assessments to provide both behavioral and quantitative explanations. The results show that coffee dominates daily consumption in Japan due to contemporary integration into routine habits, while the strong cultural roots of matcha have become a point of friction in its modern adoption. Nevertheless, certain demographics, especially younger age groups, are showing greater interest in the Westernized adaptation of matcha - matcha lattes. More significantly, the optimization model showed that cafés can simultaneously increase their weekly profits and decrease their carbon footprint by promoting tea offerings and shifting coffee sourcing to geographically closer origins such as Indonesia. Additionally, the LCA indicated that pour over coffee was approximately 1.6 times more carbon intensive than matcha lattes. Ultimately, these results provide cafés in Japan with actionable insights for profitable and sustainable initiatives to best adapt to climate-driven supply disruptions that will affect both markets in the coming years.

Nitrate-Nitrogen (NO3-N) pollution in bodies of water harms aquatic and human communities. NO3-N notably degrades drinking water quality through eutrophication, and causes hypoxic conditions that harm aquatic life. Population increase has led to a boom in agricultural production, leading to increased fertilizer use, and excess nutrients from fertilizers are found in critical water bodies. The Des Moines River in North-Central Iowa receives a significant amount of excess nitrogen due to tile drainage systems in the surrounding areas. Communities that rely on water from the Des Moines River have been impacted by orders to reduce water use. As climate change progresses, North-Central Iowa will likely experience an increase in extreme precipitation and temperature events. In this study, I assess the impact of climate change on flow, as well as NO3-N levels in the Des Moines River near Des Moines, based on three future scenarios (Shared Socioeconomic Pathways SSP2-4.5, SSP3-7.0, and SSP5-8.5). The results of this work point to an increase in the intensity and frequency of peak flow events, and depend on the selected future scenario and climate model. NO3-N results agree much less between climate model and scenario, but show concentrations above the threshold above maximum contaminant level for drinking water, suggesting a need for new management strategies and agricultural practices.

Many modern crack deflection and control strategies are material dependent and are often simplified to two-dimensional frameworks, while three-dimensional effects remain essential in fracture mechanics. By exploiting out-of-plane or three-dimensional effects without modifying the material itself, this work investigates the influence of designed lateral constraints on three-dimensional crack propagation using a quasi-static phase-field linear elastic fracture mechanics (LEFM) approach. Simulations conducted in FEniCSx under Mode I loading show that lateral constraints alter the local stress field and influence crack behavior, including crack pseudo-velocity, but do not significantly alter crack trajectory despite variations in patch size, location, and slab thickness. In contrast, tensile displacement constraints demonstrate greater potential for inducing crack deflection, although they may introduce unfavorable stress concentrations elsewhere in the structure. These results highlight the limitations of lateral constraints for crack steering and motivate further development of more effective out-of-plane lateral constraint-based strategies.

Despite advancements in hydraulic modeling, flood risk on indigenous land remains critically understudied, leaving Native American communities, who are already burdened by deficient water infrastructure and outdated floodplain maps, vulnerable to extreme flooding events. In response, this thesis creates a two-dimensional flood inundation model for the Moenkopi Wash Basin in northeast Arizona, where the Navajo Nation and Hopi Reservation are both located, to examine flood impacts under current and pseudo-global-warming (PGW) conditions. The model integrates HEC-RAS 6.6 two-dimensional unsteady flow simulations with GIS-based terrain, land cover, soil, and infiltration layers, driven by precipitation data derived from the National Oceanic and Atmospheric Administration’s AORC and Rasmussen et al.’s CONUS404 and CONUS404 PGW datasets. Streamgage data from the United States Geological Survey served as the baseline for calibration, validation, and characterization of flood-frequency behavior and risk. The resulting outputs from the model, in conjunction with a Monte Carlo resampling of median flow change, illustrate a heterogeneous pattern of PGW impacts. Seasonal analysis indicates that summer events experience both greater flow and variability, while non-summer events experience more consistent lower flows, suggesting that a warmer future climate has more pronounced effects during Arizona’s summer monsoon seasons. In addition, PGW peak flows correspond to significantly larger return periods compared to their historical counterparts, indicating a nonlinear amplification of flood risk. Ultimately, these findings suggest that global warming’s impact on flood hazard in the Moenkopi Wash Basin is event-specific, seasonally dependent, and nonlinear.

Ammonia has emerged as a promising hydrogen carrier in the transition to a low-carbon energy system, but its environmental impact depends on the accuracy of reported emissions. Recent studies suggest that emissions from industrial processes may be underestimated, raising concerns about the true scale of nitrogen release, as ammonia quickly reacts in the atmosphere to form harmful particulate matter and nitrous oxide. This thesis quantifies ammonia emissions from the AdvanSix facility in Hopewell, Virginia, using mobile atmospheric measurements and Gaussian plume modeling. A series of downwind transects were conducted to measure ammonia concentration and estimate emission rates. Under ambitious climate scenarios, hydrogen could account for 4–7% of global energy consumption by 2050 [27], with ammonia serving as the primary medium for its long-distance transport. Intercontinental hydrogen trade is projected to be primarily ammonia-based, up to approximately 65 Mt per year by 2060 [9].

Global ammonia production is projected to increase by nearly 40% by 2050 [16], driven by population growth, agricultural demand, and emerging energy applications. Leakages and undesired reactions across the ammonia value chain could result in 0.5–5% of nitrogen being lost to the environment as reactive nitrogen compounds, including NH3, NxO, and N2O. At the scale of a growing ammonia-based energy economy, even small fractional losses would represent a meaningful increase in atmospheric nitrogen loading, which makes emissions reporting from industrial facilities all the more critical.

The calculated emission rates generally ranged from approximately 9–10 g/s, with most transects falling within ±10% of the yearly reported values from AdvanSix. A final emission lower-bound of 10.1±2.4 g/s, with an upper-bound of 12.6±3.9 g/s was found. Additional analysis showed that emission estimates were relatively insensitive to source height assumptions but more sensitive to source representation and sampling completeness.

Overall, the results suggest that the facility is likely slightly under-reporting its ammonia emissions. This thesis demonstrates the effectiveness of mobile plume measurements as an independent method for validating industrial emissions reporting and highlights their importance for assessing the environmental implications of a growing ammonia-based energy system.