UWSpace

UWSpace is the University of Waterloo’s institutional repository for the free, secure, and long-term home of research produced by faculty, students, and staff.

Depositing Theses/Dissertations or Research to UWSpace

Are you a Graduate Student depositing your thesis to UWSpace? See our Thesis Deposit Help and UWSpace Thesis FAQ pages to learn more.

Are you a Faculty or Staff member depositing research to UWSpace? See our Waterloo Research Deposit Help and Self-Archiving pages to learn more.

Photo by Waterloo staff

Communities in UWSpace

Select a community to browse its collections.

Recent Submissions

  • Thumbnail Image
    Item type: Item ,
    Spatial and Temporal Variability of Herpetofauna Road Mortality and Carcass Persistence within a World Biosphere Region
    (University of Waterloo, 2026-09-23) Rose, Joshua
    Roads are one of the biggest threats to herpetofauna globally, especially in Ontario as the road network continues to expand beyond urban boundaries, and the human population is projected to add 4 million new residents by 2050. The Long Point Biosphere Region (LPBR) is home to over 50% of Ontario’s at-risk species but contains little remaining in-tact habitat, as past-century land use conversion has resulted in agriculture accounting for over 75% of the land by area. To better understand the spatial and temporal herpetofauna road mortality dynamics within an agriculturally-dominated UNESCO world biosphere, we completed 1,961 systematic walking surveys across 15 road transects between April and October in 2024 and 2025, and again in March 2026. Our research objective was to identify the drivers of temporal trends in road abundance, and assess the influence of environmental factors on road presence, absence, and abundance (Chapter 2). We expected each class to be most frequently on the road at a specific time of the year that corresponds to their annual life history pattern, and nearby landscape features to be associated with habitat use requirements. We encountered 1,606 individuals across the LPBR and found that turtles and snakes were predominantly found at wetland transects in May and snakes again at all transects in September. Anurans were observed on the road at non-agriculture transects in March and September, and salamanders at forest transects in September. Turtle and snake abundance were largely influenced by proportion of aquatic habitat (i.e., wetland and open water) surrounding the road, whereas anurans were most abundant near streams. Salamanders were mainly found at transects with high forest proportion, and were dominated by red-spotted newt migrations in both 2024 and 2025. Our findings highlight class-specific patterns that can be used to target mitigation efforts within the LPBR including exclusion fencing, wingwalls, culverts, and road closures. However, road mortality studies are known to severely underestimate the true number of collisions due to removal from the road by mechanical means or scavenger behaviour, with many studies reporting over 50% carcass removal in less than 24 hours. To quantify this magnitude for our study, we monitored 111 herpetofauna carcasses between May and September in 2025 to understand key drivers of carcass road persistence, predict the number of individuals that were missed with our survey effort, and identify how varying levels of survey effort would influence detectability (Chapter 3). We found that there was a significant difference in persistence time among species classes, with two thirds of reptiles removed from the road within 48 hours. Traffic volume was the most important environmental influence on persistence time across all classes. We recommend daily road surveys for reptiles and amphibians to reduce the underestimation of carcasses, as well as implementing more frequent surveys or continuous monitoring methods at known hotspots where possible. Future road mortality studies should apply correction factors to better represent the true mortality within their target population.
  • Thumbnail Image
    Item type: Item ,
    Toroidal Gibbs sampling using quantum and classical techniques
    (University of Waterloo, 2026-09-23) Hoque, Sehmimul
    Sampling from a Gibbs distribution for some energy function at some inverse temperature is a fundamental task in statistical physics, chemistry, and machine learning. However, it is often slow in practice for multimodal or non-convex energy landscapes. Gibbs sampling has been widely explored for discrete variable problems in statistical mechanics and machine learning, and also for continuous variable problems, where it is typically implemented by simulating Langevin dynamics, a stochastic differential equation (SDE) whose stationary distribution is the target Gibbs density. Langevin dynamics can be equivalently represented by a partial differential equation called the Fokker-Planck equation. This thesis presents two approaches to accelerate Gibbs sampling on a continuous domain on a hypertorus, using a classical technique inspired by Langevin dynamics and a quantum algorithm which solves the Fokker-Planck equation. First, we introduce diffusion-warmed MCMC, a method that uses a diffusion model trained on small lattices of the two-dimensional XY model to generate warm-start samples for larger lattices. We show that combining the diffusion generated samples with a small number of Wolff MCMC steps reduces thermalization time by approximately an order of magnitude compared to standard MCMC, across a range of system sizes and temperatures spanning the phase transition. Secondly, we develop a quantum algorithm to do Gibbs sampling (by solving the Fokker-Planck equation) for some class of Gibbs distributions on a hypertorus based on Quantum Singular Value Transformation (QSVT), extending previous work that required a warm start. We remove the warm start by constructing a temperature annealing schedule such that the full algorithm has constant success probability. We then establish a quantum-classical separation result for Gibbs sampling on the hypertorus under suitable regularity conditions, by showing that the upper bound query complexity of the quantum algorithm is lower than that of any possible classical algorithm, using a lower bound in query complexity proved by our collaborators. This establishes a quantum-classical separation result for continuous Gibbs sampling, which is, to the best of our knowledge, the first such result for this setting and has broad implications in quantum algorithms.
  • Thumbnail Image
    Item type: Item ,
    Architecture for Harsh Futures: A Spatial Response to Environmental Stewardship in Iceland
    (University of Waterloo, 2026-09-23) Conidi, Arianna
    Iceland’s volcanic eruptions, glacial retreat, and geothermal activity are already occurring at a pace perceptible within a human lifetime. In extreme environments like these, modern design logic often reduces buildings to environmental shields, structures that isolate occupants from volatile conditions in the name of safety. While vital, this defensive approach risks stripping architecture of the cultural and emotional connections that make human occupation in the landscape meaningful. This thesis challenges that standard, arguing that architecture’s more operative role is not to protect occupants from extreme environments but to make those environments legible, choreographing human encounters with geological processes that exceed ordinary perception and spatializing environmental observation as a visible, civic, and embodied act. Informed by precedent research and fieldwork at each site, this thesis proposes three field anchors, each sited at an active geological location in Iceland: a volcanic site in Grindavík, a glacial site at Svínafellsjökull, and a geothermal site at Þeistareykir. Each site already hosts scientific monitoring that remains opaque to the communities with the most immediate stake in its findings. Operating as civic structures, the buildings extend acts of environmental observation into the surrounding landscape, distinguishing a publicly accessible act of witness from a more restricted act of stewardship carried out by trained volunteers. Existing citizen science practices tend to treat participation as data collection abstracted from the environments they describe. This thesis instead spatializes that participation, aligning the site of observation with the site of interpretation as a civic and physical act rather than a digital one. Each building traces the effects of time through its own spatial logic, accumulating evidence of geological process in its materials, apertures, and position over time, acting as a register of planetary change rather than a barrier against it.
  • Thumbnail Image
    Item type: Item ,
    Characterization of Cell Migration and Descemet's Membrane Mechanical Properties in the Context of Fuchs Endothelial Corneal Dystrophy and Cell-based Treatments
    (University of Waterloo, 2026-09-23) Graham, Olivia
    Fuchs endothelial corneal dystrophy (FECD) is a degenerative disease that causes dysfunction and abnormal cell death of corneal endothelial cells (CECs). It is hallmarked by changes in the extracellular matrix (ECM) including thickening and softening of the Descemet’s membrane (DM) and the development collagenous guttae. Once the CEC count falls below a critical level, hydration levels in the cornea will increase, leading to corneal swelling, damage to other corneal layers, and eventual corneal blindness. Currently, FECD patients rely on corneal transplants to repair vision. There has been development in cell-based treatments to avoid the need for the transplantation of donor tissue including cell injection therapy (CIT) and Descemet’s stripping only (DSO). During these procedures, patients remain in a prone posture to aid in recovery. Corneal curvature mimicking substrates, placed in prone, supine, and upright positions, were used to study the effects of patient positioning on CEC wound-healing. Scratch wounds were made in monolayers of healthy and FECD affected cell-lines. Differences in wound closure between orientations were primarily observed during the early-to-middle stages of healing. Across experiments, wound closure in the upright orientation generally lagged behind the prone and supine orientations. These findings provide some evidence that gravitational orientation may influence CEC migration, particularly in the upright orientation, although the relationship was not definitive. Additionally, the mechanical properties of the DM across three regions, the central endothelium (CE), peripheral endothelium (PE), and transition zone (TZ) were measured. Previous literature has reported modulus values for human DMs with wide variation. Using three-point bend testing and measurement of the exact DM thickness, the flexural modulus for the CE, PE, and TZ were determined. The measured DM flexural moduli were on the megapascal (MPa) scale and were within the range of previously reported values from tensile inflation-based measurements but exceeded values reported using atomic force microscopy. No statistically significant differences in mean modulus were observed between the CE, PE, and TZ or between donor sex. However, in review of stiffnesses across the DM from individual donors, most had decreased PE stiffness compared to the other regions. Further studies are required to see if CECs subscribe to positive durotaxis or if substrate stiffness gradients affect migration.
  • Thumbnail Image
    Item type: Item ,
    Evaluating Machine Learning Models for Predicting PFAS Adsorption onto Activated Carbon
    (University of Waterloo, 2026-09-23) Bui, Gia Thinh
    Per- and polyfluoroalkyl substances (PFAS) are a diverse class of persistent contaminants of increasing regulatory concern. Activated carbon (AC) adsorption is among the most widely implemented technologies for removing PFAS from water, with performance governed by the combined effects of PFAS physicochemical properties, AC characteristics, and water chemistry. Although isotherm experiments are commonly used to characterize PFAS adsorption on AC, they can be time-consuming and resource-intensive. Models capable of predicting PFAS adsorption could reduce the need for isotherm testing by guiding experimental design or replacing selected experiments. This thesis evaluated the applicability of machine learning (ML) regression models for predicting PFAS adsorption onto AC in two types of systems: (1) synthetic solutions containing a single PFAS and (2) real groundwater solutions containing multiple PFAS. For the single-PFAS system, isotherm data from 21 published studies were consolidated into a curated dataset comprising 658 observations across 11 PFAS and 36 AC materials. To prevent data leakage and rigorously assess model performance, an isotherm-aware data splitting approach was implemented alongside three complementary validation regimes, including 5-fold cross-validation (CV), leave-one-PFAS-out (LOPO) CV, and leave-one-adsorbent-out (LOAO) CV. Four features were selected to predict the solid-liquid distribution coefficient (log Kd): PFAS molecular weight, AC specific surface area, the difference between AC pHpzc and solution pH, and aqueous-phase PFAS concentration. Among the models evaluated, multiple linear regression demonstrated consistent predictive performance across all three validation regimes (5-fold CV: R² = 0.77, LOPO CV: R² = 0.76, LOAO R² = 0.75), comparable to more complex ML models. For the compiled dataset and selected feature space, these results indicate that the linear model offers a robust and interpretable alternative to more complex ML algorithms for predicting PFAS adsorption in relatively simple systems. The modelling approach was subsequently extended to a more complex problem of predicting multi-PFAS adsorption on colloidal activated carbon (CAC) in real groundwater. The dataset comprised 192 observations for PFOS, PFOA, PFHxS, PFHxA, and 6:2 FTS obtained from isotherm experiments conducted with seven groundwater samples collected from PFAS-impacted sites. In each experiment, the five PFAS were added to the groundwater at equimolar initial concentrations. Because the dataset was relatively small, the allocation of groundwater samples between training and test datasets was systematically varied to evaluate the sensitivity of model performance to data partitioning and differences in groundwater chemistry. Across the data splitting scenarios, extreme gradient boosting and random forest provided the most accurate and consistent predictions. However, both tree-based models showed systematic overprediction or underprediction in some data-split cases. Reserving adsorption data collected at a single CAC concentration for calibration corrected this systematic bias and substantially improved predictive performance. This calibration strategy provides a practical workflow in which a single-point experiment can be used to verify or adjust model predictions under new water chemistry conditions, thereby reducing the need for complete isotherm experiments. Overall, this thesis demonstrates the potential of data-driven models to support the characterization of PFAS adsorption in both single-PFAS synthetic solutions and multi-PFAS real groundwater solutions. The models can inform the design of bench-scale isotherm studies, prioritize experimental testing, and reduce the experimental effort required to evaluate AC-based PFAS treatment. The contrasting model choices across the two studies further demonstrate that the predictive advantage of complex ML models over linear regression is situational rather than universal and should be evaluated rather than assumed. Beyond providing practical predictive tools, this work illustrates rigorous approaches to data splitting and model evaluation for relatively small experimental datasets, providing a framework for strengthening the reliability of data-driven modelling of contaminant adsorption.