QSpace: Queen's Scholarship & Digital Collections
QSpace is an open access repository for scholarship and research produced at Queen's University. QSpace offers faculty, students, staff, and researchers a free and secure home to preserve and present their scholarship.
Recent Submissions
Item Wind-Driven Sea Surface Wave and Coastal Water Level Dynamics Across Different Spatial Scales(2024-11-29)Wind-generated surface waves can propagate over long distances towards vulnerable coastal regions. Complex factors that impact the properties of surface waves, including wind, topography, and water levels, must be incorporated into numerical models to achieve accurate simulations of the wave climate. Predictions of the offshore surface wave conditions can inform projections of nearshore impacts in response to storms or sea level rise, such as flooding, erosion, and elevated groundwater, and aid the design limits of protective infrastructure, such as dykes. In this thesis, numerical models, physical models, and field observations are used to investigate surface waves and water levels at three scales: a coastal sea, a tidal bay, and a sandy beach. A numerical model is developed for the Strait of Georgia to investigate the influence of wind resolution on locally generated wave and storm surge. A finer scale numerical model is applied to Boundary Bay, a tidal flat within the Strait of Georgia, to simulate the combined surface wave and water level overtopping potential and intertidal reduction under future sea level rise scenarios. A series of high resolution, small-scale laboratory experiments are run to investigate the relationship between wave runup-driven infiltration into the subsurface and various beach properties. The findings of this study emphasize the importance of wind resolution in models with large domains and the error induced by model input uncertainty. Increased mean sea elevation is found to increase nearshore wave heights and heighten risk of infrastructure overtopping and changes to the intertidal zone. Detailed observations of surface wave runup on unsaturated beach face demonstrate how fluid infiltrates the subsurface and raises the phreatic surface. The overall findings of this thesis contribute to understanding and improving prediction of coastal hazards and nearshore impacts in response to storm events and sea level rise at different spatial scales.Item Federated Learning With Generalization To New Domains(2024-11-29)Federated Learning (FL) is an area of research that focuses on training machine learning models in a decentralized fashion without having the need to store all data on one central server. In this thesis, we address the challenges of data heterogeneity and label scarcity in FL by proposing two novel approaches for federated domain generalization in both unsupervised and supervised settings. First, to tackle federated domain generalization in an unsupervised setting, we introduce Federated Unsupervised Domain Generalization using Global and Local Alignment of Gradients. We establish a connection between domain shifts and gradient alignment in unsupervised federated learning, demonstrating that aligning gradients at both the client and server levels facilitates the generalization of the model to new, unseen domains. FedGaLA performs gradient alignment locally to encourage clients to learn domain-invariant features, and globally at the server to obtain a more generalized aggregated model. Extensive experiments on four multi-domain datasets—PACS, OfficeHome, DomainNet, and TerraInc—show that FedGaLA outperforms comparable baselines. Ablation and sensitivity studies highlight the impact of different components and hyper-parameters in our approach. Second, to address data heterogeneity in a supervised federated learning framework, we propose Federated Domain Generalization with Label Smoothing and Balanced Decentralized Training (FedSB). FedSB utilizes label smoothing at the client level to prevent overfitting to domain-specific features, thereby enhancing generalization capabilities across diverse domains when aggregating local models into a global model. Additionally, FedSB incorporates a decentralized budgeting mechanism that balances training among clients, improving the performance of the aggregated global model. Experiments on four commonly used multi-domain datasets—PACS, VLCS, OfficeHome, and TerraInc—demonstrate that FedSB outperforms competing methods, achieving state-of-the-art results on three out of four datasets. Collectively, these contributions address critical challenges in FL by enhancing model generalization across diverse and unseen domains in both unsupervised and supervised settings. The effectiveness of FedGaLA and FedSB in addressing data heterogeneity is evidenced by their superior performance in extensive empirical evaluations.Item Trends and Discontinuities: Charting a Canadian Research Agenda(Queen’s Centre for International and Defence Policy, 2012)Strategic foresight aims at reducing uncertainty about future events. It seeks to discern significant trends that are likely to shape future events, and explores both the probability and the impact of possible developments, events or shocks. Strategic foresight can thus be used as a useful tool to establish research priorities. Research can offer comparative insights into the intended and unintended consequences of a range of possible policy options. It can provide an evidence-base in support of decision-making, but also holds out the potential to perform a "red-teaming" challenge function. Such knowledge is particularly valuable in a time of fiscal austerity: when resources are a premium, allocation should be optimized. This concluding chapter seeks to distill the research notes and workshop discussion. The analysis is designed to enhance scholarly capacity on the intersection of transnational trends and concomitant federal stakeholder departments' priorities with respect to national and border security. It does so by highlighting key trends and discontinuities and formulating possible lines of inquiry.Item Non-Hydrostatic Modelling to Improve Nearshore Optical Remote Sensing Algorithms: Applications for Surf Zone Hydrodynamics and Morphodynamics(2024-11-28)Accurate bathymetric data is essential for predicting changes in coastal environments. Traditional surveying methods the seabed provide precise measurements but are limited in coverage and cost. Video remote sensing offers a cost-effective alternative that can capture wave, current, and bathymetric information in dynamic nearshore areas. However, reliable methods are needed to extract accurate data from imagery across various wave conditions. This thesis aims to enhance two retrieval algorithms used to estimate bathymetry and alongshore currents from optical data. Using the non-hydrostatic numerical model SWASH (Simulating WAves til SHore), synthetic wave and current fields are generated to simulate storm events at the US Army Corps of Engineers Field Research Facility in Duck, NC. The simulations complement field observations and offer a new method of identifying sources of error in optical data. The first of these algorithms is cBathy, which faces challenges during large offshore wave conditions, which causes signal interference. By incorporating SWASH-generated water surface elevation data, the errors in depth estimates were reduced, especially in areas of wave breaking. The results of different versions of cBathy are also investigated, which are discussed alongside recommendations for further refinement. The second of these algorithms calculates alongshore current speed, a key driver of bathymetric change. A new version of the Optical Current Meter method (OCM) is developed using SWASH model results and field observations. The updated algorithm lowers measurement error and improves optical measurements across a broader range of conditions. The open-source code is now included in the Coastal Imaging Research Network (CIRN) Video-Currents-Toolbox and is available to other researchers. Informed by the numerical model and field sensor observations, the results provide improved estimates of coastal bathymetry and alongshore currents from optical data, especially during storm conditions that are challenging to monitor.Item Security Implications of Demographic Change: A Canadian Perspective(Queen’s Centre for International and Defence Policy, 2012)The world is at a demographic crossroads. Hitherto, high birth rates had ensured predominantly young populations with few older people. War and epidemics, such as the plague, would intervene to depress population growth. By contrast, depressed population growth today is a function of an historically unprecedented decline in birth rates: women are consistently having fewer (or no) children than at any previous time in history (for reasons that are beyond the scope of this research note). Demographically, the world is entering unknown territory owing to historically unprecedented changes in the three variables that make up demography: fertility, mortality, and migration. Differentials in fertility and mortality are not just affecting population structure. Population structure affects political stability, and political instability tends to be a catalyst for migration. By gaining a better grasp of the demographic drivers of political and economic in/stability, Canada can take strategic action to mitigate push factors of migration. Canada's capacity to act in concert with allies, however, is constrained by the costs and stagnant tax base associated with population aging which, in an age of fiscal austerity, is bound to increase competition over scarce resources among different policy priorities and strategic objectives.
Communities in QSpace
Select a community to browse its collections.