This website stores cookies on your computer. These cookies are used to collect information about how you interact with our website and allow us to remember your browser. We use this information to improve and customize your browsing experience, for analytics and metrics about our visitors both on this website and other media, and for marketing purposes. By using this website, you accept and agree to be bound by UVic’s Terms of Use for web and social media privacy.  If you do not agree to the above, you can configure your browser’s setting to “do not track.”

Skip to main content
People Departments Maps Buildings Rooms A-Z
University
of Victoria
Apply Library
Sign in Online tools Sign out
Faculty of Graduate Studies

Miranda Reid

  • B.Sc. (Bangor University, 2024)
Notice of the Final Oral Examination for the Degree of Master of Science

Topic

Measuring and Predicting Coastal Change in the Maritimes Using Satellite Imagery and Machine Learning

School of Earth and Ocean Sciences

Date & location

  • Thursday, July 30, 2026
  • 12:30 P.M.
  • Clearihue Building, Room B017

Examining Committee

Supervisory Committee

  • Dr. Thomas James, School of Earth and Ocean Sciences, University of Victoria (Co-Supervisor)
  • Dr. Blake Dyer, School of Earth and Ocean Sciences, UVic (Co-Supervisor)
  • Dr. Gavin Manson, Coastal Geoscientist, Geological Survey of Canada (Outside Member)

External Examiner

  • Dr. David Atkinson, Department of Geography, UVic

Chair of Oral Examination

  • Dr. Tara Troy, Department of Civil Engineering, UVic

Abstract

Coastal change is of particular interest in the Maritime Provinces where coastal sensitivity is notably high. Satellite imagery and machine learning offer valuable tools for measuring and predicting coastal change. The aim of this thesis is to explore these tools and their applications to the study of coastal change in the Maritimes.

To assess the reliability of satellite imagery for measuring coastal change, the CoastSat software and Olthof et al. (2025) datasets are compared to each other and to two air photo analysis datasets. The results of these comparisons reveal that while the satellite imagery methods are strong in measuring beach environments, they struggle with measuring inland estuary and lagoon environments. Additionally, the satellite imagery methods are less equipped for monitoring extremely dynamic coastlines. These results allow for the identification of potential next steps in working with the open source CoastSat software.

The feasibility of a neural network for predicting coastal change using the CanCoast 2025 variables is also explored. A model is developed and trained to classify Prince Edward Island’s coastline as either retreating, stable, or advancing, and is constrained using the Coastal Hazard and Risk Information System dataset. The model is also used in determining the relative explanatory powers of the CanCoast variables. While the model achieves a training and validation accuracy of roughly 78%, it is unable to achieve an accuracy higher than 50% on an unseen test dataset. Despite poor accuracy, insight is still attained regarding the impact of each CanCoast variable, with relief and maximum fetch affecting the training accuracy the most. The outcomes of the model help determine a pathway for the continued development of a neural network for predicting coastal change.

Together, the results of the comparison of satellite imagery methods of measurement and the attempted development of a machine learning model allowed for a roadmap to be identified for continuing to implementing satellite imagery and machine learning into coastal change research in the Maritime provinces.

Back to oral exams