Date of Award:

8-2025

Document Type:

Thesis

Degree Name:

Master of Science (MS)

Department:

Civil and Environmental Engineering

Committee Chair(s)

Sierra N. Young

Committee

Sierra N. Young

Committee

Jeffery S. Horsburgh

Committee

Steve Petruzza

Abstract

Monitoring river water levels and flows is critical for managing water supplies, protecting against floods, supporting ecosystems, and informing infrastructure planning. However, traditional methods rely on sensors installed directly in rivers, which can be expensive, difficult to maintain, and vulnerable to damage—particularly in remote or hazardous locations. In many regions, limited budgets and access challenges result in large gaps in river monitoring networks.

This research explores an innovative alternative: using fixed cameras and artificial intelligence (AI) to monitor rivers from a distance. The approach allows for non-contact observation of water levels and flows by analyzing images captured by cameras installed along riverbanks. Deep learning models automatically identify the river surface in each image, and machine learning techniques convert these visual measurements into estimates of water level and river flow (discharge). This system eliminates the need for physical sensors in the water and can operate continuously and autonomously.

The method was tested at two river sites in northern Utah, representing different river types and environmental conditions. The results showed that the camera-based system could identify both seasonal flow patterns and rapid flow changes during hydrologic events with high accuracy compared to ground truth flow data. The technology was also successfully deployed on low-cost, energy-efficient computing devices, enabling real-time operation even in areas with limited internet connectivity.

By providing a low-cost, scalable, and resilient alternative to traditional monitoring, this AI-driven system can help expand hydrologic monitoring networks, particularly in underserved or data-sparse regions. It offers significant potential for enhancing flood forecasting, improving water management, and supporting climate resilience in communities around the world. As the impacts of climate change and growing water demands increase the need for reliable water data, camera-based monitoring systems present an exciting new tool to help meet this challenge.

Checksum

c06c32660cdf55e0530542ddd50bcf90

Download

Share

COinS

Copyright for this work is retained by the student. If you have any questions regarding the inclusion of this work in the Digital Commons, please email us at .