Abstract

Harmful algal blooms (HABs) are a serious and growing risk to waterbodies around the world. Excessive algal growth, especially when it involves cyanobacteria, impairs the functioning of aquatic ecosystems and threatens the health of humans and animals using the water. There is currently great interest in developing effective prevention and mitigation strategies for these blooms, particularly in ecologically and economically vital freshwater lakes. Such development, however, requires detailed understanding of HAB drivers and dynamics, which can be highly complex and difficult to characterize. High-frequency, high-resolution imagery from earth observation satellites provides spatially and temporally comprehensive data on global lakes--data which can be used to fill gaps in understanding left by in situ studies of HABs, which are by nature limited in spatial and temporal scope. We developed a workflow for extracting accurate, usable estimates of chlorophyll-a (chl-a, a plant pigment commonly used as an index for algal biomass), turbidity, and water temperature from the ESA's Sentinel 2 satellite series and NASA's MODIS sensor using the Google Earth Engine (GEE) platform. This workflow includes a novel approach to generating chl-a and turbidity retrieval algorithms, combining physics-based and empirical models to generate more accurate and explainable estimates of both chl-a and turbidity measurements, as well as methods for QA/QC and sub-sampling of imagery. This workflow can be easily adapted to any lake around the world, but we demonstrate it on a case study of Utah Lake, a large, shallow, eutrophic lake and vitally important natural resource in the semi-arid Utah Valley, Utah, USA. We use the dataset obtained from the GEE workflow and some additional processed Sentinel 2 imagery to analyze large-scale patterns in algal blooms on Utah Lake and the relationships between algal blooms and water column temperature and turbidity. Our results confirm pronounced spatial heterogeneity in the occurrence of algal blooms on Utah Lake and reveal distinct intra- and inter-seasonal trends. The shallow, hydrologically isolated Provo Bay on the east side of the lake was identified as a major bloom hotspot. Another bay at the south end of the lake, Goshen Bay, was also a hotspot, but to a lesser extent than Provo Bay. Both bays also appear to act as bloom "incubators" which tend to bloom earlier in the season and push water with high chl-a concentrations out into the main lake. We found that intense blooms were relatively rare, especially on the open lake away from the shoreline and the shallow bays, and were moderately associated with lower turbidity values. We use these results to make several recommendations for future bloom monitoring and mitigation strategies on Utah Lake. This research provides valuable insight for policymakers, water quality managers, and others involved in Utah Lake research and management. In addition, it supports ongoing development of best practices and methods for using satellite imagery to study water quality, and provides an accessible tool which can be applied to any waterbody in the world to conduct a similar analysis.

Degree

PhD

College and Department

Ira A. Fulton College of Engineering; Civil and Environmental Engineering

Rights

https://lib.byu.edu/about/copyright/

Date Submitted

2025-12-16

Document Type

Dissertation

Keywords

Utah Lake, chl-a, remote sensing, Sentinel, MODIS, algal blooms

Language

english

Download

Included in

Engineering Commons

Share

COinS