An assessment of present and historical (1984-2012) Lake Diefenbaker water clarity and chlorophyll-a concentration using Landsat imagery
Date
2015-01-06
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
ORCID
Type
Degree Level
Masters
Abstract
Abstract: The use of earth observing satellites can be an effective supplement or alternative to traditional field sampling. The Landsat series of satellites have been particularity useful in assessing water quality in lakes, oceans, and reservoirs. This study utilized Landsat 5 and 7 imagery to model Secchi disk depth (SDD) and chlorophyll-a concentrations (Chl-a) at Lake Diefenbaker, Saskatchewan. I used data from these Landsat satellites to answer the following questions: First, can models that predict water quality (SDD and Chl-a concentration) be developed for Lake Diefenbaker using Landsat imagery? Second, can these models identify trends that have taken place at the reservoir from 1984-2012? Third, can I determine if ephemeral events like algal blooms or flooding have an effect on the reservoir? Novel models were developed from data collected in 2011 and 2012 that could predict SDD and Chl-a concentrations in the reservoir (linear regression, model I). These models explain less variation than comparable studies, but the loss in explanatory power is made up by their ability to predict data from any Landsat image of the reservoir. My study showed that predicted SDD and Chl-a concentration were positively related, an atypical relationship in freshwater systems. During the archive study period (1984-2012), both mean seasonal SDD and mean seasonal Chl-a have significantly decreased throughout the reservoir (p<0.05, regime-shift analysis). Spatially, the greatest decrease in SDD was closest to the major inflow the SSR, while downstream areas in the reservoir have decreased minimally. There was a decline in Chl-a concentrations that was spatially consistent throughout the reservoir. There was a significant negative relationship between flow rate and both water clarity and Chl-a concentrations (P<0.05, model II linear regression). Algal blooms occurred sporadically throughout the study period. There were blooms in 9% of images analyzed. Blooms typically occurred in the Qu’Appelle arm of the reservoir in the late summer and fall. The water quality data extracted by this study can be useful to many future studies, as historical data is absent for much of the reservoir’s history.
Description
Keywords
Water Quality, Landsat, Remote Sensing, Chlorophyll, Algal Blooms, Long -term Trends
Citation
Degree
Master of Science (M.Sc.)
Department
Biology
Program
Biology