Detecting temporal variability in five water quality parameters of Kentucky Lake using time series analysis
Academic Level at Time of Presentation
Junior
Major
Biology
Minor
Applied Statistics
List all Project Mentors & Advisor(s)
Kate He, PhD; Susan Hendricks; PhD
Presentation Format
Poster Presentation
Abstract/Description
Kentucky Lake, a large man-made reservoir on the Tennessee River system, has been part of the long-term monitoring program of Murray State University since July 1988. Multiple studies have tracked the changes in water quality and productivity of the reservoir in the last three decades. In this study, we focused on five water quality parameters (Secchi depth, turbidity, total nitrogen, dissolved oxygen, and temperature) collected from the three main channel sites from 1988 to 2017. Time series analysis was used to detect the temporal patterns of these five parameters. The analysis of the data included the construction and decompositions of time series for each individual parameter, as well as fitting linear models for the trend and time components of the time series, and smoothing using LOWESS lines. We found significant increases in the Secchi depth (p = 2.24e-15) and dissolved oxygen (p = 0.04), and decreases in turbidity (p = 9.95e-16) over the total 30-year period. Total nitrogen and temperature show no significant long-term trend. Our analysis can be expanded to look at seasonal trends, lake phenology, short-term trends, or cross-correlations between parameters. By detecting the temporal changes in water quality parameters, we can explore the effects on organisms in the lake and monitor changes that may be caused by climate change, human land-use, pollution, etc. The results of our analysis can also be used to predict future changes under different global change scenarios, thus providing useful information for species conservation and lake management efforts.
Spring Scholars Week 2019 Event
Sigma Xi Poster Competition (Juried)
Detecting temporal variability in five water quality parameters of Kentucky Lake using time series analysis
Kentucky Lake, a large man-made reservoir on the Tennessee River system, has been part of the long-term monitoring program of Murray State University since July 1988. Multiple studies have tracked the changes in water quality and productivity of the reservoir in the last three decades. In this study, we focused on five water quality parameters (Secchi depth, turbidity, total nitrogen, dissolved oxygen, and temperature) collected from the three main channel sites from 1988 to 2017. Time series analysis was used to detect the temporal patterns of these five parameters. The analysis of the data included the construction and decompositions of time series for each individual parameter, as well as fitting linear models for the trend and time components of the time series, and smoothing using LOWESS lines. We found significant increases in the Secchi depth (p = 2.24e-15) and dissolved oxygen (p = 0.04), and decreases in turbidity (p = 9.95e-16) over the total 30-year period. Total nitrogen and temperature show no significant long-term trend. Our analysis can be expanded to look at seasonal trends, lake phenology, short-term trends, or cross-correlations between parameters. By detecting the temporal changes in water quality parameters, we can explore the effects on organisms in the lake and monitor changes that may be caused by climate change, human land-use, pollution, etc. The results of our analysis can also be used to predict future changes under different global change scenarios, thus providing useful information for species conservation and lake management efforts.