Date of Award

11-4-2019

Document Type

Thesis

Degree Name

Master of Science (MS)

First Advisor

Wondwosen Seyoum

Second Advisor

William L. Perry

Abstract

This research explores the use of Sentinel-2 satellite to determine the spatiotemporal patterns of lake water quality indicators (e.g. chlorophyll a) in Lake Bloomington and Evergreen Lake. Lake water quality issues related to algal blooms is a serious problem in basins with abundant agricultural lands causing harmful effects to freshwater ecosystems such as pollution of beaches, taste and odor problems in drinking water, depletion of oxygen levels causing fish kills and the issue of water exceeding safe drinking water standards. Developing monitoring techniques using various water quality indicators of algal blooms is crucial. In this project, remote sensing and field sampling methods were employed to assess the state of water quality of two small lakes, Lake Bloomington and Evergreen Lake, in Central Illinois. Water samples were collected from selected locations from the lakes to test for various water quality variables including nitrate, phosphorus and chlorophyll a. An exo sonde instrument and secchi disk was used to measure additional water quality parameters such as turbidity, secchi depth, and temperature. Concurrent satellite images obtained from Sentinel-2 with flyover with ±5 days were processed and analyzed, and the results were compared with field sampling data. Single and multiple pixel analyses were conducted on various algorithms such as Bottom-of-Atmosphere (BOA), Maximum Chlorophyll Index (MCI), and band ratios. These algorithms were tested to identify the best algorithm for estimating water quality parameters using satellite data for the two lakes. A regression analysis was conducted to derive a linear model which was used to create water quality indicator maps that showed the spatial pattern of algae in the lakes. From the results of the research, Lake Bloomington was more turbid and had higher concentrations of chlorophyll a than Evergreen Lake. Except for band ratio of B1/B2 of Sentinel-2 data, a poor regression relationship between satellite and field water quality values was observed for Lake Bloomington. This poor relationship could be due to the high turbidity of the lake. Evergreen Lake, on the other hand, showed a stronger relationship between satellite values and chlorophyll a. Generally, spatial analysis reveals that chlorophyll a distribution was heterogeneous, and it increased from downstream areas to upstream areas.

Comments

Imported from ProQuest AmbroseIgho_ilstu_0092N_11597.pdf

DOI

http://doi.org/10.30707/ETD2020.Ambrose-Igho.G

Page Count

65

Download

Share

COinS