Date Approved
8-8-2025
Graduate Degree Type
Thesis
Degree Name
Biology (M.S.)
Degree Program
Annis Water Resources Institute
First Advisor
Sean Woznicki
Second Advisor
Bopaiah Biddanda
Third Advisor
Michael Philben
Academic Year
2024/2025
Abstract
Freshwater estuaries are natural contributors to the carbon cycle including production and emission of methane (CH4) and carbon dioxide (CO2), potent greenhouse gases (GHGs); however, estimates of their contribution to regional and global GHG emissions is largely unconstrained. Few studies have examined the quantification and drivers of lake CH4 and CO2 production in the Great Lakes region and how it may differ in response to anthropogenic development. In this study, CH4 and CO2 emissions were measured from three drowned river mouth estuaries (DRMs) along the eastern shore of Lake Michigan in 2024. The DRMs exist along a latitudinal gradient ranging from mesotrophic in the north (Muskegon and White Lake) to hypereutrophic in the south (Lake Macatawa) and experience varying degrees of anthropogenic influence from forested (north) to agricultural and urban (south) riparian and watershed land cover. We deployed low-cost, autonomous floating CH4 samplers in littoral zones and conducted monthly discrete sampling for CH4 and CO2 in pelagic zones. Sentinel-3 Ocean and Land Colour Instrument, Moderate Resolution Imaging Spectrometer, and gridded meteorological datasets were then used to calculate remotely sensed environmental variable proxies used to estimate CH4 flux with machine learning via Extra Trees Regression (ETR). Model performance (R2 = 0.48, MAE = 0.41) indicated ability to capture trends in CH4 flux and extend predictions to other DRM estuaries, providing a successful framework for reducing the current uncertainty in lakes’ GHG emissions. Results from this study suggest high inter-lake variability indicating the potential linkage between anthropogenic stress and limnetic flux. While further aquatic carbon flux quantification efforts are necessary for constraining natural sources within the global carbon budget, the remote sensing and modeling techniques implemented 5 provide novel methodology to help reduce the current knowledge gap in global limnetic CH4 quantification.
ScholarWorks Citation
Greene, Jillian A., "Assessing greenhouse gas emissions from Michigan’s drowned river mouths using in situ and remote sensing methods" (2025). Masters Theses. 1162.
https://scholarworks.gvsu.edu/theses/1162
