Defense Presentation Title: Estimating and Forecasting the Spatio-Temporal Hypoxia Dynamics in Northern Gulf of Mexico

Advisor: Dr. Daniel R. Obenour, Faculty Fellow and Associate Professor in the Department of Civil, Construction and Environmental Engineering

Abstract: Hypoxia is an environmental issue affecting the northern Gulf of Mexico. Models studying the causes and consequences of hypoxia in this region rely on sparse hypoxia measurements that do not leverage all available data. To address this limitation, I enhanced an existing geostatistical hypoxia model by integrating new observations of dissolved oxygen and simulations from a hydrodynamic-biophysical model. This fusion-based approach is used to estimate hypoxic area across each summer. We also expanded the space-time geostatistical model to estimate hypoxic volume by integrating a novel transformation of hypoxic layer thickness. Finally, I developed a decision tree ensemble to predict the inshore-offshore distribution of hypoxia. This novel application of machine learning can be used to improve the spatial resolution of hypoxia forecasts, which can potentially inform fisheries and ecosystem management.