Degree

Doctor of Philosophy (PhD)

Department

Civil and Environmental Engineering

Document Type

Dissertation

Abstract

Vibrio vulnificus is a halophilic gamma proteobacterium that is autochthonous to coastal waters and is responsible for 50% of seafood-related deaths in the United States. This study presents a series of nowcasting and forecasting models for predicting vibrio vulnificus abundance in oysters by identifying the long-range dependence of vibrio vulnificus abundance on antecedent environmental conditions and detecting the environmental conditions using satellite remote sensing technology. It was discovered that vibrio vulnificus abundance exhibits a long-range dependence on antecedent environmental conditions which can be characterized by seven independent environmental predictors (stressors) including sea surface temperature (SST), water level (WL), sea surface salinity (SSS), solar radiation (SR), pH, wind speed (WS), and chlorophyll a. Advanced remote sensing algorithms were developed using the XGBoost machine learning method for the estimation of three environmental predictors (SR, SSS, and pH). The developed solar radiation algorithm is the best among all existing solar radiation models in terms of accuracy, high resolution, and global applicability. The sea surface salinity and pH algorithms are also globally applicable. By employing the environmental data for the seven independent predictors along with their time lags, vibrio vulnificus concentration data collected from three U.S. coastal waters (in Louisiana, Mississippi, and North Carolina), and XGBoost method, four ensemble models with differing lead times (including 0-day, 1-day, 6-days, and11-days) were developed for the estimation of vibrio vulnificus concentration in oysters in coastal waters of the United States. The ensemble models are capable of explaining over 75% of variations in observed vibrio vulnificus concentrations, greatly enhancing the ability to forecast the potential risk of vibrio vulnificus infection and to intervene to reduce the risk to public health.

Date

7-26-2024

Committee Chair

Deng, Zhiqiang

Download

Share

COinS