Understanding Nutrient Loads Delivered From Mississippi-Atchafalaya River Basin to Gulf of Mexico Using Tree-Based Regression Model

The objective of this study is to identify and develop predictive models of nutrient loads collected from United States Geological Survey (USGS) monitored stations in the Mississippi-Atchafalaya River Basin to (MARB) and to better understand the nature of excess nutrient loads delivered from the MARB to the Gulf of Mexico. Random Forest Regression (RFR) and Gradient Boosting Regression (GBR) are used to capture the temporal structure of nutrient load changes. The results will be used in further studies to estimate and predict nutrient concentrations and fluxes into the Gulf of Mexico. In future, innovative remote sensing techniques could be employed to gather information about nutrient loads, and engineers may contribute to the design and implementation of more intelligent systems.