Prediction of contaminant dispersion in the Gulf of St. Lawrence via Deep Learning

In this project, we developed and trained machine learning models for predicting tracer and contaminant dispersionin the Gulf of St. Lawrence, leveraging data collected by the RQM/MEOPAR Tracer Release Experiment (TReX). There have been few attempts using machine learning to study ocean dispersion and none in the Gulf of St. Lawrence. A successful machine learning model for this task would have the potential to pro-vide rapid estimates that could be used in combination with numerical models and field observations to further assess contingency plans.