Mehak Chohan
Research Mentor(s): Xiaofeng Liu
Mentor Department: Michigan Institute for Data and AI in Society
Authors:
Session: Session 6 (3:00pm – 3:50pm)
Presentation Type: Poster 93
Abstract
Freshwater ecosystems are crucial to both biodiversity and human well-being, and accurate streamflow prediction is essential for effective water resource management. However, streamflow dynamics are increasingly shaped by climate change and land use, making them more complex and harder to model. Traditional approaches, such as process-based and empirical statistical models, often face scalability issues and struggle to capture the nonlinear interactions among environmental variables. In this study, we apply a Long Short-Term Memory (LSTM) deep learning model to predict streamflow across the Great Lakes Basins. The model incorporates dynamic climate variables derived from preprocessed Climate Forecast System Reanalysis (CFSR) data, combined with static basin attributes. Our results demonstrate the potential of AI-based methods to enhance hydrological forecasting and inform sustainable water management in the Great Lakes region.
