Dr. Shuqi Lin
Research Scientist
Watershed Hydrology and Ecology
Environment and Climate Change Canada
Burlington, Ontario, Canada
Seminar: Prediction of algal blooms and dissolved oxygen via multi-model machine learning approach
Abstract:
With increasing lake monitoring data, data-driven machine learning (ML) models are able to capture the complex dissolved oxygen (DO) variations and algal bloom dynamics in lakes, which cannot be completely described in process-based (PB) models. Combined with PB models, multi-model machine learning approaches were applied to predicting algal blooms, seasonal changes in algal chlorophyll concentrations (Chl), DO concentration, as well as hypoxia. In algal bloom prediction, three ML predictive workflows were tested in a small mesotrophic lake. The performance of the ML models was superior to the PB model. The hybrid model which blending results from PB model into ML training was developed and further improved the prediction of the timing and magnitude of algal blooms. In DO prediction, two kinds of ML models and a mixed ML approach were developed and applied in 5 lakes with various trophic levels. Models showed promising results in simulating DO concentrations and detecting hypoxia. These results demonstrate that ML approaches could be used to forecast algal bloom and hypoxia to support water management under short- and long-term management plans.
About Dr Lin:
I have been a Research Scientist in Watershed Hydrology and Ecology Research Division Environment and Climate Change Canada since 2022. After getting my PhD degree in Queen’s University, I continued my academic career in Queen’s University and Uppsala University (Sweden) as a post-doc.
So far, my research interests include physical limnology, numerical simulation of lake hydrodynamics and water quality; cyanobacteria bloom and hypoxia forecast via process-based model and machine/deep learning algorithms.
Currently, I am working on multi-year hydrodynamics and water quality simulations in Lake Winnipeg/ Lake of the Wood/ Great Slave Lake.