Towards the development of integrated modelling frameworks in aquatic biogeochemistry: A Bayesian synthesis of empirical knowledge and model predictions

点击翻译按钮获取中文摘要

My research program focuses on the study of aquatic biogeochemical cycles and develops novel modelling techniques for assisting environmental management. My current research promotes a critical approach to modelling that integrates Bayesian inference techniques with mathematical modelling for rigorously assessing the uncertainty underlying model predictions. Previous support from NSERC has offered a platform to conduct fundamental research in the area of model-based environmental management. The Hamilton Harbour, a eutrophic embayment located at the western tip of Lake Ontario, is one of the main sites used to disseminate the benefits of my research program. My group has developed a network of models to capture the variability of key components of the lower food web and, most importantly, to offer an accurate representation of the cause-effect relationships underlying the eutrophication phenomenon. Building upon the findings of my recent modelling work in the Hamilton Harbour watershed and receiving waterbody, the proposed research program aims to shed light on two critical aspects of the ecosystem dynamics, i.e., phosphorus release driven by sediment diagenesis and nutritional/biochemical control of the primary producer-consumer interface. The proposed research will offer fundamental knowledge on two hot topics in the field of aquatic biogeochemistry. The eutrophic nature of the Hamilton Harbour makes the study of sediment diagenesis and biochemical control of the plant-animal interface particularly intriguing. The anticipated insights into the capacity of the two factors to modulate the response of the system will offer a unique case study in the international literature. The foundation of my modelling work on Bayesian inference techniques is perfectly suitable for integrating different submodels into one coherent framework, while rigorously assessing the uncertainty associated with model structures and parameters. The proposed research program aims to illustrate how empirical knowledge and rational modelling can be used to guide system restoration in highly disturbed settings.