Ocean deoxygenation poses a significant threat to marine ecosystems and global biogeochemical cycles. Traditional ocean biogeochemical models (BGCs), while effective at capturing large-scale processes, often struggle with parameter uncertainties and have limited capacity to resolve regional variability and emerging trends in oxygen dynamics. In this session, we welcome studies that work with  traditional models, hybrid modeling frameworks that integrate data-driven approaches—such as machine learning and statistical inference—with mechanistic ocean BGC models, and fully data-driven approaches, evaluating their effectiveness in improving the representation of key processes driving deoxygenation.

The session will explore several critical questions, including:

1. What mechanisms drive deoxygenation variability, and how does it interact with  ocean warming, stratification, and nutrient cycles? How predictable are these dynamics?
   
   
2. How well can current models capture these processes? Can data-driven approaches help reduce uncertainty and improve parameterizations?
   
   
3. How can recent advances in ocean biogeochemical models– including data-driven approaches and improved parameterizations–be leveraged to design more effective oxygen monitoring programs, especially  in under-studied regions?

• Air-Sea Interactions -
• Ocean Biology and Biogeochemistry -
• Physical-Biological Interactions -