This session, led by Véronique Garçon (LEGOS, France), highlights the benefits of using models to guide the design of observations, and vice versa, the benefits of using observations to guide the design of models. The rationale can be: to formalize conceptual ideas, to test hypothesis, to fill gaps in space and time, to make predictions for the future, and/or to quantify feedbacks (i.e. nonlinear responses). A model is a tool. Which types of models can be used for describing physical processes, chemical processes, and biological processes? How can we quantitatively assess models performance? What is meant by validation, evaluation or benchmarking? Then, we will demonstrate how an ecosystem model may mimic Oxygen Minimum Zones (OMZ) in the Eastern Boundary Upwelling Systems. Interactions between topographic features, coastal atmospheric jets, coastally trapped and internal waves, Ekman transport and submesoscale and mesoscale circulations complicate our ability to represent coastal dynamics in climate models. Quantification of the potential impacts of climate change, and the societal consequences, can be improved with recent advances in physical and biogeochemical observations. We will show how the continuous interactions between models and observations will improve our understanding of the mechanical and thermal forcing of the near-surface ocean in coastal areas, air-sea exchanges, transport and mixing at scale of kilometers or less, and biogeochemical exchanges, interactions, and cycling.