When Marine Ecosystem Models Meet Observations

Marine ecosystem models are important tools for predicting the biochemical state of the ocean. Models are often used to fill gaps in observational data, yet unconstrained model simulations can quickly diverge from realistic results. As a result, including data assimilation into modeling efforts is an essential step toward improving model predictions. A particular challenge in incorporating data into ecosystem modeling is the large number of biogeochemical tracers in models (e.g. nutrients, plankton), and the lack of available observations.

In this talk, we present data assimilation in a regional model of the US West Coast using the Regional Ocean Modeling System (ROMS) and show applications of biogeochemical parameter estimation. We highlight how increased model complexity in more advanced biogeochemical models can lead to more realistic model output while complicating data assimilation.