Development of a climate-to-fish-to-fishers model: data structures and domain decomposition

Kate Hedstrom(UAF), Jerome Fiechter(UCSC), Kenneth A. Rose(LSU), Enrique N. Curchitser(Rutgers), Miguel Bernal(Rutgers)

We are developing a model for examining the long-term population cycles of anchovy and sardine in the California Current system. This presentation will focus on the numerical methods and computing considerations for dynamically coupling the physics, lower trophic level, and upper trophic level models. The physics is represented by the Regional Ocean Modeling System (ROMS), which has been designed to run on parallel architectures using the Message Passing Interface (MPI) that is common on the current generation of supercomputers. The lower trophic level model is a version of the NEMURO nitrogen-phytoplankton-zooplankton model. The fish community is represented using a full life cycle, individual-based approach. When combining the models, which are a mix of Eulerian and Lagrangian approaches, care must be taken to ensure that the modeling system continues to function on tens of processors with many thousands of individuals without running into resource limitations – or worse, results that depend on the number of processors. We will describe several of the numerical algorithms we are using in our dynamically-coupled model, including those for spatially-locating eggs from spawning and scaling the predator-prey interactions among the fish species represented in the model and between the fishers and the fish.