U.S. GLOBEC Transitions:
Towards a fully integrated climate-to-fish model

Schematic of the redesigned CCSM data flow. Of note is the new composite ocean
merging a global POP and a regional ROMS ocean.

An effort is underway to couple the Regional Ocean Modeling System, one of the principle modeling tools in use within U.S. GLOBEC, to the Community Climate System Model (CCSM) developed and maintained at the National Center for Atmospheric Research. In large measure due to the efforts of GLOBEC investigators, the ROMS system now includes multiple options for ecosystem studies (NPZD-type models , e.g., NEMURO) with a range of complexities, as well as bio-energetic models (e.g., NEMURO.FISH). Individual based models for higher trophic levels are also currently being added to ROMS. The incorporation of ROMS within the CCSM framework thus offers immediate opportunities for regional downscaling of oceanic circulation, and the explicit inclusion of ecosystem dynamics.

All the components of the CCSM communicate with each other through a flux coupler, responsible for mediating fluxes among the components, as well as for re-gridding and time synchronization of the various components. The coupler is currently based on MCT (the Model Coupling Toolkit) with an ESMF (Earth System Modeling Framework) version under development. As part of the coupling of ROMS to CCSM, a new composite ocean was developed which couples and synchronizes the global ocean model (the Parallel Ocean Program) and ROMS. Multiple regional domains can be simultaneously coupled to the global ocean model, and each regional patch can have its own ecosystem model.

The following figure shows results utilizing the composite ocean where the Northeast Pacific is simulated with a high-resolution ROMS nested within a global POP grid. The regional model, on the right, has a more realistic representation of the upwelling physics, both in spatial scale and structure, and upwelling temperature. For this simulation, only one-way coupling was used, with the global ocean and atmosphere driving the regional model. The capability for two-way coupling back to the atmosphere is presently being tested.

An upwelling event in the Northeast Pacific as represented by POP (left) and ROMS (right).

Overall, our results show that a hybrid climate/regional model can improve the physical representation in regions of importance to ecosystems. Our approach allows for flexibility in designing a simulation with multiple patches at various resolutions. There are plans for similar coupling of regional atmospheric models using CAM and WRF, thereby enabling simultaneous regional downscaling of atmosphere and ocean.