We implement an existing marine ecosystem model, CoSiNE, within the CESM2.1.5 ocean general circulation model (POP2) framework and evaluate its performance against the default MARBL biogeochemical configuration under identical physical forcing. The objective is to assess the behaviour of the CoSiNE-based configuration in simulating large-scale marine biogeochemical and ecosystem dynamics.
The coupled system demonstrates stable performance over short- to intermediate-term integrations and reproduces the dominant global patterns of key biogeochemical tracers. In particular, the CoSiNE configuration shows notable improvements in the representation of marine nutrient distributions. It more clearly resolves the spatial contrast between oligotrophic subtropical gyres and nutrient-rich upwelling regions, and produces a more realistic vertical structure of subsurface nutrient regeneration and transport compared with MARBL.
The model also captures the large-scale distribution of phytoplankton biomass and chlorophyll, with improved responses in high-productivity regions and upwelling systems, indicating a more consistent representation of nutrient-controlled ecosystem dynamics.
For the carbon system, the model reproduces the first-order structure of surface pCO₂ and its large-scale gradients, although regional biases remain and are subject to ongoing refinement. These biases are primarily associated with uncertainties in carbon system parameterizations and air–sea exchange representation.
Relative to MARBL, the CoSiNE configuration exhibits a clearer and more structured representation of nutrient cycling and subsurface regeneration pathways, suggesting improved performance in simulating key biological pump processes.
Overall, this study demonstrates that the CoSiNE configuration within CESM provides a stable and physically consistent framework for ocean biogeochemical modelling, with particular strengths in the representation of nutrient-driven ecosystem structure, and offers a solid basis for further development and long-term simulations.