Nutrients-Plankton-Detritus framework

LOBSTER, NPZD, and ImplicitBiology are all built on a common modular framework called NutrientsPlanktonDetritus. This framework organises the biogeochemistry into five pluggable component slots:

Preset constructors

For common configurations, convenience constructors assemble these components with sensible defaults. You can build each by passing a grid:

• LOBSTER(grid) — medium-complexity model with phytoplankton, zooplankton, nitrate, ammonia, and dissolved/particulate detritus. See the LOBSTER page.
• NPZD(grid) — simple four-compartment nutrient–phytoplankton–zooplankton–detritus model from Kuhn et al. (2015). See the NPZD page.
• ImplicitBiology(grid) — nutrient-limited community productivity with no explicit plankton biomass. See the ImplicitBiology page.

All preset constructors accept the same optional keyword arguments as NutrientsPlanktonDetritus itself — in particular inorganic_carbon, oxygen, light_attenuation, sediment, and scale_negatives — so you can extend any of them without building from scratch:

using OceanBioME, Oceananigans

grid = RectilinearGrid(size = 10, extent = 200, topology = (Flat, Flat, Bounded))

model = NonhydrostaticModel(grid;
                            biogeochemistry = LOBSTER(grid;
                                                      inorganic_carbon = CarbonateSystem(),
                                                      oxygen = Oxygen()))

Custom configurations

You can assemble a fully custom model by calling NutrientsPlanktonDetritus directly and specifying each component:

biogeochemistry = NutrientsPlanktonDetritus(grid;
                                            nutrients  = Nutrients(NitrateAmmonia(), PO₄, Fe, nothing),
                                            plankton   = PhytoZoo(grid),
                                            detritus   = DissolvedParticulate(grid),
                                            inorganic_carbon = CarbonateSystem())

The framework dispatches through the components to assemble the required tracers and auxiliary fields automatically, so all standard Oceananigans boundary conditions and output writers work as normal.

For a step-by-step guide to implementing your own plankton component, see Implementing new models.