Primitive Equation Modelling ofPlankton Ecosystems
389
matically from standard climatologies of the ocean and atmosphere, or from the
user's own data.
The initial conditions describe the state of the ecosystem at the start of the integration, including the profiles of temperature and salinity (and therefore density),
nitrate and ammonia concentrations, dissolved inorganic carbon concentration, pH,
and the number of plankton in each class and their biological state. It is convenient
to initialise an integration before the start of the growing season, in winter when the
plankton populations are changing slowly; we normally choose the 1 st March. The
initial conditions are derived from world ocean climatologies like that of Levitus
(1997). The critical variables are density and nutrient concentrations. The other
variables can be computed by the model using these two initial conditions and the
boundary conditions. As the integration proceeds, errors in the initial plankton populations are forgotten within three years, as the simulation homes in on its demographic attractor (18.12). Mixing 1ayer depth adjusts immediate1y.
The boundary conditions provide the external information needed to compute the
surface fluxes of solar radiation (in 25 spectral bands), net IR radiation, sensible
and latent heat, and carbon dioxide. The internal information comes from the simulation (upwelling, irradiance, mixed layer temperature and salinity and partial pressure of carbon dioxide). The external data are derived from an astronomic al model
(for solar elevation) and atmospheric data. The latter are normally monthly mean
values from Bunker's climatology or from ERA (Re-analysis data set of the European Centre for Medium Range Weather Forecasting). It is also possible to use
ambient weather from ERA or ship meteorological observations. The external data
are computed for each time step before the model integration and stored in a
boundary condition file.
18.7 Virtual Plankton Ecosystems
18.7.1 Introduction
The re suIt of integrating the WB model by the LE method is a massive data set
called the Virtual Plankton Ecosystem. The VPE describes the state of the plankton
ecosystem each time step (typically 30 minutes) for a period of several years. Some
of our VPEs describe changes over one hundred years. The data set describes the
environment at each time step in terms of values at one metre intervals of depth
through the water column, between the sea surf ace and 500m. This Eulerian time
series describes the physical, chemical and biological components of the environment. The biological environment is computed from the plankton sub-populations
ofthe particles in each one metre layer (or passing through it during the time step).
Thus the Virtual Plankton Ecosystem derived from LE integration contains all the
information in a simulation that has been created by Eulerian integration. The environment contains about one gigabyte of data per simulated year.
In addition, the VPE contains time series of the depth and biologic al state variables of the plankton sub-population associated with each particle. For a typical
389
matically from standard climatologies of the ocean and atmosphere, or from the
user's own data.
The initial conditions describe the state of the ecosystem at the start of the integration, including the profiles of temperature and salinity (and therefore density),
nitrate and ammonia concentrations, dissolved inorganic carbon concentration, pH,
and the number of plankton in each class and their biological state. It is convenient
to initialise an integration before the start of the growing season, in winter when the
plankton populations are changing slowly; we normally choose the 1 st March. The
initial conditions are derived from world ocean climatologies like that of Levitus
(1997). The critical variables are density and nutrient concentrations. The other
variables can be computed by the model using these two initial conditions and the
boundary conditions. As the integration proceeds, errors in the initial plankton populations are forgotten within three years, as the simulation homes in on its demographic attractor (18.12). Mixing 1ayer depth adjusts immediate1y.
The boundary conditions provide the external information needed to compute the
surface fluxes of solar radiation (in 25 spectral bands), net IR radiation, sensible
and latent heat, and carbon dioxide. The internal information comes from the simulation (upwelling, irradiance, mixed layer temperature and salinity and partial pressure of carbon dioxide). The external data are derived from an astronomic al model
(for solar elevation) and atmospheric data. The latter are normally monthly mean
values from Bunker's climatology or from ERA (Re-analysis data set of the European Centre for Medium Range Weather Forecasting). It is also possible to use
ambient weather from ERA or ship meteorological observations. The external data
are computed for each time step before the model integration and stored in a
boundary condition file.
18.7 Virtual Plankton Ecosystems
18.7.1 Introduction
The re suIt of integrating the WB model by the LE method is a massive data set
called the Virtual Plankton Ecosystem. The VPE describes the state of the plankton
ecosystem each time step (typically 30 minutes) for a period of several years. Some
of our VPEs describe changes over one hundred years. The data set describes the
environment at each time step in terms of values at one metre intervals of depth
through the water column, between the sea surf ace and 500m. This Eulerian time
series describes the physical, chemical and biological components of the environment. The biological environment is computed from the plankton sub-populations
ofthe particles in each one metre layer (or passing through it during the time step).
Thus the Virtual Plankton Ecosystem derived from LE integration contains all the
information in a simulation that has been created by Eulerian integration. The environment contains about one gigabyte of data per simulated year.
In addition, the VPE contains time series of the depth and biologic al state variables of the plankton sub-population associated with each particle. For a typical