388 lohn Woods
18.6.1 Particles and snb-populations
The LE method is an example of agent-based computer modelling. The plankton
populations are represented by a set of particles, each of which behaves like an
individual plankter, and carries information about a dynamic sub-population of
identical plankton. The trajectory of each particle is computed in (typically) 30
minute time steps. During each step, the particle is displaced by three processes:
(1) advection by the ocean current or upwelling, (2) random displacement by turbulence in the mixing layer, and (3) vertical motion relative to the water by sinking or
swimming. Turbulence in the mixing layer causes trajectories to diverge.
The LE method includes particle management rules designed to control errors in
the simulated environment and demography due to local unevenness in the number
of particles and in the size of the sub-populations. As the integration proceeds, particles are split (and their sub-populations shared out), or two or more particles (and
their sub-populations) are combined. Ultimately the accuracy of the simulation
depends on the number of independent trajectories.
18.6.2 Physiology and behaviour
Phytoplankton physiology is based on energy and nitrogen pools. When the
energy pool becomes negative, ali the plankters in a particle's sub-population die of
energy starvation (natural mortality). When the pools both exceed their respective
thresholds for reproduction, the number ofplankters in a particle's sub-population
is doubled. Grazing reduces the number of plankters in a sub-population. Zooplankton physiology is based on body weight, which is increased by grazing. On
reaching the threshold weight for reproduction, the zooplankton produces offspring, which comprise a new sub-population carried by a new particle. The number of zooplankton in a particle's sub-population decreases as the result of
camivorous predation, which occurs at a rate determined by the prey visibility. The
zooplankton have simple control rules for diel migration (to reduce losses to predation) and for foraging (to maximize ingestion).
18.6.3 Ambient environment
The biological equations use a particle's ambient environment to update the biological attributes of the plankton in its sub-population. The ambient environment is
computed by interpolating the fields of environmental variables (light, temperature, nutrient concentration, prey concentration, etc.) to the particle's location.
These fields are described by an Eulerian grid with a one-metre resolution. The
profiles of physical and chemi cal variables change in response to fluxes through
the sea surface, and biofeedback in the water column. The biological environment
is computed from the plankton populations in each layer of the Eulerian grid.
18.6.4 Initial and boundary conditions
In order to integrate the WB model by the LE method, we need initial and boundary conditions for the water column. VEWTrackgen (18.19) generates these auto-
18.6.1 Particles and snb-populations
The LE method is an example of agent-based computer modelling. The plankton
populations are represented by a set of particles, each of which behaves like an
individual plankter, and carries information about a dynamic sub-population of
identical plankton. The trajectory of each particle is computed in (typically) 30
minute time steps. During each step, the particle is displaced by three processes:
(1) advection by the ocean current or upwelling, (2) random displacement by turbulence in the mixing layer, and (3) vertical motion relative to the water by sinking or
swimming. Turbulence in the mixing layer causes trajectories to diverge.
The LE method includes particle management rules designed to control errors in
the simulated environment and demography due to local unevenness in the number
of particles and in the size of the sub-populations. As the integration proceeds, particles are split (and their sub-populations shared out), or two or more particles (and
their sub-populations) are combined. Ultimately the accuracy of the simulation
depends on the number of independent trajectories.
18.6.2 Physiology and behaviour
Phytoplankton physiology is based on energy and nitrogen pools. When the
energy pool becomes negative, ali the plankters in a particle's sub-population die of
energy starvation (natural mortality). When the pools both exceed their respective
thresholds for reproduction, the number ofplankters in a particle's sub-population
is doubled. Grazing reduces the number of plankters in a sub-population. Zooplankton physiology is based on body weight, which is increased by grazing. On
reaching the threshold weight for reproduction, the zooplankton produces offspring, which comprise a new sub-population carried by a new particle. The number of zooplankton in a particle's sub-population decreases as the result of
camivorous predation, which occurs at a rate determined by the prey visibility. The
zooplankton have simple control rules for diel migration (to reduce losses to predation) and for foraging (to maximize ingestion).
18.6.3 Ambient environment
The biological equations use a particle's ambient environment to update the biological attributes of the plankton in its sub-population. The ambient environment is
computed by interpolating the fields of environmental variables (light, temperature, nutrient concentration, prey concentration, etc.) to the particle's location.
These fields are described by an Eulerian grid with a one-metre resolution. The
profiles of physical and chemi cal variables change in response to fluxes through
the sea surface, and biofeedback in the water column. The biological environment
is computed from the plankton populations in each layer of the Eulerian grid.
18.6.4 Initial and boundary conditions
In order to integrate the WB model by the LE method, we need initial and boundary conditions for the water column. VEWTrackgen (18.19) generates these auto-