Environmental Engineering Reference
In-Depth Information
These models refer to basic biological processes: dispersal and competition. It is the
variation in the strength and scale of feedbacks between cells in the automaton that
influence the outcomes in terms of structure and scale of patchiness. This illustrates
the general nature of scale-dependent processes underlying self-organized patchi-
ness in ecosystems.
8.2 Cellular Automata: The Components
Though cellular automata can handle very complex spatial situations and quite
difficult rule systems, the conceptual basis is quite simple, easy to understand and
applicable with almost any conventional or object-oriented programming language.
A cellular automaton consists of a large number of cells, which are connected to
a grid and can change their state individually. For all cells, a neighbourhood is
defined that constitutes the surrounding area that influences the state transitions of
each particular cell. Finally, there is a set of rules defining how each of the potential
states of a cell and the states of the neighbourhood will determine the transition
between cell states.
The Cells
Cellular automata models use cells as the units of operation. Cells can be consid-
ered as a storage space, with a defined number of state variables that can either be
discrete or continuous. The most simplistic CAs consist of cells that can switch
between two different states (binary), to be represented e.g. by black and white, on
and off, dead or alive, etc. But it is also possible to have a cell's state being
characterized by a larger number of variables. For example, when modelling soil
processes using a CA, the cell could represent a square meter of the ground and
have storage space for variables such as water content, organic material, tempera-
ture, etc.
The Grid
In a CA, each cell is surrounded by other neighbouring cells. The grid can be
visualized by drawing the cells as nodes and the connection to adjacent cells as
edges. A grid can be finite or infinite, (for simulation purposes only finite) and can
have different topologies. For instance, cells along a line with one neighbour to the
right and one to the left would represent a one-dimensional grid, cells with four
neighbours (North, South, East and West) would represent a two-dimensional grid,
etc. In principle, any topological structure would be possible (Fig.
8.1
).
