Multicolor particle systems with large threshold and range (Q1185798)

The author studies two classes of interacting particle systems, the Greenberg-Hastings neuron dynamics and cyclic color models. Both cases are evolutions of configurations in \(\mathbb{Z}^ d\), where each site is in one of a finite number of cyclically arranged states. Cyclic transitions occur at each site with constant rate. Depending on the model some transitions are suppressed, unless the number of sites in a neighbourhood, which are in the state a site is intending to take, is bigger than a given threshold. The main subject of this paper is the phase diagram in dependence on the number of states, the threshold and the range of the neighbourhood. The existence of parameter values, for which nontrivial stationary distributions exist, is shown for both cases. The phase diagram of the Greenberg-Hastings model is studied in more detail. The system either dies out or has stationary distributions consisting of ``debris'' or ``fire fronts''. The phase diagram for the cyclic color models is more complicated, since there exists a large number of absorbing configurations. The proofs are based on some results about oriented percolation, which are also proved in this paper.