Phase transitions for a planar quadratic contact process

TL;DR

This paper investigates phase transitions in a two-dimensional quadratic contact process, establishing bounds on critical reproduction rates for survival and stationary distribution, and linking edge expansion speeds to these critical values.

Contribution

It introduces bounds on critical values for survival and stationary distribution using block constructions and comparisons, and connects edge speeds to phase transition thresholds in the process.

Findings

Edge speeds characterize critical values for survival and stationary distribution.

Positive edge speeds imply reproduction rates above critical thresholds.

Open problem: negative edge speeds may imply system extinction from finite sets.

Abstract

We study a two dimensional version of Neuhauser's long range sexual reproduction model and prove results that give bounds on the critical values $\lambda_f$ for the process to survive from a finite set and $\lambda_e$ for the existence of a nontrivial stationary distribution. Our first result comes from a standard block construction, while the second involves a comparison with the "generic population model" of Bramson and Gray (1991). An interesting new feature of our work is the suggestion that, as in the one dimensional contact process, edge speeds characterize critical values. We are able to prove the following for our quadratic contact process when the range is large but suspect they are true for two dimensional finite range attractive particle systems that are symmetric with respect to reflection in each axis. There is a speed $c(\theta)$ for the expansion of the process in each direction. If $c(\theta) > 0$ in all directions, then $\lambda > \lambda_f$, while if at least one speed is positive, then $\lambda > \lambda_e$. It is a challenging open problem to show that if some speed is negative, then the system dies out from any finite set.