Scaling of the Cumulative Weights of the Invasion Percolation Cluster on a Branching Process Tree

Abstract

We analyse the scaling of the weights added by invasion percolation on a branching process tree. This process is a paradigm model of self-organised criticality, where criticality is approach without a prespecified parameter. In this paper, we are interested in the invasion percolation cluster (IPC), obtained by performing invasion percolation for $n$ steps and letting $n\to\infty$. The volume scaling of the IPC was discussed in detail in (G\"undlach and van der Hofstad 2023) and in this work, we extend this analysis to the scaling of the cumulative weights of the IPC. We assume a power-law offspring distribution on the branching process tree with exponent $\alpha$. In the regimes $\alpha>2$ and $\alpha\in(1,2)$, we observe a natural law-of-large-numbers result, where the cumulative weights have the same scaling as the volume, but converge to a different limit. In the case $\alpha<1$, where the weights added by invasion percolation vanish, the scaling regimes change significantly. For $\alpha\in(1/2,1)$, the weights scale exponentially but with a different parameter than the volume scaling, while for $\alpha\in(0,1/2)$ it turns out that the weights are summable without any scaling. Such a phase transition at $\alpha=1/2$ of the cumulative weights is novel and unexpected as there is no significant change in the neighbourhood scaling of the IPC at $\alpha=1/2$.