Universal statistics of fire sizes for different targeting schemes in a forest fire model

Abstract

We implement a forest-fire model with temporally periodic, spatially random, and impulsive burning. Even upon varying the dimensions of the grid and introducing external intervention via burning rules biased towards the largest or smallest tree clusters, we recovered stable decaying power-law distributions of fire sizes with scaling exponents near unity. The broad range of parameters where the stable scaling exponents are observed appears to indicate self-organization in the grid, which allows it to be more suited for modeling spatially-extended systems with self-organized (critical) mechanisms.