The modelling of wildland fire spread across a heterogeneous landscape is significant because fire dynamics are sensitive to local spatial characteristics. The development of accurate fire models and simulations is important due to the economical and social losses wildland fire can cause and the resulting need to better understand, predict, and contain fire spread. We present a methodology for encoding the spread of wildland fire in a set of interacting automata. The Circal formalism is used to explicitly describe the transmission of fire as an interaction between discrete cells of landscape. We demonstrate the potential for the methodology to accurately model spatial dynamics by giving results of our implementation of a fire spread model that includes a heterogenous environment.