We explore the basis of understanding wildland fire behaviour with the intention of stimulating curiosity and promoting fundamental investigations of fire spread problems that persist even in the presence of tremendous modelling advances. Internationally, many fire models have been developed based on a variety of assumptions and expressions for the fundamental heat transfer and combustion processes. The diversity of these assumptions raises the question as to whether the absence of a sound and coherent fire spread theory is partly responsible. We explore the thesis that, without a common understanding of what processes occur and how they occur, model reliability cannot be confirmed. A theory is defined as a collection of logically connected hypotheses that provide a coherent explanation of some aspect of reality. Models implement theory for a particular purpose, including hypotheses of phenomena and practical uses, such as prediction. We emphasise the need for theory and demonstrate the difference between theory and modelling. Increasingly sophisticated fire management requires modelling capabilities well beyond the fundamental basis of current models. These capabilities can only be met with fundamental fire behaviour research. Furthermore, possibilities as well as limitations for modelling may not be known or knowable without first having the theory.

An understanding of what variables affect the ignition of live wildland fuels is crucial to predicting crown fire spread, the most poorly understood type of wildland fire. Ignition tests were performed over the course of an entire year for ten species (three species in year one, seven in year two) to evaluate seasonal changes in flammability. Ignition delay and mass loss rate at ignition were measured for a radiant heat flux of 50 kW/m² and flow velocity of 1 m/s. Large species-to-species variation occurred in two-variable linear correlations between moisture content, ignition time, and mass loss rate at ignition. Only a few species showed the same behavior as wet wood. Due to potential physical and chemical changes that live fuels undergo during the growing season, moisture content is not a particularly useful descriptor of ignition behavior for live fuels.