Degree Name

Doctor of Philosophy


School of Earth, Atmospheric and Life Sciences


Wildfires burn in a variety of landscapes across the world and have important consequences for human populations and the environment. It is critical to understand the behaviour of wildfires in order to develop and implement effective wildfire response and management strategies.

Spotting from wildfires and wildfire rate of spread are two important aspects of fire behaviour. Spotting is a process by which small separate fires are ignited by small burning pieces of vegetation generated from a wildfire and transported by ambient wind or a smoke plume into unburnt fuels. Spot fires can significantly reduce the chances of successful fire suppression by firefighters and potentially lead to faster and more erratic fire spread. A wildfire’s rate of spread refers to the rate at which flames move through fuel. Firefighters must understand a wildfire’s rate of spread, and predict how fast the wildfire may burn, in order to make effective decisions about fire crew placements, firefighting tactics and issuance of community warnings and evacuation orders.

The overarching aim of this thesis is to describe and model specific aspects of fire behaviour, namely wildfire spotting and rate of spread, using novel methods. In this thesis, empirical data from real wildfires captured in southern Australia are used for three research chapters, along with a supplemental experimental study for one research chapter. The four research chapters explore (1) regional and within-fire variation in spotting patterns; (2) environmental drivers of long-distance spotting during wildfires; (3) the development of a Bayesian probabilistic approach to wildfire rate of spread modelling, based on wildfire observations, and (4) the interactive effect of spot fires and topography on fire rate of spread in laboratory experiments.



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.