A Burning Desire to Understand Complexity

When physicists talk about chaos, they generally mean that the outcome of a sequence of events depends sensitively on the starting conditions. The common example is the flap of a butterfly's wings in Brazil creating a hurricane in Texas. Many real-life systems are to some extent chaotic. At first glance, modelling and predicting such systems might seem hopeless because no one could ever measure every tiny detail of the starting conditions. However physicists dealing with many-body systems have long realised that a great deal of progress can be made by throwing away most of the detailed microscopic information and describing the system in a coarse-grained, statistical sense – For example, in modelling the behaviour of the atmosphere, rather than focussing on individual molecules it is more productive to divide the system into a finite number of cells or grains each containing many individual molecules. If the size of these grains is chosen correctly the chaos diminishes, or disappears altogether and predictable patterns emerge that are insensitive to the minute details of the starting conditions. This approach has led to the development a new kind of science, Complex Systems Science, which has spread far beyond physics.

Although the study of complex systems in thermal equilibrium has reached a relatively mature state, modelling of open, driven, non-equilibrium complex systems is still a frontier field of physics. One example of such a system of particular relevance to Australia is large bushfires. A bushfire is an open system because it is a localised event in the wider atmosphere. It also has a fixed beginning, limited duration and the fire generates heat which in turn, drives the system out of thermal equilibrium.

Physicists at ANU are adopting a two-pronged approach to modelling such fires. Firstly by developing a better understanding of the physics and chemistry of cellulose combustion they hope to describe the heat production mechanism more accurately. Secondly, they are designing new continuum models that are more appropriate for describing fluid systems such as bushfires than traditional many-body statistical models.

The ANU team are optimistic that progress can be made in understanding fires because their behaviour is typical of a complex system: in a well-established fire patterns emerge which can be relatively insensitive to changes in external factors such as fuel load. This can lead to dangerous and counter-intuitive behaviour such as an ability to cross firebreaks by the coalescence of small spot fires on the other side. It's hoped that with a better understanding of the emergent properties of bushfires, transitions to life-threatening situations might be better identified and avoided.

Research Highlight Poster

 
 

Back Home