Research findings

The HighFire Risk Project

Previous

PUBLICATIONS

Next
Citation Sharples, J.J., Viegas, D.X., McRae,R.H.D., Raposo, J.R.N. & Farinha, H.A.S. (2011). Lateral bushfire propagation driven by the interaction of wind, terrain and fire. 19th International Congress on Modelling and Simulation, Perth.
Title

Lateral bushfire propagation driven by the interaction of wind, terrain and fire.

Author(s) J.J. Sharples, D.X. Viegas, R.H.D. McRae, J.R.N. Raposo and H.A.S. Farinha.
Abstract The interaction of rugged terrain and extreme fire weather can have a variety of influences on bushfire propagation. These influences can sometimes result in highly atypical modes of fire propagation that can cause rapid escalation of bushfires to their most catastrophic state, thereby subjecting communities and assets to the highest levels of risk. In one such example, referred to as fire channelling, the interaction results in rapid lateral development of a fire (in a direction transverse to the synoptic winds) in addition to enhanced downwind fire development and the production of expansive flaming zones, which may occur as a precursor to the formation of pyrocumulus, or even pyrocumulonimbi. In this paper, we give an overview of the fire channelling phenomenon and report on some of the recent modelling and simulation efforts directed at providing a more formal understanding of it.
The fire channelling phenomenon was first noted in connection with the extreme bushfires that burnt into Canberra on the afternoon of 18 January 2003. In this instance, fires were observed spreading laterally at a number of locations within the rugged terrain in the Brindabella Ranges to the west of Canberra. Subsequent observations of the 2003 Thredbo fires revealed additional examples. Moreover, in the time since studying the 2003 fires a number of other likely instances of fire channelling have been noted in the Blue Mountains, California, Sardinia and Portugal.
The fire channelling phenomenon has been attributed to the (hypothetical) interaction of a horizontal vortex that forms due to separation of the ambient winds over a sufficiently steep lee slope with an active lee slope fire. The interaction causes the fire to spread laterally within the vortex as a turbulent finger of flame. The increased turbulence enhances ember production and some of these embers are incorporated into the synoptic flow above the vortex and deposited downwind. Over all the process results in a type of ‘bidirectional’ fire propagation. This hypothetical mechanism has been tested and supported by a series of combustion tunnel experiments which are reported briefly in the paper.
The existence of the atypical lateral fire propagation across a lee slope is now firmly established and so our concerns in this paper turn to the question of how best to model the effect. We report on a number of preliminary modelling efforts using a two-dimensional numerical model that describes the interaction of the horizontal vortex with a surface heat source through coupled fluid and heat conduction equations. The model results are by no means authoritative, however, and we discuss some possible ways to progress the modelling of the fire channelling phenomenon.
Keywords Dynamic fire spread, wind-terrain interaction, atypical fire spread, bushfire, topography.
Access http://www.mssanz.org.au/modsim11/a2/sharples.pdf