Images of capital in 2003 offer first proof of fire tornado
In a helicopter over fire-ravaged Canberra in January 2003, Alan Walker and Stephen Wilkes photographed the black ruins of homes and trees below, unaware their images would create the world's first confirmed case of a fire tornado.
In the worst flying conditions imaginable and in poor light at sunset, they had limited opportunities before the pilot was ordered to return to base.
Video and still cameras had captured the fire's ferocity earlier, including some images from a man on the southern end of Kambah playing fields who tracked the tornado from kilometres away.
Researcher Rick McRae said bushfire science leapt forward that day, thanks to cameras on the ground and in the air, a newly commissioned Captains Flat weather radar and satellite imagery.
On the other side of Mount Coree on the north-west border of the ACT, the tornado formed in a plume of the McIntyres Hut Fire about mid-afternoon. It crossed the Brindabella Ranges, then moved through Uriarra and Pierces Creek pine plantations and grazed the edge of Chapman.
It raged for about an hour before finally fading on the northern edge of Kambah near Mount Taylor.
Mr Walker and Mr Wilkes, both researchers for the NSW Rural Fire Service on the day of the fires, later joined Mr McRae, a special risks analyst at the ACT Emergency Services Agency, in piecing together data. When combined and analysed it revealed a phenomenon known as pyro-tornadogenesis, the technical term for a large fire which produces a genuine tornado.
''Researchers had speculated about the ability of a fire to produce a tornado, but this is the first documentation of the creation of a true tornado by the convection column of a large fire,'' Mr McRae, the lead researcher, said. The two-year project also involved Dr Jason Sharples from the University of NSW, Canberra, also a volunteer firefighter.
''Our analysis indicated that the tornado had a rating of at least a 2, on the Enhanced Fujita scale of tornado severity. It had major effects on the behaviour of the fire on the urban edge and had enough force to remove roofs from houses and to blow cars off the road,'' Mr McRae said. ''It moved at over 30km/h across the ground and had a basal diameter of nearly half a kilometre when it reached Chapman. It was a major tornado, but was barely noticed given the setting.''
A fire tornado is fundamentally different from a fire whirl, which is commonly associated with fires, said Dr Sharples. ''Tornadoes are associated with thunderstorms and as such they are anchored to a thundercloud above, and are able to sporadically lift off the ground. Fire whirls, on the other hand, are anchored to the ground and do not require the presence of a thunderstorm.''