WHITE BEACH, TASMANIA – As a child, Ian Weir used to set fire to the bush. In his small hometown outside Bremer Bay in Western Australia, dozens would partake in mass clearings by fire in the 1960s.
First, large bulldozers would clear about 300 acres of trees and shrubs and push it all into a massive pile. Once the vegetation dried, residents would line up around the perimeter and light it on fire. With just a slight breeze, the entire landscape would ignite.
The clearing was part of a soldier resettlement scheme where 133 farms were doled out to white military veterans, including Weir's father. The experience speaks to the enduring convention, which still underscores regulations, that in order to live in Australia's fire-prone bush you have to get rid of it.
But after a summer of hellish fires, a debate has sprung up over how to rebuild in the bush in a way that preserves human life and the land. It's one of the "unresolved paradoxes" of bush-fire architecture, Weir said. To resolve it, he wants to emphasize innovative design, not land clearing.
"Landscape combined with bush-fire safety is powerful enough, with the right level of attentiveness, to create a whole new typology of architecture," said Weir, 56. "Something that is just outside people's imaginings."
Fires this summer razed more than 50 million acres, killing at least 34 people and destroying around 6,000 houses and other structures. As climate change pushes fires to extremes, the vast majority of homes in at-risk areas are woefully underprepared for the next inferno.
Generally speaking, bush-fire safety follows a simple rule of thumb: The more land you clear, the less resilient the design needs to be. But clearing land around a house cannot protect against embers, the primary starter of house fires, which can be blown as far as 25 miles, Weir said.
Clearing land puts a house more at risk, he says, because it allows for conventional homes to be built in fire-prone areas. Plus, clearing can cause soil erosion, hurt biodiversity and create dry microclimates.