Using past fire records, we identified planned and unplanned fire occurrences between 2004 and 2015 in south-east Queensland. Unplanned fires covered a larger extent, were more frequent and occurred mostly in spring, whereas planned fires occurred mostly in winter. The use of accurate fire records in mapping enhances fire management capabilities.

Australia’s sparsely settled rangelands cover 78% of the nation, where contemporary fire regimes range from being infrequently burnt in the arid centre following significant rainfall events, to being frequently burnt in northern savannas following annually reliable monsoonal rains. We show that limited prescribed fire management is undertaken today, and, with reference to the Northern Territory, illustrate the immense demographic, institutional, logistical, and economic challenges involved. Proposed solutions include supporting and contracting the developing network of remote Indigenous Ranger Groups to deliver front-line preparedness and response services, building on and taking advantage of commercial savanna burning projects, and nurturing cross-sectoral and -institutional arrangements generally.

Historically unmanaged fire regimes are implicated in the detected decline of savanna biodiversity. Analysis of a contemporary primarily carbon market funded indigenous fire management program shows that most defined ecological thresholds are being addressed. While there is room for improvement, and work is required to more appropriately define and monitor ecological fire regime thresholds, these results demonstrate what is possible through well-resourced co-operative fire management.

We present a statistical method to quantify and map wildfire spread to assets and the reduction gained by removing fuel from treatment blocks based on fire spread patterns in historical fires. The likelihood of spread to assets was highest in areas with high forest cover, but spread reduction was highest in treatment blocks at the wildland–urban interface.

We use fire behaviour simulations and Bayesian networks to estimate the risk mitigation effects of prescribed burning for area burnt, house loss, life loss, length of powerline and road damaged, and area burnt below minimum tolerable fire interval. Our methods can be used to quantify and compare risk across regions.

Fire managers need to make decisions about expenditure in a more transparent manner. Here we undertake an economic analysis of fire management in the Australian Capital Territory to identify cost-effective management approaches. Absence of fuel treatments in this landscape will increase the risk of loss in the long term.