Using Animal Biotelemetry we showed that Gouldian finches preferentially usedareas which had undergone 2–3 years of early dry season prescribed burning.

To understand mainland quokkas’ use of habitat in relation to prescribed burns, and the effectiveness of fire exclusion within quokka range and habitat, we tracked movement patterns of 20 quokkas before and after prescribed burns and examined the important interactions between quokkas and prescribed fire in a flammable and fragmented landscape.

We project future numbers of simultaneous very large wildfires in Western USA based on outputs from climate simulations. In all regions, we project more fires and longer seasons of simultaneity, and years with high simultaneity becoming both more frequent and more extreme.

Burn severity influences post-wildfire hazards such as debris flows, which can occur soon after fires. We modelled burn severity based on pre-fire landscape and fire-weather conditions, and we integrated burn severity predictions with debris-flow models to inform hazard mitigation and planning.

After snowmelt, northern grasslands typically consist of highly flammable grass litter. We analysed a series of experimental grassland fires, as well as Swedish wildfire dispatch data, to isolate the impeding effect of grass phenology on fire propagation. A complete grassfire danger index is presented, likely applicable to most regions with cold (snow-covered) winters.

A Transformer-based model was proposed to identify the forest fire from surveillance camera images. This model proves to be superior in reducing the false alarm rate caused by the interference of different types of smoke, enabling to improve the efficiency of real-time monitoring and early warning of forest fires.