This paper examines household preparedness and responses to the 2009 ‘Black Saturday’ bushfires in Victoria, Australia. Results from a mail survey indicate that although leaving early is the safest response to bushfires, staying to defend can be a viable alternative to evacuation for some people.

The relationship between wildfire risk perceptions and mitigating behaviours of wildland–urban interface residents is examined. Exposure to a wildfire program website, having a wood roof, previous experience living in a fire-prone area, age and income are positively related to wildfire risk-mitigating behaviours.

This article describes an analysis of two studies that look at trust between community members and local land management agencies in the context of fire management in Australia and the United States. Findings suggest common characteristics exist and can be focussed on to build trust in both countries.

We determined the soil heat pulse beneath operational burn piles ranging in fuel composition and pile size. Piles dominated by large wood produced extreme temperatures and heat durations of concern to post-fire soil properties. In comparison, the heat pulse beneath slash piles containing a mixture of fuel sizes was moderate.

We tested the idea that native annual plant species would be better for seeding onto post-fire habitats in the western USA compared with commonly planted perennial species. Our results suggest that native annual plant species have potential for use in restoring post-fire habitats but longer-term data are needed to determine future outcomes in plant community composition.

Post-fire white ash cover correlated significantly to surface fuel consumption at eight prescribed fires across four fuelbed types. White ash cover measured immediately post-fire may provide a physically based, retrospective measure of fuel consumption and fire severity with utility for fire managers, ecologists and modellers of fire behaviour and fire effects.

Approaches for reporting on forest fire-related greenhouse gas emissions in southern Europe were experimentally applied to estimate burnt biomass in forest fire events occurring in Italy during 2008–2010. A reliable range of biomass loss can be derived by methods able to assess variability of local fire effects across different forest types.

The Weather and Research Forecasting model was evaluated for its ability to simulate fire weather in south-east Australia between 1985 and 2009. The model showed considerable skill and is a valuable tool for downscaling climate projections from global climate models. The causes of model error and the effect of model grid spacing were also examined.

Post-windstorm salvage logging and a combination of salvage logging and piling and burning of slash significantly reduced average fuelbed depths, fine woody fuels, fuel consumption and pollutant emissions. We did not find clear evidence treatments reduced potential fire behaviour parameters such as rate of spread, reaction intensity and flame length.

This paper presents estimates of potential future wildfire suppression cost savings that result from allowing a current wildfire to burn on a landscape in central Oregon. Under some conditions, estimated savings were large, suggesting that the benefit of allowing a wildfire to burn may, in select cases, outweigh the additional risk of loss.

We found an overall positive perception of safety climate among members of US federal wildland fire agencies with recent incident experience, with the degree of connection to other positions and units having the strongest influence. Findings indicating that mid-level supervisors are the least connected have implications for high performance and safety.

The equilibrium moisture content (EMC), a key parameter for fuel moisture modelling, was determined for litter from four major tree species in central Europe using three different experiments. Differences between species, sorption directions and experiments were examined and a comparison to literature values was made. Test procedures are recommended for future experiments.

We assessed wildfire effects on a differently thinned young Douglas-fir afforested area of Patagonia, Argentina. Thinning varied from Reineke’s Stand Density Index (SDI) of 900, 700, 500 and unthinned control. Tree mortality decreased as thinning increased, and no trees survived in the control. Plots thinned to SDI 500 or 700 appear to better withstand severe fires and also achieve the highest growing rates.

We estimated the total carbon release emitted from forest fires in Japan over a 30-year period. While carbon emission varied widely from year to year based on the area burnt, it decreased dramatically from the 1980s onward and was substantially lower than the mean annual net primary production of Japanese forests and carbon releases in other countries and regions.

We developed fuel load and fire risk models using measurements of fuel loadings and remote sensing data. We modelled the extensive fires that occurred in the Mojave Desert in 2005, and validated our risk model at a secondary site. This fire risk prediction method could extend to other arid systems.

Formulation of a physics-based model capable of predicting the spread of laboratory-scale fires is described. The model is used to investigate the importance of spatial distribution of bulk density, and of moisture content on the fire spread within a shrub.

The WRF-Fire coupled atmosphere–fire model is used to investigate atypical wildland fire spread on steep leeward slopes. Intermittent rapid lateral fire spread is seen close to the mountain ridge line, owing to an interaction between strong pyro-convection and terrain-modified winds.

The El Niño–Southern Oscillation was the primary driver of historical drought and upper elevation fire occurrence, including high-severity fire, in the mixed conifer–aspen, and spruce–fir forests of the south-western US. This relationship is potentially modulated by the Atlantic and north Pacific Oceans, which could provide longer-term lead forecasting.

We present a technique for estimating conditional probability of burn patterns for large fires. Patterns quantify the conditional probability that points near an ignition will burn and illustrate the probabilistic shape and extent of a fire. Predictive ability is good and estimated coefficients are consistent with expectations.

Traditional remote sensing approaches to assess fire severity provide information for post-fire management; however, they do not provide a useful approach to adjust combustion completeness (CC) and carbon emissions. We show a remotely sensed approach that provides a measure of fire severity that can be used to refine carbon emissions.

Salvage logging is a widely applied post-fire management that implies the removal of nutrients in the remaining wood. We assess the value of the partially charred wood as a nutrient reservoir for the ecosystem reserves. Wood contributions were particularly relevant for N and micronutrients compared to the soil nutrient stocks.

This study investigates the selectivity of fire in relation to land cover type and topography in Southern Europe. Shrubland and grassland are more fire prone, whereas agricultural areas and artificial surfaces are less susceptible to burn. There are significant differences between countries and regions. Slopes >25% and north facing ones were less fire prone.

We quantified the effects of post-fire logging on forest fuel profiles as influenced by harvest intensity (percentage basal area cut) and biophysical setting (mesic to drier forests). Post-harvest fuel loads varied significantly across each of these gradients, and suggested a broadly applicable management tool relating post-fire logging intensity (basal area cut) to subsequent surface fuel loads.

Effects of two post-fire stabilisation techniques (grass seeding and straw mulching) on chemical properties of eroded sediments, and amount of nutrients lost with them, were evaluated. The highest concentrations of extractable base cations in the sediments occurred during the first 3 months after fire. Post-fire mulching application can be effective for reducing nutrient loss with sediments.

This paper reports the results of laboratory experiments to test the effects on cation leaching after thermal shocks and rainfall simulations on unaltered soil samples. The amount of cations leached increased with the severity of the thermal shock; however, under moderate conditions, there was a decrease in cation leaching, mainly of monovalent ions.

We modelled the distribution of wildfire ignitions across three eastern states (New Jersey, Ohio and Pennsylvania) for a 12-month period using maximum entropy. Models were parameterised with four integrated landscape variables and observed wildfires between 2000 and 2009 were used to predict the probability of fire occurrence. Monthly models performed better than a 12-month average of ignition probabilities.

Firebrand characteristics of untethered samples of the bark of messmate stringybark (Eucalyptus obliqua) were measured in the modified airflow of a vertical wind tunnel. The flameout and burnout times of the samples, together with the quantities of this bark available in a forest, confirm the species capability for intense spotting.

Anthropogenic burning has become a common phenomenon in pine-dominated forest ecosystems in Thailand. We investigated fuel loads and fire behaviour in a degraded pine forest and a pine–oak forest at Nam Nao National Park. Based on this study, a pine forest here requires more than 1 year of fire-free period to recover back to the pre-burn conditions.

Using current cost data for the Sierra Nevada area of California, this paper adds to mounting evidence that increases in housing lead to increases in fire suppression costs.

In Mediterranean gorse shrublands, total woody fuel diminished with repeated fires. In addition, there was no positive selection for elevated dead fuel since a second fire encouraged a community with less dead fuel. In contrast, herbaceous species were enhanced by repeated fires.

Through small-scale burning experiments, we determined the amplifying effect of fuel loading on fire behaviour and thermocouple probe temperatures at the surface of fuelbeds dominated by masticated saw palmetto and gallberry shrubs. Soil temperatures were also influenced by fuelbed properties; however, temperatures never reached 60°C as shallow as 2 cm beneath the masticated fuels.

Based on remote sensing data for the period 1990–2008, median fire intervals in tropical savannas in western Arnhem Land, Australia, ranged from 1 to 3 years. Closed forests experienced repeated fires within 4 years. Hazard functions were non-monotonic with peak values at 2 to 3 years. A discrete lognormal model is more appropriate than a Weibull model for fire frequency analysis in these tropical savannas.

This study reports the relationships between forest fire activity and meteorological parameters in the eastern Euro-Mediterranean region through the analysis of long-term meteorological and fire data spanning more than a century. Outputs underline the effect of precipitation in fire spread by controlling both fuel production and fuel moisture.

A review of the literature suggests at best a minor effect of live foliar moisture on the spread rate of crowning fires in conifer forests and shrublands. None of the present model functions used to adjust the spread rate for the relative effect of live foliar moisture are considered satisfactory, including application to dead foliage.

Initial attack success models and decision guides have been developed for grassfires in agricultural regions of southern Australia using incident and response data. These can be used by fire managers to make quick assessments of the need for extra resources at early stages of a fire.

The study addresses long-term effects of single and recurrent fires on soil water repellency and organic matter content. Results indicate a significant prolonged decrease in water repellency following recurrent fires, depending on organic matter content. Soil organic matter can serve as an indicator for soil rehabilitation following recurrent fires.

Factors influencing lightning and arson ignitions were examined within the densely populated Sydney region of south-eastern Australia. Patterns observed were consistent with those reported for forested regions in the northern hemisphere. Climate change is likely to increase ignition frequency but population growth resulting in urban expansion is likely to have a greater effect.

Fire behaviour at two sites was moderately enhanced by the desiccation of tamarisk (Tamarix spp.), a flammable invasive tree, in one case from herbivory by the tamarisk leaf beetle (Diorhabda carinulata Desbrocher), and another in which herbivory was simulated with herbicide. Physical characteristics and season had a greater influence on fire behaviour than did foliar desiccation at both sites.

We analysed spatial patterns of wildfire exposure in relation to key social and economic features on Sardinia. Historical data and wildfire simulations were used to estimate burn probabilities, flame length and fire size. Outputs were summarised for the features of interest, and related to spatial variation in fuels, ignitions, topography and weather.

Peavine Métis settlement, an Aboriginal community in Northern Alberta Canada, has a settlement-led wildfire mitigation program that includes both residential and community level activities. Local residents supported this program and engaged in additional residential mitigation activities. Residents’ wildfire experiences were found to encourage support for community level activities.

A cellular automata model integrating surface fire spread, firebrand lift-off and dispersal, as well as spot fire ignition and propagation was constructed to investigate spot fire behaviour. This study uses the model to examine how spotting under various environmental conditions may affect (i) fuelbreak jumping and (ii) the main fire’s rate of spread.

This paper presents a method to generate fire risk maps making extensive use of geographic information technologies. It describes how the variables were generated and integrated, and how the final index was validated using two years of fire occurrence.