Progress and prospects for predicting wildfire spread through the wildland-urban interface

Wildfires that cross the wildland urban interface (WUI) are a major cause of property loss around the world. Recent disasters are likely exacerbated by climate change and continued expansion of urban areas. There is urgent need for better models WUI fires to improve prediction of major events, understand how fires propagate through the built environment, enabling evaluation of the vulnerability of individual buildings and the testing mitigation options. While there are a wealth of models for fire spread in purely wildland and urban settings, little progress has been made in modelling WUI fires. Scientific understanding of wildfire spread into urban areas is challenging to develop because WUI fires burn a complex mix of fuels derived from native vegetation, gardens, and the built environment. Here, we briefly review the state of knowledge around modelling fire spread in wildland and urban settings and then discuss recent research and options for modelling in the WUI, drawing on published analyses of wildfires have impacted built communities. Based on this review we propose an empirical approach to understand how wildfires can ‘propagate’ into the built environment by estimating the probability of a building being destroyed by WUI fires that would as a function of the characteristics of the wildfire as it enters the built environment, the fuel arrangement in nearby buildings and the garden and the construction details of the buildings. We suggest that our proposed model is feasible given available data sets, computationally efficient, and worthing of further research and development. • Wildland fires that cross the Urban Interface are poorly understood, frustrating risk evaluation and mitigation • We review past methods applied to predictive impact modelling for wildland, urban and wildland urban interface fires. • We propose a practical empirical modelling approach that would improve our ability to predict and mitigate these disasters.