Using extreme value analysis to enhance defendable space for fire fighters and residents

While wildland fires (bushfires) are an anticipated event with each fire season, it is important that the extent of likely annual fire weather conditions and those of more extreme conditions are identified as part of community preparations for fire season. Historically extreme value analysis has been used for floods, storms, temperature, and wind, however, little work has been produced for extreme fire weather. This may be because historically a fire weather index is a composite of differing weather conditions, which, at their extreme, individually may not be related to wildland fire alone. Fire danger index systems generate non-dimensional parameters and in Australia, this has generally focussed on the forest fire danger index or grassland fire danger index. While it is difficult to ascertain individual extremes related to individual parameters for wildland fire, the use of the Generalised Extreme Values distribution is a suitable process for fire danger indices. This can be applied to deterministic fire behaviour assessments whether through the identification of rate of spread, flame length, intensity or suppression effort. This paper offers new insights in the use of extreme value assessments for both MacArthur Forest Fire Danger Index (FFDI) equations so as to determine suitable defendable space for buildings under both existing and planned urban developments in NSW, Australia. Treatment options for existing homes can be more effectively determined and quantified under such an approach so as to better balance resident and fire fighter safety as well as recognising environmental assets. The approach may therefore have broader application for other countries' danger indices when preparing communities and fire fighters for extreme fire weather conditions.