Koondoola Regional Bushland (KRB) is a 137-hectare remnant bushland site of the Swan Coastal Plain in Western Australia, particularly important for social, ecological economic and cultural reasons.

KRB is roughly quadrangular with maximum extensions N-S and W-E of ≈1.7 Km and 1.2 Km, respectively. The reserve has a network of walking trails, of various width, linearity, evenness, clearance, ground firmness etc. Some of these trails may be used by relatively small vehicles for emergency response. KRB is surrounded by a densely populated area, 15 km from the Perth CBD. Residential areas border the western and northern sides, with industrial/commercial precincts found to the South-West. The site also features a community centre, and a Water Pump Station.

Regarding bushfires, the proximity of a large and densely vegetated site to highly urbanised mixed-use areas, requires a proper management strategy, including risk assessment, prevention and reaction plans, also under different scenarios. The present study was motivated by such requirement, and its aims are to assess the current bushfire risk level, anticipate changes under different scenarios and provide management recommendations to the City of Wanneroo and relevant stakeholders.

For baseline assessment, field data on fuel load, vegetation condition, topography, human interference was collected, in September 2019 (late winter), from four zones and n=89 sites. Perth summer and autumn wind roses (n ≈ 7,000 observations) were obtained from the Australian Government Bureau of Meteorology (Bureau of Meteorology, 2019), adopting and an arbitrary 20% frequency threshold. Additional measurements were obtained in ArcGIS v.10.5.1, which was also used for Kriging-interpolations of baseline overall fuel hazard, and other geospatial analyses.

Four scenarios were defined by the combination of the levels of each of two drivers: population density and season. The choice of drivers was motivated by their expected large impacts on bushfire -risk and -management (i.e., prevention/response), and their high temporal variability.

The model also assumes that higher population density will increase bushfire risk (i.e., positive linear correlation), mainly due to a higher:
• prevalence of arsonists using concealed internal tracks
• pressure on water supplies suitable for firefighting
• fuel load as anthropogenic litter
• fuel load as anthropogenically introduced weeds
• fuel load due to higher incidence of anthropogenically-spread dieback infections
• fuel load from dead vegetal biomass due to human disturbance
• accidental ignition from cigarette butts, sparks from road traffic and machinery, home fires from barbecues, appliances etc.