A broader perspective on the causes and consequences of eastern Australia's 2019-20 season of mega-fires : a response to Adams et al.

In a response to our Letter on the causes and consequences of the 2019–20 forest fires in eastern Australia (Nolan et al., 2020), Adams, Neumann, and Shadmanroodposhti (2020) argued that fuel loads were causal to the occurrence and size of the fires, along with antecedent dryness. They state that fuel levels were ‘extreme everywhere’, resulting from a lack of fuel reduction via use of prescribed fire. Their proposition is, however, problematic for various reasons. First, they provide no details of the source, methodology and sampling date of their fuel data, nor how it accounted for the wide variations in fire history across the forested landscapes of New South Wales (NSW; Figure 1a), and variations in fuel dynamics among forest types (e.g. Thomas, Watson, Bradstock, Penman, & Price, 2014). Given the strong influence of time since last fire on fine fuel accumulation (Thomas et al., 2014) and the limited scope of sampling, their data do not provide a comprehensive summary of fuels across the fire grounds. Fuel loads, based on the recent fire history and known fuel dynamics, were likely highly variable, with a considerable area recently burnt (Figure 1a), with corresponding inferred low fuel loads. Second, their narrative about ‘extreme’ fuels included coarse woody debris as an implied driver of rate of spread of fires. Fire behaviour models for Australian forests include surface fine and near surface fine fuels (e.g. Cheney, Gould, McCaw, & Anderson, 2012) but not coarse woody debris, as key predictors of spread rate and fire-line intensity. Third, they provide no evidence to support the claim that the fires ‘nearly all burnt at high intensity’ indiscriminately across land tenures. Fire severity (an indicator of intensity) was highly variable in relation to land tenure and time since fire, including records of crown fire in recently burnt areas (Figure 1a,b), with weather a likely key determinant as found in many previous studies (e.g. Collins, Bradstock, & Penman, 2014).