Initial growth of fires in eucalypt litter, from ignition to quasi-steady-state rate of spread: A.G. McArthur’s field experiments

Background There are currently no operational fire spread models intended for predicting nascent fire spread such as of point ignitions or spotfires in Australia. Aims To determine if previous findings from point ignition experiments in dry eucalypt litter in a combustion wind tunnel are similar to those in the field. Methods A dataset of historical experiments conducted by A.G. McArthur in dry eucalypt forest, consisting of 1982 fire spread observations, was fitted to two functions for time to quasi-steady-state spread and analysed using statistical methods applied previously to the wind tunnel data. Key results For higher winds, predicted quasi-steady-state times were similar for both datasets. In lower winds, however, times for the field data were 2–3 times longer than the wind tunnel data. Conclusions The Van Wagner function was the most robust, predicting growth times as short as 7 min under high wind/low moisture conditions and as long as 42 min under low wind/high moisture. In drier fuels, lower fuel load doubled the growth time but had no effect in moister fuels. Implications Being able to predict the growth times of new ignitions in forests, including spotfires, will aid prescribed burn ignition planning and wildfire initial attack response.