Bushfire behaviour is well known to be sensitive to the weather. What is less well appreciated is that the behaviour is affected by all scales of atmospheric behaviour, from the large and long (drought), through the medium (weather systems such as fronts and wind changes) to the small (short term fluctuations in humidity and wind). Ongoing improvements in numerical weather prediction have given us the ability to resolve features as small as a few hundred metres, and provide an unprecedented window into atmospheric structures that may affect a bushfire.

This talk will summarise the past two years research by our group into fine-scale modelling of severe fire weather. In that time, we have modelled and verified the meteorology of events including Black Saturday of 2009, the Eyre Peninsula fires of 2005, the Margaret River fires of 2011, and the Blue Mountains fires of 2001. We have identified a variety of small-scale meteorological behaviour that has the potential to significantly escalate a fire, and in one case strongly suspect that the small-scale meteorology was a major contributor to the fire intensity and spread. We will summarise some of these behaviours, and indicate the prospects for predicting them in the future, thereby improving firefighter effectiveness and safety.

This research has been part of the FIRE-DST project funded by the Bushfire CRC. Our modelling has also been a key input into fire modelling, smoke dispersion and risk assessment research under that project. We will indicate a few ways in which these data may be used.