An attribution study of southeast Australian wildfire risk

Abstract

Extreme weather and climate-related events often have a serious impact on our economy, environment and society. This is particularly true in Australia where recurring heat waves, floods, droughts and wildfires have resulted in the loss of life, property and livelihoods. The 2009 'Black Saturday' wildfires in southeast Australia provides a tragic example, having resulted in the death of 173 people and the destruction of over 2000 homes. While there are a number of recorded attribution studies for Australian temperature and precipitation-related events, no such study exists for fire weather.

This thesis presents a new climate modelling system for regional climate simulation and the attribution of weather and climate extremes over Australia and New Zealand. The system, known as weather@home Australia-New Zealand, uses public volunteers' home computers to run a moderate-resolution global atmospheric model with a nested regional model over the Australasian region. By harnessing the aggregated computing power of home computers, weather@home is able to generate an unprecedented number of simulations of possible weather under various climate scenarios. This combination of large ensemble size with high spatial resolution allows a range of extreme events to be examined over Australia and New Zealand with well-constrained estimates of sampling uncertainty. The model is seen to be capable of resolving many climate features that are important for the Australian and New Zealand regions, including the influence of El Nino-Southern Oscillation (ENSO) on driving natural climate variability.

Using the new weather@home modelling framework, this thesis presents the first known attribution study of southeast Australian fire weather. By applying the McArthur Forest Fire Danger Index to large ensembles of regional climate model simulations generated for factual and counterfactual climate scenarios, this thesis reveals that anthropogenic climate change increased the likelihood of elevated wildfire risk over southeast Australia during the 2008-2009 fire season. Furthermore, the influence of anthropogenic climate change on wildfire risk is found to be greater in spring than summer. Through a series of further modelling experiments, this thesis also demonstrates a novel approach for separating the influence of ENSO and anthropogenic climate change within the context of an attribution study. Across southeast Australia, the increase in wildfire risk due to a change in ENSO phase (from La Nina to El Nino conditions) was identified to be much greater than the increase attributed to anthropogenic climate change. This was largely due to the strong increase in drought factor, and decrease in relative humidity, from La Nina to El Nino conditions.