Australia experiences 50,000 bushfires annually; bushfires have been integral to the Australian environmental ecosystem. However, in recent decades, the damages caused by unplanned bushfire to lives and livelihood have been burgeoning along with associated hazard events. Current annual damage estimates are above $1.5 billion AUD. Some well-known facts are cases in point – the most recent major black summer 2019-2020 bushfire led to a destruction of more than 18 million hectares destroying homes and affecting billions of animals, in its wake, while the 2009 bushfires killed 173 people. Beyond the obvious and acute impacts on lives and economy, the effects of increases in frequency and intensity of bushfires are often far reaching, leading to loss of vegetation, soil erosion and soil quality degradation (impacting agricultural productivity), and green-house gas emission. A recent study has estimated that 25% of Australian greenhouse emissions are attributable to bushfires.
Quantitative risk profiles of bushfires are beginning to emerge regarding local factors that influence emergence, propagation, and periodicity of unplanned bushfires. These factors can broadly be grouped into two classes
- Anthropogenic
- human driven aspects that include systematic factors of increases in land clearing for housing, related urban development accompanying more sporadic factors such as arson (reference). The Australian Institute for Criminology estimates that up to 50% of all unplanned bushfires could be related to human activities. The 2009 bushfire were linked to arson.
- But also, interventions in the form of cool burnings to reduce fuel load. There are multiple approaches to these (reference).
- Environmental
- These are more complex interplays of climate drivers – trade wind patterns (EL Nino, La Nina) that lead to flood and drought cycles, on a susceptible landscape under prevailing vegetation (e.g – grass versus eucalyptus), soil condition (bulk density and such), humidity, and lightning density (ignition). These factors are further affected by an evolving climate regime, including global warming.
The intertwining of these factors makes it difficult to separate natural patterns from definitive changes in regime, based on evidence. Such understanding is critical for two reasons – a) to better explain the uncertainty of fire frequency and fire intensity, in presence of exogenous
Aim
The intertwining of these factors makes it difficult to separate natural patterns from definitive changes in regime, based on evidence. Such understanding is critical for two reasons:
- to better explain the uncertainty of fire frequency and fire intensity, in presence of exogenous conditions,
- further such phenomenological understanding would help us project these events into future helping policy.Using 100 years of bushfire data, this HDR project would apply contemporary methods in statistical science and develop novel spatio-temporal uncertainty models to uncover some of these challenges.
Thus, a key objective of this research is to use observations (s_i,t_j) of past bushfire records to construct the probability density for clusters of spatial and temporal risk for unplanned bushfire events, driven by an intensity function λ(s,t). A general representation of such densities is given by:
- f(s,t)∝{█(∑i▒∑_j▒〖log{λ(s_i,t_j )}〗-∫({s∈A,t∈T})▒〖λ(s,t) ds dt〗@λ(s,t)∝g{X(s,t);θ} )┤
Objectives
The project would seek to combine best practices in computational statistics and signal processing by incorporating soil and related environmental analytic methods. There is also scope to investigate predictive analytics and intervention methods. The project would create a bespoke repository, using probabilistic data-fusion between multiple correlated data sources. These include bushfire extent data of Geoscience Australia, soil and vegetation index data obtained from the Commonwealth Scientific and Industrial Research Organization (CSIRO), and weather information obtained from the Bureau of Meteorology (BoM) and World Climate research program.
Significance
The impact of a changing climate is altering the intensity and frequency of all natural hazards. This is beginning to have serious impacts on lived experiences-including all facets of economy, life on land, and all ecosystems at large. Driven by an heating climate unplanned bushfires are a significant Australian natural hazard whose impacts on urban and rural life have been changing rapidly. This project that will sit at the interface of statistical science and signal processing, will quantify and revise our knowledge of uncertainty of these events to establish best practices for disaster preparedness.
Ideal Candidate
Content knowledge, one or more of the following:
a. Likelihood principles in statistics
b. Bayesian computation
c. Stochastic processes/time series/Markov chains.
Desirable: Spatial statistics, spectral analysis, MCMC and computational statistics methods.
Proficient in one or more of the following software – R, Python, Matlab.Additionally, the applicants should meet the eligibility criteria for entry into a PhD program at Curtin University.
Additionally, the applicants should meet the eligibility criteria for entry into a PhD program at Curtin University.
This project is open to Domestic applicants only.
Scholarship
If you are identified as the preferred candidate for this project, you may be considered for an RTP scholarship.
Enquires and How to Apply
For enquires about this opportunity contact Dr Sourav Das at Sourav.Das@curtin.edu.au
To formally apply submit an Expression of Interest to Dr Sourav Das during the Central Scholarship round (July 1st – July 31st 2026)