During the initial 7-years of the Bushfire CRC, a considerable amount of work was undertaken by CSIRO Ecosystem Science in quantifying the physical attributes of buildings and their surroundings in a way to help predict probable damage or loss in a bushfire. At the same time, the University of Melbourne developed a fire characterization model, PHOENIX RapidFire, as part of a risk assessment process. PHOENIX used the nature of fuels across the landscape, the topography and weather conditions to characterize fire characteristics such as flame height, fireline intensity, fire size, fire rate of spread, and ember production. These fire characteristics are known to be associated with house and life loss
The body of work involves the characterisation of non-vegetative elements found within the interface zone (vegetative elements and ground fuels). The research focus provides consistent parameters that adequately describe the element properties and spatial relationship hence their behaviour during the vulnerability assessment process.
The vulnerability assessment framework requires the definition of elements in two fundamental property categories:
- physical response to the effects of fire,
- potential to and severity of fire effects if ignited.
Review of data from the Black Saturday Bushfires will support these parameters and provide guidance as to appropriate default values relevant to that region. Investigations into other case study areas will also occur to establish appropriate regionally specific urban element assumptions. These will be influenced by local building and landscape practices as well as local policies.
The project will address the is a need to combine the broader fuel definition used in bushfire behavior prediction with the very detailed fuel definition used in house damage and loss research to come up with a system of defining fuels in semi-urban areas that useful for contributing to the prediction of house loss from bushfire at the scale in which Phoenix RapidFire is typically run.
This body of work will focus on the use of time based simulations of actual case study scenarios created by the project. The assessments will simulate a wide range of input assumptions such as wind directions and fire weather intensity, as well as a number of ‘what if’ scenarios regarding modified building and vegetation layouts. Hence the case study vulnerability assessment becomes a working assessment tool for future policy options consideration.
The project work is broken up into a series of process:
- Refinement of the model for application in the specific case study area, eg for Black Saturday fire we draw on the post bushfire survey dataset to refine parameters and assumptions in the model. These assumptions will be developed in close consultation with stakeholders with regard to methods, and scale of scenario definitions. An example would be to describe the prevalence of wind as a damaging element to buildings in the Black Saturday fires and to decide whether this effect is incorporated into the vulnerability assessment model.
- Creation of case study scenarios. This involves to development of 3D scenes using spatial data previously captured of the relevant case study areas and the range of variation in the case studies that represent the ‘what if’ alternatives. The ‘what if’ scenarios will also be developed in close consultation with stakeholders.
- Initialisation of the scenes ready for time base analysis. This initialisation process creates a database of the 3D relationships each element has with the others in its surroundings. The output of this process can be used to understand radiation exposure implication that various elements present to critical elements such as occupied dwellings.
- Time based analysis of case studies. This process involves the actual time based simulation of advancing fire fronts whose characteristics are determined in other parts of the project (such as the use of Phoenix Rapidfire etc). The time based analyses will run at much faster than real time, allowing many runs to be performed statistically aggregated results. For demonstration purposes the model can be slowed down to demonstrate how a simulation/scene develops.