In a physically based model, interest is in the underlying causes of the hazard event as well as in the manner in which the landscape is affected. Although each hazard is manifested differently and requires different model parameters, the basic framework is the same. As a result, the problem can be divided into two parts:
- The probability of the event or hazard scenario occurring; and
- the propagation of the event through the atmosphere (eg: windstorm), the earth's sub-surface (eg: earthquake) or on the surface (eg: flood).
In the case of an earthquake, the rate at which earthquakes of various magnitudes happen at any location defines the probability of occurrence while a specific event is expressed through the pattern of ground shaking resulting from seismic waves travelling though the earth.
Techniques for modelling various hazards are generally well developed. However, hazard models are only as good as the data used to define them and historical catalogues for earthquakes, floods, fires, landslides, tsunamis, volcanoes, severe storms and tropical cyclones are critically important. Because the historical record in Australia extends only about 150 to 200 years, analysis often requires extrapolating the effects of historical events to potentially catastrophic or probable maximum events. If the underlying hazard process is understood, the knowledge can then be used to define the model parameters which allow extrapolation in a realistic and scientifically credible manner. In addition, the information from hazards which have occurred in a specific location can be used by applying appropriate modifications to estimate the effects when applied to another scenario.
The ability to adapt models to local characteristics is highly dependent on the availability of detailed maps with data on geology, vegetation, elevation and slope and other physical information as well as basic meteorological information such as rainfall and temperature. To determine future trends in hazardous events, models also need to incorporate potential effects of climate change and urbanisation, which may differ substantially from the past. As a result of these factors, its is necessary to ensure the ongoing collection of comprehensive hazard data and access to integrated databases so that the changing requirements of risk model development can be met.