Earthquakes pose a major risk across Asia and recent events have demonstrated their power to take lives, destroy buildings and disrupt societal functioning. In what follows we outline a methodology for building realistic earthquake disaster scenarios for major cities using Adelaide, Australia, as a case study. We analyse an earthquake of magnitude 6.0 occurring on the Para fault, which runs underneath Adelaide’s CBD.
This study is part of a broader research programme from Risk Frontiers and the Bushfire & Natural Hazard Cooperative Research Centre (CRC). It seeks to deliver a suite of scenario simulations that investigate realistic disaster scenarios, scenarios that lie well beyond the emergency managers’ recent experience. They aim to help emergency managers stress test their operations by revealing blind spots and vulnerabilities in their strategic planning.
Despite its low seismic activity, Australia is more vulnerable to earthquakes than one would expect by virtue of its concentration of population and the value of its building stock, much of which is structurally unable to withstand even moderate seismic shaking. The 1989 M5.6 Newcastle earthquake rates as one of the costliest natural disasters in Australia, despite its low magnitude. Its epicentre was close to the city’s CBD.
An important question arising from the Newcastle event is: what would happen if one of Australia’s more populous cities were struck by an earthquake similar to the Newcastle event?
The 1954 M5.6 Adelaide earthquake affords an example of a close call. Until the Newcastle event, this was the most destructive earthquake to have hit Australia. Its epicentre lay far from developed areas at the time, but today, with the growth of the city, would now lie within densely developed areas.
No lives were lost in the 1954 earthquake and there were only three recorded injuries. Many houses were cracked and heavy pieces of masonry fell from parapets and tall buildings in the city. One of the settlement’s earliest buildings, the Victoria Hotel, partially collapsed. Other major buildings that were severely damaged included the local church, St Francis Xavier Cathedral, the Adelaide Post Office clock tower and a newly completed hospital in Blackwood, which sustained major damage to its wards and offices. The Britannia statue in Pirie Street, Adelaide was badly damaged, and since it had also been similarly damaged in the 1897 Beachport and 1902 Warooka earthquakes, the clock in the statue was permanently removed. The Troubridge Island Lighthouse off the southeast corner of Yorke Peninsula shut down after the quake damaged its generator while the Cape St. Auburns Lighthouse on Kangaroo Island began flashing irregularly.
In order to develop a scenario we employ a suite of models that consider risk a function of:
- Hazard: in the case of earthquake, this describes how seismic waves cause ground motions in locations away from the fault as a function of its magnitude and distance from the epicentre
- Vulnerability: how buildings are impacted by seismic ground motions
- Exposure: where buildings are located and whether or not they are occupied at the time of the event.
What is special about this exercise?
- Hazard Modelling
Most earthquake loss models tend to use a single-variable such as peak ground acceleration (PGA) to estimate damage. PGA corresponds to the lateral impulsive force exerted by ground motions on a building; however, damage is also a function of how quickly or slowly that force is applied, which is in turn related to the distance from the source and local soil conditions as well as a building’s construction details and its height. For example, low frequency ground motions of 1sec period will affect a 10 storeys building more than shorter buildings which will respond more to high frequency ground motions. This modelling exercise takes all of this into consideration by using a multi-parameter vulnerability model, which corresponds to the full response spectrum of the ground motion – this is the only earthquake loss model of its kind.
- Exposure Modelling
Another important feature of our modelling is high-resolution exposure information using the Geocoded National Address File (G-NAF). For construction categories we used the National Exposure Information System (NEXIS) from Geoscience Australia. This exercise also requires information on soil types, location of essential facilities and population from a number of sources (ABS, Roadnet and Risk Frontiers’ in-house information)
A large number of buildings are expected to sustain severe damage: Figure 1 shows the extent of the damage to residential addresses while Table 1 summarises the Numbers of Home Equivalents Destroyed (Sum of the damage ratios) for the whole area. For example if 10 buildings are each 10 % destroyed, this is equivalent to one Home Equivalent destroyed. It is assumed that a building is located at each address.
An earthquake can occur at any time of the day and this is important when estimating casualties. For this exercise we have modelled an event that occurred at night when most people are at home and another during a regular working day. Maps below show the distribution of casualties under each of these scenarios. The damage state probabilities for the building stock are used in conjunction with the estimated population distribution to calculate casualties. Figure 2 shows the median spatial distribution of severe injuries and deaths.
We have also modelled damage and disruption to infrastructure and essential facilities – hospitals, schools and emergency services facilities.
The analysis portrays a potentially devastating scenario for the population and the economy of Adelaide. It serves as a tool for emergency managers to examine a plausible event that is well beyond recent experience and help them plan for such an occasion.
For more information, go to www.bnhcrc.com.au/research/economics-policy-and-decision-making/234