Another Earthquake in the Cadell Fault Scarp

At least five times in the past, the Cadell Fault Scarp has ruptured with an earthquake of Mw1 7.0-7.3.

What if a similar earthquake happened today?

One of the questions we might ask is what would the effect be if an earthquake similar to those in the geologic record happened today? It has happened at least five times in the recent geological past, so the fault is certainly capable of hosting a future event.

What Geoscience Australia studied

Geoscience Australia has been studying the scarp for the last five years. This investigation involved: seismic reflection profiling to determine fault geometry, detailed geomorphological mapping to determine earthquake magnitude and effects on the environment, and dating of regolith units to determine earthquake timing.

About the Cadell Fault Scarp

The Cadell Fault scarp is situated on the Riverine Plain, a low relief region of the Murray-Darling Basin in southern New South Wales. It is the best studied example of a multiple-event Quaternary fault scarp in eastern Australia.

The scarp is almost 80km long and up to 15m high, and reflects the accumulated displacement resulting from at least five earthquakes. These events occurred over roughly a 50ka time interval between 70ka - 20ka, and involved a total slip in the order of 25m.

What if we don't have the historical information?

As these earthquakes occurred long before European settlement in Australia, there is no record of their effects on people or the built environment. The potential effects must therefore be modelled based upon an understanding of the magnitude of earthquake the fault is capable of producing, the propagation of earthquake energy through the earth from source to the site of interest, and the properties of the structures exposed to ground shaking.

Figure 1: Cadell Fault Scarp. The modelled rupture and epicentre

Figure 1: Cadell Fault Scarp.
The modelled rupture and epicentre

Model uncertainties and their implications

Ground-motion attenuation is the change in the intensity of ground-shaking as the seismic waves propagate away from the earthquake rupture (Figure 2). Although there are a number of variables involved, as a general rule, earthquake waves will move fast with low amplitude through rock, and will slow and increase in amplitude when passing through regolith. Buildings of different construction type are susceptible to different combinations of wave speed and amplitude of ground shaking.

The uncertainties in ground motion models for Australia is the largest single factor limiting our ability to accurately estimate the impact of earthquakes such as ones on the Cadell fault. We have the computational tools to model the impact of future earthquakes but at present we only have preliminary fundamental physics models to underpin these tools. These physics models are essential for earthquake risk modelling, and are a current focus for research within Geoscience Australia.

Figure 2: Propagation of earthquake energy through the Earth

Figure 2: Propagation of
earthquake energy through
the Earth

Population and infrastructure at risk

Geoscience Australia's geological and geomorphological study of the fault determined the geometry of the fault and the likely magnitude of events that might be expected on the fault. This information, together with knowledge of how earthquake energy propagates through the earth (Figure 2), was used to model the distribution of strong ground shaking that would result from an earthquake event.

There are four significant population centres close to the Cadell fault that would be exposed to strong ground shaking from a large earthquake on the Cadell Fault (Figure 1):

  • Bendigo ~60 000 people
  • Shepparton ~30 000 people
  • Echuca ~10 000 people
  • Deniliquin ~8 500 people.

In order to calculate the impact of an event on these communities, the vulnerability of various elements of the built environment (e.g. infrastructure, dwellings) to strong ground shaking must be combined with the exposure of these elements to certain levels of ground shaking during an event.

What the modelling has shown

Figure 3: Mw 6.8 Atkinson and Boore Model

Figure 3: Mw 6.8 Atkinson and
Boore Model

Geoscience Australia studied the possible impacts to residential structures from both the maximum credible earthquake (Mw 7.2) as well as smaller but potentially more frequent earthquakes (Mw 6.8) on the Cadell fault. Figures 3 and 4 show the impact of earthquakes measuring Mw 6.8 and 7.2 using the attenuation model of Atkinson and Boore.

The results obtained from the study on this region demonstrated that a maximum magnitude event would have a catastrophic impact on the local area, with over 1 000km² sustaining greater than 50 per cent loss of residential structures. Even smaller earthquakes (Mw 6.8) have the potential to significantly impact on regional communities in south-eastern Australia, and due to their relative frequency are found to be more of a risk than the maximum credible earthquake. It is important to note, however, that the largest uncertainties arise in the choice of attenuation model, reducing the confidence that we can place in our estimates.

Figure 4: Mw 7.2 Atkinson and Boore Model

Figure 4: Mw 7.2 Atkinson and
Boore Model

Uses of this Information

Understanding how the subsurface and the built environment responds to ground shaking is important to understanding the potential impacts of an earthquake. This can have significant flow-on effects to building codes in an area, because each building is built to withstand a certain level of shaking.

The modelling demonstrated above has potential implications for:

  • revising building codes so that the built environment can be designed to withstand plausible levels of ground shaking
  • Emergency Management planning for disasters. i.e. prioritisation of emergency response
  • modelling and estimating damage and loss for economic and insurance purposes.

1 Mw is a logarithmic measure of earthquake size, similar to the Richter scale but better suited to very large events.

Topic contact: Last updated: March 12, 2014