Intraplate earthquakes in Australia - Intraplate Earthquakes - page 20

hundred metres (e.g., Sandiford, 2003b ; Quigley et al ., 2006 ) . Further east, within non-
cratonic Australia ( Figure 2.4 -D4), faults are commonly found in en echelon arrangements,
with fault complexes extending for several hundred kilometres along strike (Beavis, 1960 ,
1962; Moye et al ., 1963 ) . As little as 15 m (e.g., Canavan, 1988 ; Robson and Webb,
2011 ; McPherson et al ., 2012b ) and as much as several hundred metres of neotectonic
displacement has been documented on several of these features (Beavis, 1960 , 1962; Moye
et al ., 1963 ; Abell, 1985 ) , and there is no clear relationship to historic seismicity. Faults
within the extended continental crust fringing the Australian continental margin ( Figure
2.4 - D5, D6), a remnant of the Cretaceous breakup of the supercontinent Gondwana (e.g.,
Ve eve r s , 2000 ) , have some of the larger throws of any Australian neotectonic faults (greater
than 100 m; Holdgate et al ., 2003 ) . Extensional structural architecture is largely preserved
(e.g., Hill et al ., 1994 ) , and neotectonic faults are often spatially associated with earthquake
epicentres.
2.3.1 Variation in fault scarp length and vertical displacement
The neotectonic data compiled by Clark et al . ( 2012 ) contain two semi-quantitative variables
useful for characterising fault behavior - length and vertical displacement. The population
distributions for Australian fault scarp length and vertical displacement data are presented
in Figure 2.5 a , b. Fault length is defined as the along-strike distance (tail to tail) of discrete
geomorphic features (most often fault scarps) that are considered to represent one or more
surface-rupturing earthquake events. Vertical displacement is the vertical separation across
a topographic scarp or fold. Many of the fault-length values reported by Clark et al . ( 2012 )
might be expected to be underestimates, as vertical displacement tapers towards the tails
of ruptures, resulting in these scarp sections being less discoverable in digital elevation
data: the primary tool for their identification and characterisation (Clark, 2010 , Clark
et al ., 2011a , 2012). It is also reasonable to assume that relatively short scarps are under-
represented as they are less discoverable. This factor may explain the positive skew in both
length and height data ( Figure 2.5 a , b). Furthermore, the resolution and noise content of
digital elevation data from various sources might be expected to affect the precision of both
length and vertical displacement measurements.
Interpolated surfaces for Australian neotectonic data demonstrate the spatial variation
in these fault parameters ( Figure 2.5 c , d), and these spatial patterns are borne out in the
statistics for each of the neotectonic domains ( Figure 2.5 e , f). The cratonic domains (D1,
D3) are characterised by the lowest vertical displacement values ( Figure 2.5 d , f). In view
of the extremely low erosion rates in these parts of Australia (Bierman and Caffee, 2002 ;
Belton et al ., 2004 ; Chappell, 2006), this is indicative of very low time-averaged rates
of morphogenic seismicity. Scarp lengths in D3 are up to 100 km greater than in D1
( Figure 2.5 e ), raising the possibility that relief in the Proterozoic mobile belts is built in
fewer, larger earthquakes. Five fault scarps have been subject to detailed paleoseismological
investigation in this crustal type: (1) Meckering (Clark et al ., 2011a ) , (2) Hyden (Crone