Intraplate earthquakes in Australia - Intraplate Earthquakes - page 28

magnitudes below M W 5.0 (e.g., Leonard, 2008 ) , and thus require significant extrapola-
tion up to magnitudes at which the most damaging ground-shaking might be expected
(Wheeler, 2009a , b). Australia is uniquely suited to assessing the validity of M max estimates
derived from the instrumental record of seismicity by virtue of its extraordinary neotectonic
and paleo-earthquake record spanning many tens of thousands of years (e.g., Clark et al .,
2011a ) .
Where a fault has been studied paleoseismologically, single-event scarp lengths and
single-event displacements can provide two independent estimates of the characteristic
magnitude (cf. Schwartz and Coppersmith, 1984 ) for that fault, with scarp length being
considered the more reliable measure (e.g., Wells and Coppersmith, 1994 ; Hemphill-Haley
and Weldon, 1999 ; Leonard, 2010 ) . Where a fault has not been studied in detail the
“characteristic” rupture of the entire scarp length during each morphogenic event, subject to
appropriate caveats regarding segmentation and dip, might be assumed in order to estimate
future large earthquake potential (e.g., Hemphill-Haley and Weldon, 1999 ; Stirling et al .,
2002; Wheeler, 2009a ) .
In most regions of Australia the neotectonic record is either incomplete, or under-
explored (Sandiford, 2003b ; Clark et al ., 2011a ) . As such, it is not possible to assert
with confidence that future large earthquakes will be restricted to known scarps/faults with
documented source characteristics. It is therefore desirable to aggregate characteristic event
magnitudes from a group of faults with similar source characteristics from within a region
of interest (e.g., neotectonic domain or aggregation of domains [Leonard et al ., 2012 ; Clark
et al ., 2012 ] ) to provide an estimate of M max . This can be achieved with varying levels of
confidence, depending upon the estimated completeness of the paleo-record.

2.5.1 Scarp length as a proxy for paleo-earthquake magnitude

A relatively large area of 10 m resolution DEM data in the southwest corner of the Yilgarn
Craton (D1) ( Figure 2.4 ) allowed for mapping of fault scarps in unprecedented detail
(Clark, 2010 ) . Based upon an assessment of erosion and landscape modification rates, it
was estimated that most scarps representing events of M W 6.5 that had occurred in the last

100 ka were captured in the catalogue (Leonard and Clark, 2011 ) . The scaling relations
of Leonard ( 2010 ) were used to develop a paleo-seismicity catalogue comprising 65 events
(Leonard and Clark, 2006 , 2011), which was subsequently combined with the catalogue
derived from instrumental seismicity. The data were found to exhibit typical truncated
Gutenberg-Richter recurrence characteristics with a slope ( b ) of 0.9-1.0 between M W 6.5
and 6.9. A rapid roll off in recurrence occurred above M W 6.9, with an asymptotic value of
M W 7.25

0.1 considered to be the M max for non-extended cratonic crust typified by the
YilgarnCraton (Leonard andClark, 2006 , 2011). A less well constrained M max of M W 7.65

±

±
0.1 was determined for scarps representing extended crust in the rift basins flanking the
western margin of the continent (D6).