Conference PaperPDF Available

Earthquake clusters in the southwest Australia seismic zone, June 2016 - May 2017

Authors:

Abstract and Figures

This report is the 4th in a series of annual reports describing the annual seismicity of the South West Australia (SWA) zone since June 2013. In previous reports, 36 earthquake cluster localities were identified. In this report, covering the period June 2016-May 2017, 101 of the 207 events located by Geoscience Australia in the SWA zone have been relocated using extra data. Of the 207 events, about 170 have been allocated to one of 19 centres identified as being active during the year. Ten of these centres were identified in previous annual reports, and nine have been noted for the first time. However, of these nine, at least five were also active prior to 2013. The relocations have resulted in a number of good focal depth determinations, all less than 7 km, and it is implied that most focal depths in the zone are less than 5 km. Identification of additional events not located by GA suggests that the magnitude-completeness level in SWA is above the ML 1.6 level previously suggested.
Content may be subject to copyright.
Australian Earthquake Engineering 2017 Conference, 24-26 November, Canberra
Earthquake clusters in southwest Australia seismic zone,
June 2016 May 2017
V.F. Dent
University Associate, Curtin University, Perth, W.A.
Email: vic_dent@yahoo.com
Abstract
This report is the 4th in a series of annual reports describing the annual seismicity of the
South West Australia (SWA) zone since June 2013. In previous reports, 36 earthquake
cluster localities were identified. In this report, covering the period June 2016-May 2017,
101 of the 207 events located by Geoscience Australia in the SWA zone have been relocated
using extra data. Of the 207 events, about 170 have been allocated to one of 19 centres
identified as being active during the year. Ten of these centres were identified in previous
annual reports, and nine have been noted for the first time. However, of these nine, at least
five were also active prior to 2013. The relocations have resulted in a number of good focal
depth determinations, all less than 7 km, and it is implied that most focal depths in the zone
are less than 5 km. Identification of additional events not located by GA suggests that the
magnitude-completeness level in SWA is above the ML 1.6 level previously suggested.
Introduction
1.1 Preamble
The South West Australia (SWA)
seismic zone, to the east of Perth, as
defined by Leonard (2008) (Figure
1), is of interest because it is one of
the best regions in the world to study
the less-well-known phenomenon of
intra-plate seismicity. It saw three
ground-rupturing events in the 11
year period 1968 1979. Since 2014
the author has been examining the
region in 12 month periods. The aim
has been to demonstrate the clustered
nature of the seismicity, and to
pinpoint as exactly as possible the
location of each cluster. We are
attempting to determine if seismicity
in particularly active areas is
recurring from the same location, or
if it is migrating to a slightly different
location, as might be expected if
seismicity was occurring along a
fault line
1.2 Overview of seismicity in 2016-
2017
In simple numerical terms, 2016 was
more active than the previous 3 years
studied (Table 1). However, much of
this can be attributed to the ML 4.2
Figure 1 Seismicity in the SWA zone, Jun2016- May 2017
Australian Earthquake Engineering 2017 Conference, 24-26 November, Canberra
Table 1 Summary of yearly earthquake numbers
Magnitude
(ML) Range
1994
2013-
2014
2014-
2015
2015-
2016
2016-
2017
All events
132
290
167
73
209
ML 2.0 2.9
53
214
124
48
168
ML 3.0+
6
4
2
2
9
ML 4.0+
1
1
event in January 2017 near Bonnie Rock,
(the first ML4 event in SWA since May
2009), and which had ~ 50 aftershocks.
Ten events occurred at nearly the same
location (foreshocks?) in November 2016.
Even without this large event and its
aftershocks, there were more magnitude 3
events than in recent years, although the
number of active cluster locations
seems to be similar. As in previous
years, the region 30o 31oSouth has
been by far the most active part of
the SWA zone.
1.2 Earthquake zonation and
classification
As in the previous reports, the
SWA seismic zone has been
divided into a number of
rectangular zones (Figure 1) to aid
analysis. Two of the 8 zones used
in the previous reports have been
subdivided, so there are now 10
zones. A new term introduced here
is the “proto-cluster” – this is for a
small number of earthquakes which
may define a cluster, but there is a
degree of uncertainty, either that
they are a cluster, or in the correct
location of that cluster.
To aid in comparing this year with
earlier data, events for the previous
12 months are also plotted in the
zonal analyses. Also, the author
noticed that in 1994, many of the
cluster locations noted here were
also active, so events in that year
are also plotted.
1.3 Relocations and the Public
Seismic Network (PSN)
As has been done in previous
reports, about half of the
Geoscience Australia (GA)
locations for events in the study
year (2016-17) have been revised
(i.e., relocated) using the EQLOCL
location program, and with the
addition of phase data from Public
Seismic Network (PSN) stations in
the SWA region. The relocations
generally have much lower
Location
zone
New
name
Old
name
Ref.
Latit-
ude.
Longit-
ude.
N of Burakin
A
A1
A
(1)
-30.46
116.95
W of Burakin
A
A2
B
(1)
-30.53
117.05
SW of Burakin
A
A3
C
(1)
-30.63
117.10
W of Cadoux
A
A4
D
(1)
-30.79
117.10
Kalannie
A
A5
E
(1)
-30.15
117.17
New Norcia
A
A6
Mu
(4)
-30.93
116.30
W of Kulja
A
A7
new
N of Beacon
B
B1
G
(1)
-30.24
117.75
Nw of Beacon
B
B2
P
(2)
-30.40
117.75
Kalannie
B
B3
Alpha
(3)
-30.26
117.28
NW of Beacon
B
B4
Nu
(4)
-30.31
117.65
NE of Kalannie
B
B5
new
N of Koorda
C
C1
F
(1)
-30.64
117.47
N of Bencubbin
C
C2
J
(1)
-30.64
117.80
NE of Koorda
C
C3
K
(1)
-30.62
117.60
S of Bencubbin
C
C4
Q
(2)
-30.95
117.90
Yorkrakine
D
D1
H
(1)
-31.45
117.67
Wyalkatchem
D
D2
I
(1)
-31.25
117.45
N of Cunderdin
E
D3
M
(1)
-31.45
117.15
Nungarin
D
D4
Gam
(3)
-31.24
118.24
NW of Kellerb
D
D5
Omi
(4)
-31.43
117.50
N of Kellerb
D
D6*
Pi
(4)
-31.49
117.69
W of Wyalkcatch
D
D7
new
SE of Wyalkcat.
D
D8
new
Youndegin
E
E1
Delta
(3)
-31.77
117.36
W of Bruce Rock
E
E2
Rho
(4)
-31.84
117.84
N of Doodlakine
E
E3
new
North of Hyden
F
F1
R
(2)
-32.30
118.90
Kokeby
F
F2
U
(2)
-32.25
116.98
Morbinning
F
F3
V
(2)
-32.14
117.19
East of Pingelly
F
F4
W
(2)
-32.49
117.31
S of Quairading
F
F5
Epsil
(3)
-32.24
117.35
SE of Dumblyun
G
G1
S
(2)
-33.46
117.66
N of Kojonup
G
G2
Theta
(3)
-33.69
117.03
Newdegate
G
G3
A
(1)
-33.21
119.16
Nyabing
G
G4
B
(1)
-33.50
117.92
N of Rocky G.
H
H1
T
(2)
-34.35
117.05
Bonnie Rock
L
L1*
Beta
(3)
-30.56
118.57
Barbalin
L
L2
new
Mt Jackson
L
L3
new
S of York
M
M1
L
(1)
-31.98
116.72
N of Meckering
M
M2
N
(1)
-31.62
117.08
S of Meckering
M
M3
O
(1)
-31.70
117.04
Meckering
M
M4
new
Australian Earthquake Engineering 2017 Conference, 24-26 November, Canberra
uncertainties in their parameters than the original GA locations.
The PSN network is a group of low-cost seismographs set up by the author, and had its
beginnings in 2006 (Dent, Harris & Hardy, 2010). There are now more PSN stations in the
SWA region than there are GA stations, and, while they may have significant periods of
“down-time”, they usually provide sufficient data to allow significantly more reliable
locations to be achieved.
1.4 Precis of previous findings and research directions
Results from previous years point to the possibility that almost all SWA events belong to
cluster groups, and 36 such locations were defined in the papers mentioned above. The simple
alphabetical naming system used in those papers is proving inadequate, and the cluster
locations have been renamed, as shown in Table 2, to a new, zonal-based nomenclature. The
cluster locations which have been identified appear to be active for variable periods, and then
become dormant for variable periods, in a similar fashion to palaeo-seismic activity as
described by Clark et al., (2012), but on a very different time-scale. It is obviously necessary
to study the whole SWA zone for a relatively long period, to establish is this is in fact the
case. At first appraisal, the seismicity seems to be consistent with the above hypothesis,
although the 4 year time period so far studied is relatively short.
The rupture mechanisms, and the faults responsible for the clustered nature remain uncertain.
So far, no clear connection with pre-existing mapped has been established. Data from this
year might be an exception, as the Bonnie Rock cluster of January 2017 may be connected to
a small mapped fault in the nearby Jouerdine Nature Reserve (Dent, 2017, this volume).
Dawson et al., (2008) suggested that the Kalannie sequence of 2005 was related to a group of
en-echelon faults, but these faults had not been previously mapped, and seem to have no
surface manifestation.
Table 3 Cluster properties
Location
zone
New
name
Old
name
status
Latit-
ude
Longit-
ude
# of
event
Max
mag
Most active
period
Burakin
A
A3
C
old
5
2.4
21 Sep 2016
West of Kulja
A
A7
--
new
30.47
117.18
7
2.4
19 Jun 2016
N of Kalannie
B
B3
alpha
old
7
2.3
August 2016
NE of Kalannie
B
B5
new
30.10
117.41
3
3.3
Jun/Jul 2016
WNW of Beaco
B
B2
P
old
~27
3.0
Sep-Dec 16
Bonnie Rock
L
L2
new
30.66
118.42
~65
4.2
Jan 2017
Barbalin
L
L3
new
30.87
118.04
3
2.1
29 Aug 2016
Mt Jackson
L
L4
new
30.39
118.74
1
2.4
10 Sep 2016
Nw of Kellerb
D
D6
Pi*
old
30.495
117.695
9
2.5
Jun/Jul 2016
W of Wyalkat.
D
D7
new
31.18
117.28
3
2.4
Jan. 2017
SE of Wyalk
D
D8
new
31.26
117.63
2
2.4
31 Oct 2016
Doodlakine.
E
E3
new
31.52
117.91
3
2.3
April 2017
Meckering
M
M2
N
old
4
3.0
18 Apr 2017
Meckering
M
M4
new
31.648
117.025
7
1.6
26 Nov 2016
N of Hyden
F
F1
R
old
12
2.7
Jan/Feb 17
N of Brookton
F
F2
U
old
6
2.6
Jun/Jul 2016
E of Beverley
F
F3
V
old
2
E of Pingelly
F
F4
W
old
3
2.1
Newdegate
G
G3
sigma
old
3
2.6
Oct Jan 17
Australian Earthquake Engineering 2017 Conference, 24-26 November, Canberra
2 Analysis of 2016-17 data by zone
2.1 Zone A “Burakin”, 30.0o S- 31.0o S (Figure 2)
This zone includes the centre of Cadoux, site of a Magnitude 6.1 event in 1979. It is not clear
if present day seismicity in the area is related to this fault. Since 2001, much of the activity
seems to be related to the three ML 5 events near Burakin, about 20 km north of Cadoux, in
2001-2002 (Leonard, 2002). In the current period, there were 14 events in the zone, and a
north-south trend seems to be present. The largest event, in December 2016, was an ML 3.4
event south of Cadoux,
about 5 km south of
Location A4. It seems
to be an isolated event,
as are some other
largish Cadoux events
of recent years. Three
other small events in
the area are not clearly
related to the ML 3.4
event.
Two events northwest
of Wongan Hills
(January and March,
2017) could be called a
cluster, but since there
are only two, and small
(both ML 1.9), they are
here called a “proto-
cluster”.
The GA locations of 5
events (largest ML
2.4), ~10 km west of
Kulja, between 18- 20
June 2016 plot in zone B, but the relocations move them into Zone A. They define a new
cluster, location A7, which is listed in Table 3. This table summarises the activity of each
cluster centre active between June 2017 and May 2017.
GA locations show a cluster of 5 events close to Burakin in late Sep-mid Oct 2016 (largest
ML 2.4). However, the residuals are large, indicating large uncertainties in the locations. The
relocations are also not good, but suggest that the true locations are some 10 km south of the
ML 2.4 event. This puts them close to cluster location A3, and they are tentatively assigned to
that location. This location was the centre of the significant cluster, north of Cadoux, in Sep-
Oct 2000, described by Leonard & Boldra (2001).
2.2 Zone B “Beacon” (Figure 3, 34 events, 4 cluster groups)
Figure 2 Zone A seismicity June 2015 May 2017
Australian Earthquake Engineering 2017 Conference, 24-26 November, Canberra
The two largest events (both ml 3.3)
were from a cluster of 3 events north
of Kalannie in late June-early July
2016. The GA locations of the two
ML 3.3 events are about 20 km apart,
but the relocations bring them much
closer together. The approximate
centre of the group is used to define a
new cluster (B5). This cluster is about
20 km ENE of the centre of the large
Kalannie swarm of 2005 (Location
A5), which was the subject of an in-
depth study by Dawson et al., (2008).
Numerically, seismicity in this zone
is dominated by a cluster of events
northwest of Beacon (~19 events,
largest ML 3.0). The most active
period was September to December
2016.The GA epicentres have a
distinct NE trend, but the relocations
of several events at the extremities of
the trend suggests that the trend is
unreal, and that the events are
clustered in a region a few km south
of the predefined location, B2
A cluster of seven events closer to
Kalannie in September 2016 appears
to be at location B3 and the events
represent a continuation of activity seen near there in the previous 12 month period.
2.3 Zone C “Koorda” (Figure 3, 11 events largest ML 2.3)
The original zone C has been subdivided, to put events near Bonnie Rock into a unique zone
(Zone L). The Koorda region was very active between 2012-2015, and 4 cluster centres have
been defined in the zone (C1-C4). In the current period, there are only 11 small and relatively
scattered events. 5 of these join clusters in adjacent zones when relocated (2 events southeast
of Bencubbin on 29th August, when relocated move into the new Zone L, as described below).
Zone L Bonnie Rock 73 events, largest ML 4.2 (Figure 4)
There is only one defined cluster location in this (new) zone, location Beta, identified in Dent
(2015) but here renamed to location L1. This was noted as being an approximate location
when defined, and more work, presented in Dent & Collins (2017) has suggested it be moved
about 8 km to the southeast, to an active location which was well-defined by a temporary
network in June 2000.
The significant activity following the ML 4.2 event north of Mukinbudin on 3rd January 2017,
represents about 1/3rd of all the 2016-2017 seismic events studied in this report. The activity
in 2016-17 presented below is summarised from the more detailed report of Dent & Collins
(2017) mentioned above. The new seismograph (BR4) installed near the epicentre of the main
event, 10 days after it happened, has assisted greatly in providing better locations for the
events.
Figure 3 Zones B and C seismicity June 2015 May 2017
Australian Earthquake Engineering 2017 Conference, 24-26 November, Canberra
There were two significant
cluster groups in the 12
month period, the first being
a group of 8 events northeast
of Mukinbudin in November
2016 (largest ML 2.6).
Relocations in Dent &
Collins (2017) suggest that
these events came from the
same location as the more
significant cluster some six
weeks later, in Jan 2017.
This later cluster was
initiated by an Ml 4.2 event
(the largest in SWA for 7
years) on 3rd Jan 2017. Dent
& Collins (2017) suggest
that the cluster centre was at
-30.66, 118.42, and this was
relatively well determined,
as it used data from the close
and newly-installed
seismograph (BR4). This
location is about 20 km southwest of location L1.
The November 2016 events (probable foreshocks) seem to be “swarm-like” (i.e., no
particularly dominant event), whereas the events of January 2017, starting with the ML 4.2
event of 3rd January 2017 are more main-shock/ aftershock in character.
Two events in the southeast of Zone C, on 29th August 2016, when relocated, move to the
southwest of Zone L, close to another event on 17th June 16. This group location coincides
with that of a larger cluster in January 2001, identified as “near Barbalin” in Dent & Collins
(2017). This location has been named L3.
A single event about 50 km northeast of location L1, in November 2016, is apparently at the
same location as a group in 2013 (“southwest of Mt. Jackson”) as described in Dent (2017).
This cluster location is here defined as location L4, at the location given in Dent 2017.
Figure 5 Zones D, M and E
Figure 4 Zone L Bonnie Rock seismicity, June 2016 May 2017
Australian Earthquake Engineering 2017 Conference, 24-26 November, Canberra
There are other cluster groups in the area noted in Dent 2017 (near Mukinbudin and
Welbungin), and together they seem to define a NE-SW trend, and this possible structure was
therein named “the Bonnie Rock Seismic Lineation”.
2.4 Zone D, “Wyalkatchem” 31.0oS - 31.5oS (18 events, largest ML 2.9 Figures 5 & 6)
Northwest of
Kellerberrin The
GA locations of this
group of 9 events
(largest ML 2.5) in
June and July 2016
puts them close to a
smaller group 12
months earlier. This
group was given the
label Pi in last year’s
report, with the
caveat that it was not
well determined.
Some of the
relocations suggested
the group was
actually about 8 km
further east, and
relocations of the June/July 2016 events confirm that the locations are east of the GA
locations. Location Pi, or D6 in the new convention, has been moved to this new location in
Tables 2 and 3.
West of Wyalkatchem
This cluster of three events (Dec 2016-Jan 2017, largest ML 2.2) plots about 5 km southwest
of a cluster of 12 events (largest ML 2.5) in March/April 1994, noted in Dent (2008).
Relocations of the 2016-17 events suggest the cluster is about 8 km to the northeast of the GA
location, and close to the 1994 cluster location. A new cluster location called Location D7
has been defined for these events.
Southeast of Wyalkatchem
Two events southeast of Wyalkatchem (ML’s 2.3, 2.4) on 31 Oct 2016 have been used to
define a new cluster location, D8.
Events near Southern Cross. (Two events in September 2016, Ml 2.6 and ML 2.9). These
events are designated as a proto-cluster, because there are only two events, and their locations
are poor. Note that more intense activity returned to this location in late 2017, and the events
will be reviewed in next year’s report.
2.5 Zone M Meckering (Figures 6 & 7, 14 events, largest ML 3.0).
Three cluster centres have been defined previously in this (new) zone, two near Meckering,
and one south of York (Figure 5). Seismicity in the zone in 2016-2017 seems to consist of
two distinct clusters, one about 7 km ENE of Meckering (consisting of a mainshock on 18th
April 2017, ML 3.0, and 4 small aftershocks on the same day) and the other is a group of
Figure 6 Enlarged Wyalkatchem Kellerberrin region
Australian Earthquake Engineering 2017 Conference, 24-26 November, Canberra
small events about 2 km southeast of
Meckering (all on November 26th, 2016,
largest ML 1.6). Good relocations for events
in this zone are now possible, since the
installation of a PSN seismograph at a local
business in December 2015.
Events ENE of Meckering, April 2017
This group has a mainshock/ aftershock
character. Relocations bring the small
aftershocks back towards the main event, and
closer to cluster centre M2, which was
defined in Dent (2012) on the basis of two
events in late 2011. The events are allocated
to that cluster centre in Table 3.
Events south of Meckering, 26th Nov 2016
Relocations of these events bring them closer
together, and in particular, two events just
north of the town move south to join this
group, identified as a new location, M3, in
Table 3. The focal depths average about 4
km. Note that a solitary event about 8 km
south of Meckering, on 24th September, also
relocates close this location.
Focal depths of Meckering events
The focal depths of Meckering events may be
reliable because of the closeness of the MECK
station, and are listed in Table 4. This table lists
what are considered better focal depths obtained
by relocation, as discussed later in this paper.
The two cluster centres discussed above are
both about 4 km east of the Meckering Fault
scarp (Figure 7), and the focal depths are
consistent with the assumed easterly dip of the
fault (Gordon & Lewis, 1980).
2.6 Zone E, “Kellerberrin” 31.5oS 32oS
(Figures 5 & 6, 5 events, largest ML 2.3.)
This zone was previously called “Meckering
but is now renamed “Kellerberrin” since the
Meckering region has been excised to form the
new zone “M”. There are two pre-defined
cluster locations in this zone (E1 and E2, Figure
6), but the only grouping of significance in the
2016-17 year is one of three events north of
Doodlakine (August 2016 and April 2017).
Figure 7 Meckering region, enlarged
Table 4 List of better-determined focal
depths determined by relocations
Date & time
Nearest
Town
Dist.
to stn
Depth
(km)
16-11-26 0133
Meckering
2.1
4.3
16-11-26 0125
Meckering
2.5
3.3
16-11-26 2348
Meckering
2.5
4.3
16-09-24 1607
Meckering
3.9
5.8
17-04-18 2224
Meckering
8.1
6.2
17-01-15 2231
Bonnie Rk
8.1
0.4
17-02-12 0811
Bonnie Rk
8.5
0.3
17-04-18 1258
Meckering
8.7
3.9
17-03-05 1803
Bonnie Rk
9.2
1.9
17-04-18 0826
Meckering
8.3
5.9
17-02-22 1318
Bonnie Rk
10.1
2.0
16-07-15 1533
Cadoux
10.6
1.3
16-10-25 0933
Quairading
10.6
4.3
16-06-18 2343
Kellerberin
11.7
2.7
16-07-12 0055
Kellerberin
12.0
2.8
16-06-25 1946
Kellerberin
12.2
3.7
16-06-25 1837
Kellerberin
12.5
3.6
16-06-25 1837
Kellerberin
12.9
2.7
16-08-30 1050
Kellerberin
16.2
1.4
17-06-05 1020
Bencubbin
16.3
6.2
17-06-05 1519
Bencubbin
16.5
0.3
17-06-05 1519
Bencubbin
17.6
6.7
16-10-12 1825
Quairading
18.5
11.2
17-06-03 0750
Mukibudin
19.8
1.3
Australian Earthquake Engineering 2017 Conference, 24-26 November, Canberra
Relocations tighten the group, and the location appears to coincide with a cluster in June 1994
(grey events in Figure 6), which was noted in Dent (2008). This is cluster location has not
been named before, and is here defined as location E3.
2.7 Zone F, “Brookton-Hyden” 32oS - 33oS (Figure 8, 30 events largest 2 events
both ML 2.7, north of Hyden)
North of Hyden
Activity in Zone F in 2016-17 is dominated by events which are here allocated to Location
F1, north of Hyden, a centre which was quite active in early 2013. The events north of Hyden
were in two cluster groups, in late January-early February 2017 (6 events), and the 6 events in
the later group were all on 22nd May 2017. They are typical earthquake swarms i.e., there is
no single dominant event. Other smaller events from this location were detected by the PSN
network, but not located by GA.
North of Brookton
This cluster of ~ 6 events (June-Aug 2016, largest ML 2.6) may emanate from the same
location (F2), near the village of Kokeby, as a cluster in December 2013. Both the current
cluster and the 2013 cluster were main/aftershock in character as the first events were large
relative to the following events.
East of Beverley, February and April 2017
Two small events in February and April, 2017, may come from location F3, near Morbinning,
about 20 km east of Beverley, where a cluster of events occurred in July 2013.
Figure 8 Zones F and G
Australian Earthquake Engineering 2017 Conference, 24-26 November, Canberra
East of Pingelly, May 2017
This group of 5 small events (largest ML 2.0), mostly in May 2017, probably comes from
location F4, first identified from a group of three events in March 2014.
2.8 Zone G “Katanning”, 33oS to 34oS (Figure 8, 10 events, largest ML 3.1)
There are 5 previously defined cluster locations in this zone, shown in Figure 8 as G1
G5.The most significant group of events in the zone in 2016-17 is a cluster of 4 events
southeast of Newdegate, (largest ML 3.1 on 13th May 2017). The activity represents a renewal
of activity at location G3, which was defined in Dent (2016) on the basis of two events in
June 2015. Relocation of the ML 3.1 event brings it closer to Location G3. Note that event
locations in this region will have large uncertainties because of the poor distribution of
seismographs.
Examination of seismic data from PSN stations has identified numerous relatively small
events not located by GA, but for which approximate locations can be determined using S-P
times (Table 5). One of these (2nd Jan 2017 at 2035 UTC) also seems to be from the G3
cluster, and is also plotted on Figure 8. Its magnitude is estimated at ML 2.2
Three small events within ~ 20 km of Lake Grace (Sept. 2016, March & May of 2017) could
represent a cluster, with the disparity of locations being attributed to the difficulty of locating
events in this region. However, they not defined as a cluster here. Other small unlocated
events suspected of being in the Lake Grace area were detected by the PSN network (Table
5), and future work may result in the definition of a cluster centre there.
A single event about 20 km northeast of Nyabing plots very close to an event in December
2014, and relatively close to clustered events in 1994 and 1995. A “proto-cluster” is declared
for this location, and future events may help to define its location more precisely.
Other events near Gnowangerup
Random examination of seismograms from the seismograph GNOC, (located about 20 km
northeast of Gnowangerup), revealed a group of small, unlocated events, mostly in mid-
August, 2016, as indicated in Table 5. A composite location, using S-P times from PSN
stations in the region, suggests their origin is about 20 km north of Gnowangerup (largest
event about ML 2.0). This is also declared a “proto-cluster” location. Examination of
Table 5 Some of the unlocated earthquakes seen on PSN seismograms during 2016-17
Date/time
Estimated
area
est.
ML!
Date/time
Estimated area
est.
ML!
01/08/2016_09:41
Lake Grace?
1.9
15/08/2016_10:55
N of Gnowangerup
1.2
01/08/2016_10:02
Lake Grace?
1.8
15/08/2016_11:40
N of Gnowangerup
0.9
10/08/2016_23:38
N of Gnow
1.8
29/08/2016_1435
SW of Cadoux
1.5
11/08/2016_00:30
N of Gnow
2.3
13/09/2016_1426
Muja
2.4
11/08/2016_22:07
N of Gnow
1.6
29/12/2016_2332
Westonia
2.4
12/08/2016_19:14
N of Gnow
0.9
02/01/2017_2036
SE of Newdegate
2.2
12/08/2016_19:15
N of Gnow
1.2
03/01/2017_1621
Bonnie Rock
2.0
13/08/2016_18:39
N of Gnow
1.3
31/01/2017_2316
SE of Meckering
1.3
14/08/2016_12:01
Lake Grace?
1.6
05/04/2017_1825
N of Hyden
15/08/2016_09:45
N of Gnowa
1.4
28/04/2017_1952
N of Hyden
1.8
15/08/2016_10:01
N of Gnowa
1.7
16/05/2017_1523
Lake Grace
1.9
28/05/2017_2140
E of Hyden
! estimated ML ( Local magnitude) - estimated from amplitudes on PSN stations.
Australian Earthquake Engineering 2017 Conference, 24-26 November, Canberra
historical data shows an ML 2.4 event at almost precisely this location in April 2013. This
cluster resembles a cluster of probably hundreds of small (but felt) events from a location,
thought to be just north of GNOW, in late 2007, which was described in Dent (2008). Only
three events in that cluster were large enough (largest 20th Nov 2007, ML 2.3) to be located
by the regional (GA) seismograph network. Local farmer’s concerns following that cluster
prompted the installation of the GNOW and GNOC seismographs.
2.9 Zone H, Cranbrook, 34o S 35o S
There were no events in Zone H in the 12 month period, although, a small event (ML 1.9)
near Ongerup relocates to 34.05oS, which moves it into the Zone H region. The last cluster to
be observed in the Zone H region was that north of Rocky Gully in May-June 2013 (location
H1), largest ML 2.9.
3 Summary and Discussion
3.1 Summary of the zonal analysis of 2016-17 seismicity
Of the 207 events reviewed here, Table 3 shows ~170 have been assigned to 19 cluster
locations. Nine of these locations are defined here for the first time, including location L2,
source of the Bonnie Rock cluster, which was by far the most significant seismic event of the
year. Although it is defined here for the first time, work by Dent (2017) suggests that the
location has shown at least occasional activity over the last 25 years. Of the other new cluster
locations, at least four been active in earlier years, but also were not defined as cluster
locations. These are L2 (active in 2001), L4 (active in 2013), D7 (active in 1994), and E3
(also active in 1994). Ten of the cluster locations active in this year were defined in earlier
reports, and the most significant being B2 (northwest of Beacon, 20 events), and F1, (north of
Hyden, 12 events).
In total now, 44 cluster locations have been defined in SWA. Of the remaining ~ 40 events
not assigned to a cluster location, later relocations may move some to a known cluster
location. Alternatively, they may represent remnant activity from cluster locations not yet
defined. Some, probably a minority, will remain as “isolated” events. The most significant
isolated event in the 2016-17 period was the ML 3.4 event southwest Cadoux, in December
2016. Note however that it is very close to cluster location A4.
3.2 Earthquake focal depths.
Focal depths are usually less well
determined than
latitude/longitude, and close
seismographs (<~20 km) are
needed in order to determine a
depth with some confidence. GA
depths have been constrained to
10 km since December 2016, and
many of the PSN relocations of
events plotted here have been
constrained at 2 km depth. In
some instances in this year,
events are close to PSN
seismographs, e.g. Bonnie Rock,
Meckering, and the depths
computed have reduced
Figure 9 Distribution of well-determined focal depths
Australian Earthquake Engineering 2017 Conference, 24-26 November, Canberra
uncertainties, estimated to be about +/- 4 km. These more-reliable depths (i.e., nearest
recorder is < 20 km distant) are listed in Table 4, and the depth-distribution is shown in
Figure 9. Based on this data, it is suggested that most earthquakes in the SWA zone are less
than 5 km deep.
The number of defined cluster locations seems to be increasing by about 10 locations each
year. Whether this will decrease over time, as most “recurring” locations become identified, is
yet to be seen.
3.3 Discussion
The analyses presented in these reports were inspired by the Leonard & Boldra (2001) report
on a field survey conducted to monitor an earthquake swarm north of Cadoux in September-
October 2000 (largest event ML 3.6). They identified four cluster locations (A1 to A4 in
Table 2), and concluded that the majority of the ~ 1,700 events detected by the survey came
from location A3. This location seems to have been relatively quiet since 2000, but location
A2 seems to have been the focus of the extremely active Burakin swarm of 2001 2003,
which included three magnitude 5 events. Activity at A2 seems to have continued, but was
only minor or non-existent in the current period.
Leonard & Boldra (2001) quoted a personal communication from Gibson, saying that such
clusters occurred in SWA every 1 2 years. Major clusters following the Cadoux 2000
swarm were at Burakin, as mentioned above, and also near Koorda in 2004 (Location C1),
Kalannie in 2005 (location A5), Beacon in 2007 (location B1), again near Beacon in 2012
(location B2), and near Bonnie Rock in 2017. Gibson’s observation might be correct if you
consider only “major” clusters, but results over the last 4 years indicate that at least 10 cluster
locations generally show activity in any given year, and clusters can be of all “sizes”, from
“minor” to “major”.
Interestingly, the “major” clusters in the above paragraph are all in the northern part of the
SWA zone, between 30 and 31 degrees south. Understanding the behaviour of clusters over
time is also probably important. The report of Ghasemi et al., (2013) would seem to assign
them to being “extended aftershock sequences”, but in many cases, there seems to be no
obvious “main-shock” present. In the case of the January 2017 sequence near Bonnie Rock,
there was a main-shock, but there was a cluster of events at the same location some 2 months
earlier, without a mainshock, and the report by Dent (2017) suggests that this location (L1)
has seen active periods back as far as 1968. It seems it is possible that clusters from a single
location may show a wide variety of magnitude distributions, from fore-shocks/main-shock
and aftershocks to a true swarm-like nature, where there can be several or many events near
the maximum magnitude of the group.
Notable from the studies so far is the fact that activity seems to return to the same location,
after long periods of quiescence, possibly decades or longer. This would not be unusual if
faults were observed at these locations, but this is not the case so far. More data is required to
confirm this observation, and obviously earthquake location errors will need to be reduced,
from the order of +/- 10 km, to about +/- 2 km or less.
Completeness of GA earthquake catalogue
The density of the PSN network, and the relative nearness of many of the stations to the
cluster centres, has revealed numerous events not located by GA, the larger ones of which are
listed in Table 5. Approximate magnitudes can also be determined, and these are also shown.
Leonard (2008) suggested that the GA earthquake catalogue in the SWA region was probably
complete for events > ML 1.6 since 2009. However, this study, and also that of Dent (2009)
Australian Earthquake Engineering 2017 Conference, 24-26 November, Canberra
suggest that events up to magnitudes ~ ML 2.2 may escape location by GA. This may not
have a significant effect on the Seismic Hazard Analysis for the region, because, as they are
usually small events within a larger cluster, they are usually eliminated as being “dependent
events”.
Events outside the SWA seismic region
There are few events around the perimeter of the SWA region, possibly partly because of the
inability of the existing seismic network to detect small events (i.e. below ~ ML 2.5) in these
areas. However, there were significant activity north of Ravensthorpe in March and
September 2017 (three events between ML 3.2 and 3.5), some 50 km east of the SWA
boundary used here (119.5oE). The Ravensthorpe region has seen some large earthquakes in
the past (e.g. ML 5.2 in October 2000), and also has a significant palaeo-scarp (Estrada, B.
2008). This suggests that there is little difference in the seismic characteristics of this region,
and the rest of the SWA region to its west.
Acknowledgements
The author wishes to acknowledge the significant contributions of Clive Collins and Virginia
Ward in preparing this paper, and also Eddie Leask of GA for providing necessary data.
4 References
Clark, D., McPherson, A. & Van Dissen, R., (2012). Long-term behaviour of Australian
continental region (SCR) faults. Tectonophysics, 566, 1-30.
Dawson, J., Cummins, P., Tregonning, P., and Leonard, M. (2008). Shallow intraplate
earthquakes in Western Australia observed by Interferometric Synthetic Aperture Radar.
Journ. Geophys. Res. V.113, B11408.
Dent, V. F., (2008) Improved hypocentral estimates for two recent seismic events in south-
western Western Australia using temporary station data. In Proc. AEES 2008 Conference,
Canberra, ACT.
Dent, V. F., (2009). The Beacon, WA, earthquake swarm of 2009. In Proc. AEES 2009
Conference, Newcastle, NSW.
Dent, V. F., Harris, P, and Hardy, D., (2010). A new seismograph network in the southwest
seismic zone of Western Australia. In Proc. AEES 2010, Conference, Perth, WA.
Dent, V. F., (2010). A NE lineation of epicentres northeast of Perth fact or fiction? A
review of earthquake hypocentres in the region, 2005. In Proc. AEES 2010, Conference,
Perth, WA.
Dent, V. F., (2012). Evidence for shallow focal depths and denser locations for three
southwest seismic zone earthquake clusters, 2011. In Proc. AEES 2012 Conference, Gold
Coast, Qld.
Dent, V. F. (2014). Earthquake clusters in southwest Australia in 2013-14. In Proc. AEES
2014 Conference, Lorne, Vic.
Dent, V.F. (2015) Clustered seismicity in the Southwest Australia seismic zone, 2014-15. In
Proc. AEES 2015 Conference, Sydney, N.S.W.
Australian Earthquake Engineering 2017 Conference, 24-26 November, Canberra
Dent V.F., (2016) A preliminary map of cluster locations in southwest Western Australia, 1990
2016 in Proc. AEES 2016 Conference, Melbourne, Vic.
Dent V.F., & Collins C.D.N., (2017) . A review of seismicity in the Mukinbudin Bonnie
Rock area of WA. In Proc. AEES 2017 Conference, Canberra ACT
Estrada, B. (2008). Neotectonic and Palaeoseismological studies in the southwest of Western
Australia PhD thesis, University of Western Australia.
Ghasemi, H., Leonard, M., Robinson, D., Owen, K. & Horspool, N. (2013) Aftershock
probability model for Australia Proc. of the AEES 2013 Conference, Hobart.
Gordon, F.R., and Lewis, J.D., (1980). The Meckering and Calingiri earthquakes, October
1968 and March 1970. Geological Survey of Western Australia. Bulletin 126.
Leonard, M. & Boldra, P. (2001) Cadoux swarm September 2000 an indication of rapid
stress transfer? In Proc. AEES 2001 Conference, Canberra, ACT.
Leonard, M. (2002). The Burakin WA Earthquake sequence Sept 2000 Jun 2002. In Proc.
AEES 2002 Conference, Adelaide, SA.
Leonard, M. (2008). One Hundred Years of Earthquake Recording in Australia. Bull. Seismol.
Soc. Am. 98, 14581470.
... These papers attempt to identify and provide some detail about the frequent earthquake clusters that occur in the region. Approximately 15 -20 clusters have been identified in each year, and a total of 44 cluster locations have been defined in previous reports by Dent (2014aDent ( , 2015Dent ( , 2016Dent ( & 2017. These cluster locations are shown in Figure 1. ...
... were recorded, correlating with an epicentral distance of about 21 km (Figure 6), assuming a shallow focal depth. Using these derived epicentral distances, the intersection of the circles ( Figure 6) suggests a common epicentre at 30.64 o S, 117.43 o E, which is about 3 km WNW of the location assigned to the events which occurred in 2013-2014(location F in Dent (2014a, or C1 in Dent (2017). This may represent a better location for the C1 cluster. ...
... The next largest was ML 2.4 northwest of Gnowangerup, on 19 April 2013. This event also appears isolated, but, as noted in Dent (2017), its location correlates with a cluster of small events in 2016 (none located by GA). It also correlates with a cluster of three events in November 2017 (largest ML 2.4), and is close to GA's location of an ML 3.7 event in October 1996 (-33.9, 117.909). ...
Conference Paper
Full-text available
Clustered intra plate seismicity in south west Australia
... Location map. Earthquake Zones of Dent (2014;2017). Earthquakes from July 2022 to June 2023 inclusive. ...
... four such regions in Australia for which he discussed the seismicity. Leonard's South West Australia region was divided by Dent (2014) into 8 latitudinally-based sub-zones, and later revised (Dent 2017) to the 10-zone configuration shown in Figure 1. The seismicity of SWA between July 2012 and June 2022 has been analysed from a "clustered" viewpoint in 8 previous reports by Dent and co-workers (eg. ...
Conference Paper
Full-text available
The wheatbelt region of southwest Western Australia is an area of elevated intraplate seismicity. The seismicity is highly clustered, with 80 active clusters identified over the last 20 years. This report reviews the seismicity between July 2022 and June 2024, which includes continuing seismicity from the Arthur River region, and a damaging magnitude 5.0 event near Gnowangerup. Most of the seismicity in this period occurs in 21 clusters; 10 were previously active and 11 are new. Of interest is the contemporaneous seismicity of 5 neighbouring cluster pairs, suggesting a possible causative link.
... It did not receive much more attention until Dent (2012) started listing and naming cluster centres. Fifty-one such centres have been defined so far (Dent 2017, and they are plotted on Figure 1. Some of the more significant clusters (i.e. containing at least one ML 4 event) are listed in Table 1 and are also shown (as black filled hexagons) in Figure 1. ...
... This was divided by Dent (2016) into eight zones by latitude, A-H (see Figure 1). Two of these zones were subdivided in Dent (2017), creating new zones L and M. In this paper, a new zone, (zone I), is appended to the east of the region to accommodate seismicity north of Ravensthorpe, and some other events which occurred during this study period. Also, zone A is extended northwards to include seismicity near Latham (Figure 1). ...
Conference Paper
Full-text available
Earthquake clustering has long been noted as a significant feature of the seismicity of southwest Western Australia. The cluster locations are of significance because they are sites of repeating seismicity for over a decade or more. Fifty-one cluster centres have been previously defined, mostly from seismicity during the 2012-2017 period. Seismicity during 2017-2020 is reviewed here, resulting in the identification of 28 cluster centres active during this period. Of these, 14 had been active prior to 2017 and 14 are new. The most significant of these were northwest of Beacon, representing ongoing activity, and north of Ravensthorpe, which is a new location. The cluster centres in southwest Western Australia appear to define a NNW trend, as has been noted in earlier studies, and may correlate with the granitoid components of the geology in that region of the Yilgarn craton. The northern region of the study area has more cluster locations than the southern region, which is consistent with earlier studies of the seismicity of the region.
... Cluster location G5 was named inDent (2017) on the basis of an ML 2.4 event in April 2014, and about 20 small events (max ~ ML 1.5) recorded by the seismograph GNOC in August 2016. It is about 10 km west of the 2019 activity, and its estimated uncertainty is about +/-5 km. ...
Conference Paper
Full-text available
Much of the seismicity in southwest WA occurs in localised groups of recurring earthquakes, or clusters. Four potential cluster sites have been identified in the Great Southern Region of WA in the period January to November 2019. A group near Newdegate contained the largest event of the year (ML 4.3), and it is suggested these events belong to an earthquake group first identified in 2016. The three other groups identified are near Denmark, Gnowangerup and Ongerup. Relocations of many of the events using the VIC5A earth model suggests tighter groupings for the events than the published Geoscience Australia locations. The Gnowangerup and Ongerup groups are also close spatially, and there may be a causal link between the groups. The region has been categorised as a region of relatively low seismicity, and the higher numbers of events in 2019 may be related to the occurrence of two ML 5 event in Lake Muir in late 2018. .
Conference Paper
Full-text available
Earthquake clusters in southwestern Western Australia between 1990 and 2016 have been mapped using a visual perusal of the Geoscience Australia earthquake catalogue. Also plotted are 36 relatively well-determined clusters (Dent 2014, 2015, 2016), from re-evaluation of epicentres mostly between 2014 and 2016. Approximately 100 centres have been identified, but further examination may show that some closely-spaced centres are actually coincident; a cluster of apparent clusters. Many smaller cluster groups may also be present in the data-set. Significant non-clustered activity in the period is also discussed. The majority of the seismicity in the area over the 26-year period examined occurs in clusters, and this pattern may also reflect the distribution of seismicity over a much longer term. The presence of cluster groups in a presentation of historical WA activity by Everingham (1966) is noted.
Article
Australia boasts arguably the richest Late Neogene to Quaternary faulting record in stable continental region (SCR) crust anywhere in the world. Variation in fault scarp length, vertical displacement, proximity to other faults and relationship to topography permits division of the continent according to fault character. Six onshore “neotectonic domains” are recognised, with an additional offshore domain proposed by analogy with the eastern United States. Each domain relates to a distinct underlying crustal type and architecture, broadly considered to represent cratonic, non-cratonic and extended environments. In general, greater topographic expression associated with faults occurring in extended crust relative to non-extended crust suggests a higher rate of seismic activity in the former setting, consistent with observations worldwide. Using the same reasoning, non-cratonic crust might be expected to have a higher rate of seismic activity than cratonic crust. This distinction, together with the variation in fault character between domains, should be recognised in attempts to identify analogous systems worldwide.
Article
A comprehensive new catalog of Australian earthquakes has been compiled and used to review the seismicity of Australia. The catalog contains 27,000 events, of which 17,000 are considered to be mainshocks. The catalog is com-plete for all Australian events with a magnitude greater than M L 5:5 since 1910, M L 5 since 1960, M L 4 since 1970, and M L 3:5 since 1980. It is complete for events in southern Australia above M L 3:5 since 1965 and M L 2 since 1980. Before the devel-opment of local magnitude scales for Australia, around 1990, the Richter magnitude scale was generally used. At 600 km (a typical hypocentral distance for much of Aus-tralia) the Richter formula overestimates magnitude by around 0.5 units. Thus, the results in the catalogs before and after the early 1990s are potentially discrepant. Most well-located Australian earthquakes are in the southern areas of the conti-nent, where the seismometer density is greatest. In general, the location uncertainty of Australian earthquakes is high. Only 60% of events are located with an uncertainty of 10 km or less. This percentage is smaller for earthquakes before 1980, and before 1960 very few events were located to within 10 km. The hypocentral depths of Aus-tralian earthquakes range mostly between 8 and 18 km, except for the southwest cor-ner of the continent where they are typically shallower than 5 km. The seismicity in some areas of Australia has been steady for at least 100 yr (in-cluding the southeast corner, the Flinders Ranges, and the northwest corner). In con-trast, seismicity in the southwest corner jumped by at least a factor of 6 in the 1940s and has been steady since then. Much of the rest of Australia is characterized by epi-sodic seismicity. These episodes begin with a period of high activity lasting 1–10 yr and they are normally associated with a large (M >6) earthquake. Following the large earthquake, there is often a period of moderate activity lasting from a few years to several decades. Before and after each episode is a quiescent period of low activity lasting 0.1–10 ka, during which the seismicity is more than an order of magnitude lower than during the period of high activity. Frequency-magnitude relations were calculated using events since 1970 from the new catalog. Gutenberg–Richter a-and b-values were calculated on an 85-km grid, and maps of the probability of an earthquake of M ≥ 4:9 occurring per year were derived. These results are very similar to the Global Seismic Hazard Assessment Program (GSHAP) map for Australia. The results were used to define four large (> 20; 000 km 2) seismogenic zones (Fig. 1). There are also several other small zones, some of which appear to reflect recent episodes, while others appear to be long-lived. The expected number of earthquakes M ≥ 5 and M ≥ 6, strain rate, and deformation rate is given for the four large zones, the remainder of Australia, and the whole Aus-tralian continent. The combined estimates of strain using seismic Global Positioning System (GPS) and Satellite Laser Ranging (SLR) data suggest east–west compressive deformation across southern Australia of 0:65 2:0 mm per year, likely to be in the 0.5–1.0-mm-per-year range.
A new seismograph network in the southwest seismic zone of Western Australia
  • V F Dent
  • P Harris
  • D Hardy
Dent, V. F., Harris, P, and Hardy, D., (2010). A new seismograph network in the southwest seismic zone of Western Australia. In Proc. AEES 2010, Conference, Perth, WA.
A NE lineation of epicentres northeast of Perth -fact or fiction? A review of earthquake hypocentres in the region
  • V F Dent
Dent, V. F., (2010). A NE lineation of epicentres northeast of Perth -fact or fiction? A review of earthquake hypocentres in the region, 2005. In Proc. AEES 2010, Conference, Perth, WA.
Evidence for shallow focal depths and denser locations for three southwest seismic zone earthquake clusters
  • V F Dent
Dent, V. F., (2012). Evidence for shallow focal depths and denser locations for three southwest seismic zone earthquake clusters, 2011. In Proc. AEES 2012 Conference, Gold Coast, Qld.
Earthquake clusters in southwest Australia in 2013-14
  • V F Dent
Dent, V. F. (2014). Earthquake clusters in southwest Australia in 2013-14. In Proc. AEES 2014 Conference, Lorne, Vic.
A review of seismicity in the Mukinbudin Bonnie Rock area of WA
  • V F Dent
  • C D N Collins
Dent V.F., & Collins C.D.N., (2017). A review of seismicity in the Mukinbudin Bonnie Rock area of WA. In Proc. AEES 2017 Conference, Canberra ACT