Sie sind auf Seite 1von 11

Hypocenter and Magnitude Analysis

of Aftershocks of the 2018 Lombok,


Indonesia, Earthquakes Using Local
Seismographic Networks
Annisa Trisnia Sasmi*1, Andri Dian Nugraha2, Muzli Muzli3,4, Sri Widiyantoro2,5, Zulfakriza Zulfakriza2,
Shengji Wei3, David P. Sahara2, Agus Riyanto6, Nanang T. Puspito2, Awali Priyono2, Tim Greenfield7,
Haunan Afif8, Pepen Supendi1,9, Daryono Daryono4, Ardianto Ardianto1, Devy Kamil Syahbana8,
Yayan Mi’rojul Husni10, Billy S. Prabowo10, and Achmad Fajar Narotama Sarjan1

Abstract
The island of Lombok in Indonesia is located between the Indo-Australian and Eurasian
subduction trenches and the Flores back-arc thrust, making it vulnerable to earth-
quakes. On 29 July 2018, a significant earthquake Mw 6.4 shook this region and was
followed by series of major earthquakes (M w > 5:8) on 5, 9, and 19 August, which
led to severe damage in the northern Lombok area. In this study, we attempt to reveal
the possible cause of the sequences of the 2018 Lombok earthquakes based on after-
shock monitoring data. Twenty stations were deployed to record earthquake waveform
data from 4 August to 9 September 2018. In total, 3259 events were identified using
28,728 P- and 20,713 S-wave arrival times during the monitoring. The aftershock hypo-
centers were determined using a nonlinear approach and relocated using double-differ-
ence method. The moment magnitude (M w ) of each event was determined by fitting
the displacement spectrum amplitude using a Brune-type model. The magnitudes of the
Cite this article as Sasmi, A. T.,
aftershocks range from M w 1.7 to 6.7. The seismicity pattern reveals three clusters A. D. Nugraha, M. Muzli, S. Widiyantoro,
located in the Flores oceanic crust, which fit well with the occurrences of the four events Z. Zulfakriza, S. Wei, D. P. Sahara,
A. Riyanto, N. T. Puspito, A. Priyono, et al.
with M w > 6. We interpret these events as the main rupture area of the 2018 Lombok
(2020). Hypocenter and Magnitude
earthquake sequence. Furthermore, an aseismic zone in the vicinity of Rinjani extending Analysis of Aftershocks of the 2018
toward the northwestern part of Lombok was observed. We propose that the crust in Lombok, Indonesia, Earthquakes Using
Local Seismographic Networks, Seismol.
this area has elevated temperatures and is highly fractured thus inhibiting the gener- Res. Lett. XX, 1–11, doi: 10.1785/
ation of large earthquakes. The aseismic nature is therefore an artifact of the detection 0220190348.

threshold of our network (M w 4.6). Supplemental Material

Introduction the maximum intensity of VIII–IX modified Mercalli intensity


Lombok is situated in the southern part of central Indonesia, (MMI) in west, east, and north Lombok, and were strongly felt
directly adjacent to the Indian Ocean. Recently, a series of sig- by residents in the whole of Lombok. Because of this disaster,
nificant earthquakes shook Lombok starting from end of July The National Disaster Management Agency of Indonesia
2018 and caused damage to structures and killing hundreds
of people. This earthquake sequence began with an M w 6.4 1. Geophysical Engineering, Institut Teknologi Bandung, Bandung, Indonesia;
earthquake on 29 July 2018, followed by earthquakes on 5 2. Global Geophysics Research Group, Institut Teknologi Bandung, Bandung,
Indonesia; 3. Earth Observatory of Singapore/Asian School of the Environment,
August (M w 7.0), 9 August (M w 5.9), and 19 August (M w 6.3
Nanyang Technological University, Singapore; 4. Meteorology, Climatology, and
and M w 6.9). The Badan Meteorologi Klimatologi dan Geofisika Geophysics Agency (BMKG), Jakarta, Indonesia; 5. Universitas Kristen Maranatha
(BMKG; Meteorology, Climatology and Geophysics Agency, (Maranatha Christian University), Bandung, Indonesia; 6. Meteorology, Climatology,
and Geophysics Agency (BMKG), Mataram, Indonesia; 7. Bullard Laboratories,
Indonesia’s national body for reporting earthquakes) reported University of Cambridge, Cambridge, United Kingdom; 8. Center for Volcanology and
that the strong earthquake hypocenters were located to the Geological Hazard Mitigation (PVMBG), Bandung, Indonesia; 9. Meteorology,
Climatology, and Geophysics Agency (BMKG), Bandung, Indonesia; 10. Laboratory of
north of Lombok and to the south of Flores back-arc thrust Volcanology and Geothermal, Institut Teknologi Bandung, Bandung, Indonesia
(FBT). Robiana, Minarno, et al. (2018) explained that the *Corresponding author: trisnia24@gmail.com
earthquakes that occurred from 5 to 19 August 2018 reached © Seismological Society of America

Volume XX • Number XX • – 2020 • www.srl-online.org Seismological Research Letters 1

Downloaded from https://pubs.geoscienceworld.org/ssa/srl/article-pdf/doi/10.1785/0220190348/5049242/srl-2019348.1.pdf


by Univ Bergen Library user
(Badan Nasional
Penanggulangan Bencana
[BNPB]) reported 560 deaths,
80% of which were in the
North Lombok Regency
(BNPB, 2019).
There are two major geologi-
cal structures that mainly con-
trol the seismicity of Lombok:
a subducting megathrust
between Indo-Australian and
Eurasian plate to the south
(Hamilton, 1977, 1979), and the
FBT to its north (Hamilton,
1977, 1979; Silver et al., 1983).
The FBT extends in a near
east–west direction (Fig. 1).
Hamilton (1977, 1979)
suggested that this fault extends
from the north of central Flores
Island to the northern part of
central Sumbawa. Silver et al.
(1983), using seismic reflection
profiles, extends FBT farther
west toward the Bali basin.
Using Global Positioning
System (GPS) modeling,
Koulali et al. (2016) extends far-
ther FBT to the west until east
Java. The convergence rates of
FBT is estimated at around
5:6–6:0 mm=yr (Koulali et al.,
2016; Susilo et al., 2018). There
are also two strike-slip faults in
the west and east of Lombok,
namely the Lombok and
Sumbawa faults, respectively. In
the northern part of Lombok,
the currently active Rinjani vol-
canic complex is located. It’s the
most recent eruption occurred
in September 2016 (Global
Volcanism Program, 2016)
(2 yr prior to the Lombok
Figure 1. Map of study area; inset map is the Indonesian region. Blue inverted triangles show the 2018 earthquake). These geo-
local seismic stations in this study. Red triangles show volcanoes in the region. The black line logic and tectonic features make
depicts local active faults taken from Irsyam et al. (2017). Topography data were taken from Shuttle Lombok very vulnerable to the
Radar Topography Mission 30 Plus (Becker et al., 2009). Stars represent significant earthquakes seismic and volcanic hazard.
that occurred from 1979 to 2013, obtained from Supartoyo et al. (2014) and Badan Meteorologi
The last recorded major
Klimatologi dan Geofisika (BMKG, 2019), and 2018 significant earthquakes (BMKG, 2019). The
color of each star indicates the earthquake depth according to the color palette in the lower left events around Lombok are
corner. The contour dashed line represents slab depth around Lombok region (Hayes et al., 2018). dated to at least the nineteenth
The color version of this figure is available only in the electronic edition. century. On 25 July 1856, a de-
vastating earthquake occurred

2 Seismological Research Letters www.srl-online.org • Volume XX • Number XX • – 2020

Downloaded from https://pubs.geoscienceworld.org/ssa/srl/article-pdf/doi/10.1785/0220190348/5049242/srl-2019348.1.pdf


by Univ Bergen Library user
and caused a tsunami in the southern part of Lombok Figure 2. The three-component seismogram of the event that
(Wichmann, 1918). The BMKG reported that a major earth- occurred on 5 August 2018, at 18:36 UTC. From top to bottom:
quake with M w 6.7 occurred on 10 April 1978. In 1979, three the seismograms recorded at stations LM05, LM08, LM10, LM14,
and LM13. Red and blue lines indicate the selection of P- and S-
earthquakes were recorded on 21 May with M w 5.7 (BMKG,
wave arrival times. The color version of this figure is available only
2019), 30 May with M w 6.1 (Supartoyo et al., 2014; BMKG, in the electronic edition.
2019), and 20 October with M w 6.0 (BMKG, 2019), which had
quite a destructive effect. During the 2000s, large earthquakes
were recorded on 2 January 2004 (M w 6.2) and 22 June 2013 analysis of gravity and seismic active data. Using seismic velocity
(M w 5.4), as reported by Supartoyo et al. (2014). Most of the tomography, Widiyantoro et al. (2011) depicted the south-dipping
shallow events in northern Lombok had reverse mechanisms lithosphere beneath northern Lombok and Bali. Zubaidah et al.
(Widiyantoro and Fauzi, 2005; Ekström et al., 2012). This con- (2014) also confirmed this structure using a combination of gravity
dition is related to the existence of FBT, which is an important and magnetic surveys. However, the geometry of the FBT is still
manifestation of the subduction of the Indo-Australian plate not well defined.
with the Sunda block. The earthquake focus position of the We hypothesize that the 2018 Lombok earthquake sequence
1979–2018 Lombok earthquake can be seen in Figure 1. was caused either by the reactivation of the FBT or rupture of
Geophysical studies in the northern Lombok have been car- new faults located to the south of FBT. Taking advantage of the
ried out to image the FBT. McCaffrey and Nabelek (1987) newly deployed 20 temporary stations in the Lombok region
revealed the existence of Flores oceanic crust (FOC) to the starting 4 August 2018, we aim to reveal the probable sources
south of FBT that subducts beneath Lombok Island using joint of Lombok earthquakes using the accurate aftershock locations

Volume XX • Number XX • – 2020 • www.srl-online.org Seismological Research Letters 3

Downloaded from https://pubs.geoscienceworld.org/ssa/srl/article-pdf/doi/10.1785/0220190348/5049242/srl-2019348.1.pdf


by Univ Bergen Library user
significantly slower in the upper
20 km than the ak-135 model,
and dramatically reduces the
travel-time residuals.
Hypocenters are further refined
using the double-difference
approach method (Waldhauser
and Ellsworth, 2000) imple-
mented in the hypoDD pro-
gram (Waldhauser, 2001). This
method pairs earthquakes and
minimizes the travel-time dif-
ference residual to further
improve earthquake positions.
We use the 1D seismic velocity
model output from VELEST
to calculate earthquake hypo-
Figure 3. Wadati diagram for 3259 events. The V P =V S ratio value is 1.71. centers. The double-difference
method has been successfully
conducted to investigate earth-
quakes before in Indonesia, in-
and their M w distribution. The geometry of the FBT inferred cluding in Sumatra (Ramdhan and Nugraha, 2013), West
from the aftershock will be discussed. Furthermore, an aseis- Java (Supendi, Nugraha, Puspito, et al., 2018), Sulawesi
mic zone close to the Rinjani volcanic complex was also (Supendi, Nugraha, and Widiyantoro, 2018), and North
observed. We note that this is the first study of seismicity using Molucca (Nugraha et al., 2018).
local monitoring network on Lombok. To determine the moment magnitude of earthquakes, we
follow the processing steps outlined in Stork et al. (2014).
Data and Method The instrument response is removed from each trace and is
We deployed the local seismographic network soon after the converted from ground velocity to displacement. The data
first large event of the 2018 Lombok earthquake sequence that are then windowed around the P- and S-wave arrival times
occurred on 29 July (M w 6.4). The seismic waveform data used and detrended. The seismograms are then rotated to acquire
in this study were recorded by 20 seismographs from the P, SV, and SH components, with the incident angle and
Institut Teknologi Bandung, the Earth Observatory of Singapore/ back-azimuth values obtained through raytracing and the
Nanyang Technological University, and the Center of Volcanology shooting Snell method. In this case, the ray path of the P
and Geological Hazard Mitigation from 4 August to 9 September and S waves was assumed to be the same. The length of the
2018 (Fig. 1). We used the Seisgram2K (Lomax and Michelini, P-wave time window is half of the S and P travel-time
2009) to visualize the P- and S-wave arrival-time phases of each differences (t S − t P ), and the S-wave time window is 1.5 times
recorded event, as shown in Figure 2. We kept only those events of the length of the P-wave window. The maximum P-wave
recorded by at least three stations for further studies. time window is 3 s. This length is sufficient as the events have
The duration of deployment was for approximately 34 days epicentral distances of less than 50 km, and the time window
and captured several strong earthquakes, such as the 5 August must exclude later arriving phases. The signal was tapered
(M w 7.0), 9 August (M w 5.9), and 19 August (M w 6.3 and using a Hanning window to make the main energy of the wave
M w 6.9) events. We successfully identified 3259 local earth- positioned in the middle of the window.
quakes, using 28,728 P and 20,713 S wavepicks. A Wadati plot The scalar seismic moment can then be estimated using
(Fig. 3) is used to estimate the V P =V S ratio value (∼1:71). We information from ground displacement data in the frequency
then determined the initial location of these events using a non- domain, determined by the following equation (Havskov and
linear location method proposed by Lomax and Michelini Ottemöller, 2010):
(2009). The 1D seismic velocity model ak-135 (Kennett et al.,
1995) is used to initially generate earthquake hypocenters. We 4πρv3 rΩ0
M0  ; 1
use the initial locations as an input into VELEST (code in Rc
EQ-TARGET;temp:intralink-;df1;320;132

Fortran to invert simultaneously for 1D V P and V S and hypo-


central parameter ) (Kissling et al., 1994) and invert for the best- in which ρ is the density value near the source, v is the wave
fitting 1D velocity model. The new 1D model for Lombok is velocity around the source, and r is the length of the ray path.

4 Seismological Research Letters www.srl-online.org • Volume XX • Number XX • – 2020

Downloaded from https://pubs.geoscienceworld.org/ssa/srl/article-pdf/doi/10.1785/0220190348/5049242/srl-2019348.1.pdf


by Univ Bergen Library user
This value is obtained through
the raytracing. Rc is the radia-
tion pattern correction, and
Ω0 is the low-frequency level
of the ground displacement
amplitude spectrum. The low-
frequency spectral level was
estimated by fitting the spec-
trum amplitude with the Brune
(1970) model (Fig. 4), stated in
the following equation:

Ω0 e−πf t=Q
Ωf   ; 2
1  f =F c  2 
EQ-TARGET;temp:intralink-;df2;433;600

in which t is the travel time, f is


the frequency, F c is the corner
frequency, and Q represents
the quality factor.
On this occasion, the sys-
tematic search procedure was
performed to obtain the spec-
tral level at low frequency, cor-
ner frequency, and considering
the rock quality factor (Q).
From equation (2), it is clear
that as the Q-value increases,
the curvature decreases, and the
attenuation becomes smaller.
By completing this process,
the value of each parameter
could be determined more
accurately to calculate the seis-
mic moment. We then use the
resultant of SV and SH spectra
at low frequencies to represent
the S waves spectral level.
The seismic moments
Figure 4. (a) Position of LOM-06 and LOM-15 station, shown by the blue inverted triangles.
The yellow star depicts the location of the event, with origin time 6 September 2018 06:43:09.47
obtained from the P and S waves
UTC. (b) P and S wavepicks on rotated seismograms for the event depicted in (a). Red lines show were then used to calculate the
P- and S-wave arrival time. The gray line represents window length for the noise, and the black M w , using the equations given
line represents window length for the data. (c) Spectral fitting of the P-wave phase on the by Hanks and Kanamori (1979).
LOM-06 station. The black graphic shows the displacement spectrum amplitude. The blue line Data at each station are used to
shows the fitted Brune model. The gray line shows the spectrum of the noise. Vertical dashed
calculate M w value for each
line represents the frequency corner of each spectrum. (d) Spectral fitting of the S-wave phase in
LOM-06 station. (e) Spectral fitting of the P-wave phase in LOM-15 station. (f) Spectral fitting of the event. The final moment magni-
S-wave phase in LOM-15 station. Ω is the low-frequency spectral level, F c is the corner frequency, tude representing the size of an
Q is the quality factor, and Mw is the resulted moment magnitude. Subscript of P and S distinguish event was acquired by averaging
the parameters for P- and S-wave phases. The color version of this figure is available only in the the magnitude estimations from
electronic edition. P and S waves. The M w values
from all the stations were aver-
aged to acquire the M w value
from an event.

Volume XX • Number XX • – 2020 • www.srl-online.org Seismological Research Letters 5

Downloaded from https://pubs.geoscienceworld.org/ssa/srl/article-pdf/doi/10.1785/0220190348/5049242/srl-2019348.1.pdf


by Univ Bergen Library user
Hypocenter
Distribution
Analysis
We successfully determined the
location of 3259 events, using
28,728 P- and 20,713 S-wave
arrival times. These results
reveal that earthquakes are pre-
dominantly located to the north
of Lombok and are distributed
in an east–west orientation. We
also determined the location of
the large events of 5 August
(M w 7), 9 August (M w 5.9), and
19 August (M w 6.3 and
M w 6.9), as shown in
Figure 5. These large earth-
quakes were generally located
beneath the northern Lombok.
Using hypoDD, we successfully
relocated 3085 out of 3259
events (Fig. 6) and compared
the relocated results with their
initial locations, as shown in
Figure 6b. The residual travel
times after relocation shows sig-
nificant improvement (Fig. 7),
indicating that the hypocenters
of the earthquakes had been
optimized. The locations of
the aftershocks were primarily
distributed to the north of
Lombok, adjacent to the FBT.
Our results show that the depth
of the events is concentrated
between 5 and 25 km. The M w
of the aftershocks varied from
1.7 to 6.7 (Fig. 6a).
The four large-magnitude
earthquakes in Lombok, de-
picted in the cross section,
were shallow earthquakes with
depths less than 35 km. Based
on the aftershock locations, we
suggest that the three major
Figure 5. (a) Map of the initial earthquake locations beneath Lombok and the surrounding area.
Colored dots depict the distribution of the epicenters. Yellow stars signify strong earthquakes clusters are distributed off-
obtained from the seismograph data. Yellow diamond shows the position of 29 July Mw 6.4 shore to the north of Lombok,
earthquake from the BMKG earthquake catalog. Blue inverted triangles show the local seismic close to the FBT trench
stations. Each significant earthquake has a focal mechanism that is included in the cross section, (Fig. 6a). In general, the hypo-
with data from Global Centroid Moment Tensor (CMT; except earthquake Mw 6.3 [8/19] which is
centers delineate a southward-
not listed in Global CMT catalog). Red triangles depict Mt. Rinjani, and the black line depicts local
active faults and strike-slip faults as taken from Irsyam et al. (2017). Blue boxes signify the location dipping fault geometry with a
of the cross section. (b) Vertical cross section of the Lombok earthquakes; the locations of the cross small dip (30°), as depicted in
sections are shown in (a). The color version of this figure is available only in the electronic edition. Figure 6b. We show that the

6 Seismological Research Letters www.srl-online.org • Volume XX • Number XX • – 2020

Downloaded from https://pubs.geoscienceworld.org/ssa/srl/article-pdf/doi/10.1785/0220190348/5049242/srl-2019348.1.pdf


by Univ Bergen Library user
We interpret the large earth-
quakes on 5, 9, and 19 August to
be generated by movement
between the FOC subducting
plate and overriding island of
Lombok. This interpretation
also corresponds with the defor-
mation study that was done by
Susilo et al. (2018). Based on
GPS observations, Susilo et al.
(2018) found a clear coseismic
deformation as a result of the
M w 7.0 Lombok earthquake on
5 August. Based on these results,
it was possible to detect the
deformations to the north of
Lombok. This northward defor-
mation is a result of the relative
movement of the Indo-
Australian plate and generated
uplift of up to 70 cm
(Agustan et al., 2019). This
shows that the earthquake activ-
ity in the north of the Lombok
region is closely related to the
uplift movement. Furthermore,
the focal mechanism solution
from the Global Centroid
Moment Tensor (Fig. 6a) rein-
forces the fact that the strike
of the Lombok earthquakes
generally oriented in an east–
west direction (parallel to the
FBT). However, the events are
distributed farther south of the
FBT position (Fig. 6b) and show
Figure 6. (a) Relocated Lombok earthquakes calculated using the double-difference method. The a pattern that can be identified
Mw values had been determined, denoted with filled colored circles. Yellow stars signify the strong as thrust-fault appearance in
earthquakes from the seismograph data. Yellow diamond shows the position of 29 July Mw 6.4 front of the back-arc thrust.
earthquake from BMKG earthquake catalog. Blue inverted triangles are the locations of local
To analyze our results, we
seismic stations. The focal mechanisms are taken from the Global CMT catalog. Black lines denote
active faults (Irsyam et al., 2017). The imaged aseismic zone is delineated by the black, dashed plot the earthquakes during
ellipse. Black dashed rectangles indicate the distribution of earthquake clusters. (b) Vertical section three periods: between the 5
of line B through the 5 August event, taken from Figure 5a. The hypocenter cluster distribution and the 9 August 2018 events,
shows a pattern that can be identified as the thrust-fault appearance in front of the back-arc thrust between the 9 and the 19
(black dashed line). The color version of this figure is available only in the electronic edition. August 2018 events, and after
the 19 August 2018 to 9
September 2018 events. The
three mainshocks trigger after-
5 August M w 7.0 earthquake aftershocks were primarily concen- shocks close to their epicenters. The three earthquake clusters
trated in front of the FBT uplift region (Fig. 6b). It is likely that are shown in Figure 6a. The temporal and spatial clustering
this cluster indicates a rupture area on the FBT. Our interpreta- of the seismicity suggests that the FOC to the north of
tion is that the events related to the subduction system of the Lombok is segmented. This is similar to the results of a geodetic
FOC beneath Lombok are with a back-arc thrusting mechanism. model proposed by Wang et al. (2020). We propose the failure

Volume XX • Number XX • – 2020 • www.srl-online.org Seismological Research Letters 7

Downloaded from https://pubs.geoscienceworld.org/ssa/srl/article-pdf/doi/10.1785/0220190348/5049242/srl-2019348.1.pdf


by Univ Bergen Library user
of 1. The completion magni-
tude (M c , a measure of the
detection threshold of the local
network) is 4.6. We calculate
this value using a b-value sta-
bility approach (Greenfield
et al., 2018) for which the
average b-value in a forward
looking, five bin width wide
window is the same (within
error) as the calculated b-value
at a specific trial M c . This is
typically a higher M c than cal-
culated using the more stan-
dard maximum curvature
approach, which for this cata-
log is closer to 4, but the
Figure 7. Comparison residual of travel times (a) before and (b) after relocation, using the double- b-value is significantly more
difference method. robust. In this catalog, there
is a kink in the cumulative dis-
tribution around a magnitude
of 4.6, which is picked out by
of each successive segment loads the next segment in the the stability approach. Such a kink is most likely due to the
sequence, triggering a large earthquake shortly afterward. The errors in calculating M w increasing for smaller earthquakes.
initial event in this sequence would be the M w 6.4 earthquake This could generate artifacts in the distribution.
on 29 July 2018. A similar sequence of earthquakes occurred in An earthquake magnitude represents the size of the earth-
1954 during the rupture of the Dixie Valley–Pleasant Valley, quake. However, the impact of an earthquake does not only
Nevada fault system (Zhang et al., 1991). depends on the magnitude but also on the earthquake focal
depth, geological site effect, and the epicenter distance.
Magnitude Determination Results These impacts are qualitatively stated in the MMI index in
We calculate the M w of each earthquake using the spectral which the values range from I to XII. The higher MMI scale
fitting approach outlined earlier (Fig. 4). Of the 3085 events represents a more severe impact from an earthquake that
relocated using hypoDD, 2620 could have their moment mag- occurs in a region. The Lombok earthquake reached intensity
nitude determined. The fitted values are dependent on the sig- of VIII–IX (Robiana, Minarno, et al., 2018) and can be catego-
nal-to-noise ratio (SNR) of the event. Figure 4c,d shows the fits rized as a severe earthquake. The strong ground movement in
to high SNR stations, whereas Figure 4e,f shows the fit to low Lombok was likely driven a combination of the soft sedimen-
SNR stations. It is clear that F c and Q are sensitive to the noise, tary and volcanic units found on Lombok (Mangga et al.,
whereas the calculated value of M w (thus Ω0 ) is less affected. 1994), which act to increase the amplitude of earthquake
The robustness of our calculated M w is further confirmed by waves; and the shallow depth (5–25 km), large magnitude
the recovered M w for the example earthquake shown in (M w > 5:8), and the relatively small epicentral distance
Figure 4. It is the same (within error) as M w calculated by (<50 km) of the earthquakes.
the U.S. Geological Survey (5.0) and the BMKG (4.9). In gen-
eral, for three of the mainshocks in this study (06:25 UTC 9 The Occurrence of an Aseismic Zone
August, 04:10 UTC 19 August, and 02:56 UTC 19 August), A prominent relatively aseismic zone is found to the northwest
our M w is within 0.2 magnitude units of the published BMKG of Lombok (dashed line, Fig. 6a). Aseismic zones can be
magnitudes. However, we were not able to calculate the mag- formed by a fracture zone (i.e., Jordahl et al., 1995), unconsoli-
nitude of the 5 August event; the stations that were deployed at dated sediments (i.e., Byrne et al., 1988), a large area of coseis-
that point were all clipped and were unsuitable for the spectral mic slip, or the upflow of partial molten material (i.e., Hanus
analysis. and Vanek, 1979). In northern Lombok, we suspect that the
The earthquake magnitude distribution is plotted in cause of the aseismic zone is the presence of elevated temper-
Figure 8. Magnitudes range from 1.7 to 6.7 and fit the atures and an extensive fracture system related to volcanism on
Gutenberg–Richter distribution well (Fig. 8) with a b-value Lombok. The aseismic region is close to the extinct Samalas
of 1:13  0:09, close to the global average and expected value volcano (Vidal et al., 2016) and on the western flanks of

8 Seismological Research Letters www.srl-online.org • Volume XX • Number XX • – 2020

Downloaded from https://pubs.geoscienceworld.org/ssa/srl/article-pdf/doi/10.1785/0220190348/5049242/srl-2019348.1.pdf


by Univ Bergen Library user
threshold in this region to be locally higher than the regional
average.

Conclusions
We have successfully determined and relocated the aftershocks
of the 2018 Lombok earthquakes using a local seismographic
network. The hypocenters were generally distributed to the
north of Lombok, with an average depth of ∼20 km. Overall,
the strong earthquakes and aftershocks that occurred in
Lombok from 5 August to 9 September 2018 had a depth
of less than 35 km. From the distribution of hypocenters,
we have discovered that the 2018 Lombok earthquakes mostly
occurred as a result of relative FOC movement to the south,
generating a series of earthquake aftershocks with M w 1.7–
6.7. Three temporal and spatial clusters are observed, which
image the segmentation of the FOC in this region and limit
the maximum magnitude of any earthquake along this fault
(assuming that multiple segments cannot be simultaneously
triggered). A clear aseismic zone northwest of Lombok is
imaged, which may be related to a low velocity, thus high-
temperature region associated with volcanism on Lombok.

Data and Resources


Figure 8. Earthquake magnitude distribution plotted cumulatively All figures presented were plotted using Generic Mapping Tools
(black crosses) and as a histogram (red triangles). The magnitude (Wessel and Smith, 1998). The focal mechanism data were obtained
of completion (Mc ) is indicated by the open inverted triangle and from the Global Centroid Moment Tensor (Global CMT) catalog
the maximum-likelihood fit to the Gutenberg–Richter distribution
(Dziewonski et al., 1981). The local fault data (e.g., thrust fault and
is delineated by the blue line. The color version of this figure is
strike-slip fault) plotted in the figures were taken from Irsyam et al.
available only in the electronic edition.
(2017). Topography data used in the figures were acquired from
Shuttle Radar Topography Mission (SRTM) 30 Plus (Becker et al.,
2009). Hypocenters used in this study are tabulated in the supplemen-
tal material. The unpublished manuscript by H. Afif, A. D. Nugraha,
the currently active Rinjani complex. This suggests a relation- M. Muzli, S. Widiyantoro, Z. Zulfakriza, S. Wei, D. P. Sahara, A.
ship between the volcanism and the aseismic zone. Riyanto, N. T. Puspito, A. Priyono, et al. (2019), “Tomography of
A recent seismic tomography study (H. Afif et al., unpub- the source area of the 2018 Lombok earthquake, Indonesia, from local
lished manuscript, 2019; see Data and Resources) images a low seismographic network,” submitted to Geophys. J. Int.
V P and V S anomaly in the same location as the aseismic vol-
ume. Low-seismic velocities are indicative of high tempera- Acknowledgments
tures, the presence of fluids, and an increase in the amount The authors would like to thank their colleagues at Institut
of fracturing. Together, we would expect an increase in the Teknologi Bandung (ITB), the Earth Observatory Singapore (EOS),
amount of seismicity in such a region because of the low nor- the Agency for Meteorology, Climatology, and Geophysics of
mal stresses across fractures driven by high pore-fluid pres- Indonesia (Badan Meteorologi Klimatologi dan Geofisika
sures. However, the earthquakes triggered in such a region [BMKG]), and the Center for Volcanology and Geological Hazard
are usually small because the complex fracture network inhibits Mitigation of Indonesia (PVMBG) for their collaboration. They also
the creation of longer faults, which in turn generate large- express their gratitude for the Pendidikan Magister Menuju Doktor
Untuk Sarjana Unggul (PMDSU) Batch IV scholarship program
magnitude earthquakes. Typical earthquake magnitudes in vol-
(2018) established by the Indonesian Ministry of Research,
canic regions are less than 3 (Greenfield et al., 2018), and no
Technology, and Higher Education and for the support and research
significant magnitude increases are observed when an external opportunities given to A. T. Sasmi, Kementrian Riset dan Teknologi /
stress is applied (Green et al., 2015). We therefore suggest that Badan Riset dan Inovasi Nasional (KEMRISTEK/BRIN) Indonesia
the aseismic hole (Fig. 6) is in fact seismically active, but the research funding 2019 awarded to A. D. Nugraha and N. T.
triggered seismicity is below the detection threshold of our Puspito, and the World Class Research funding to S. Widiyantoro.
network (M w ∼ 4:6, Fig. 8). We note that the aseismic hole Our thanks to R. Ryannugraha, A. D. Sasongko, F. T.
coincides with a significant gap in the seismograph network Dasahruddyn, G. S. Fatih, and A. W. Baskara for their help during
deployed in Lombok (Fig. 6), which likely causes the detection the data preparation process.

Volume XX • Number XX • – 2020 • www.srl-online.org Seismological Research Letters 9

Downloaded from https://pubs.geoscienceworld.org/ssa/srl/article-pdf/doi/10.1785/0220190348/5049242/srl-2019348.1.pdf


by Univ Bergen Library user
References Irsyam, M., S. Widiyantoro, D. H. Natawidjaya, I. Meilano, A.
Agustan, , R. N. Hanifa, Y. Anantasena, M. Sadly, and T. Ito (2019). Rudyanto, S. Hidayati, W. Triyoso, N. R. Hanifa, D. Djarwadi,
Ground deformation identification related to 2018 Lombok earth- L. Faizal, et al. (2017). Peta sumber dan bahaya gempa
quake series based on Sentinel-1 data, Earth Environ. Sci. 280, Indonesia tahun 2017, Pusat Penelitian dan Pengembangan
012004, doi: 10.1088/1755-1315/280/1/012004. Perumahan dan Permukiman, Kementerian Pekerjaan Umum
Badan Meteorologi Klimatologi dan Geofisika (BMKG) (2019). dan Perumahan Rakyat (in Indonesian).
Katalog Gempa Bumi Signifikan dan Merusak Tahun 1821–2018, Jordahl, K. A., M. K. McNutt, H. F. Webb, S. Kruse, and M.
Pusat Gempa dan Tsunami BMKG, Retrieved from https://www Kuykendall (1995). Why there are no earthquakes on the Marquesas
.bmkg.go.id/gempabumi/katalog-gempabumi-signifikan.bmkg Fracture Zone, J. Geophys. Res. 100, no. B12, 24,431–24,447.
(last accessed January 2019) (in Indonesian). Kennett, B. L. N., E. R. Engdahl, and R. Buland (1995). Constraints on
Badan Nasional Penanggulangan Bencana (BNPB) (2019). Gempa seismic velocities in the Earth from traveltimes, Geophys. J. Int.
Lombok, available at https://bnpb.go.id/uploads/24/info-bencana 122, 108–124.
-agustus-2018-1.pdf (last accessed June 2019). Kissling, E., W. L. Ellsworth, D. Eberhart-Phillips, and U. Kradolfer
Becker, J. J., D. T. Sandwell, W. H. F. Smith, J. Braud, B. Binder, J. (1994). Initial reference models in local earthquake tomography,
Depner, D. Fabre, J. Factor, S. Ingalls, S.-H. Kim, et al. (2009). J. Geophys. Res. 99, 19,635–19,646.
Global bathymetry and elevation data at 30 arc seconds resolution: Koulali, A., S. Susilo, S. McClusky, I. Meilano, P. Cummins, P.
SRTM30_PLUS, Mar. Geodes. 32, no. 4, 355–371. Tregoning, G. Lister, J. Efendi, and M. A. Syafi’i (2016). Crustal
Brune, J. N. (1970). Tectonic stress and the spectra of seismic shear strain partitioning and the associated earthquake hazard in the
waves from earthquakes, J. Geophys. Res. 75, 4997–5009. eastern Sunda-Banda Arc, Geophys. Res. Lett. 43, 1943–1949,
Byrne, D. E., D. M. Davis, and L. R. Sykes (1988). Loci and maximum doi: 10.1002/2016GL067941.
size of the thrust earthquakes and the mechanics of the shallow Lomax, A., and A. Michelini (2009). M wpd : A duration-amplitude
region of subduction zones, Tectonics 7, 833–857. procedure for rapid determination of earthquake magnitude and
Dziewonski, A. M., T.-A. Chou, and J. H. Woodhouse (1981). tsunamigenic potential from P waveforms, Geophys. J. Int. 176,
Determination of earthquake source parameters from waveform 200–214, doi: 10.1111/j.1365-246X.2008.03974.x.
data for studies of global and regional seismicity, J. Geophys. Res. Mangga, S. A., S. Atmawinata, B. Hermanto, and T. C. Amin (1994).
86, 2825–2852, doi: 10.1029/JB086iB04p02825. Geologi Lembar Lombok, Nusatenggara, Lembar 1807, Pusat
Ekström, G., M. Nettles, and A. M. Dziewonski (2012). The global Penelitian dan Pengembangan Geologi (in Indonesian).
CMT project 2004–2010: Centroid-moment tensors for 13,017 McCaffrey, R., and J. Nabelek (1987). Earthquakes, gravity, and the
earthquakes, Phys. Earth Planet. In. 1/9, 200–201. origin of the Bali Basin: An example of a nascent continental
Global Volcanism Program (2016). Report on Rinjani (Indonesia), fold-and- thrust belt, J. Geophys. Res. 92, no. B1, 441–460, doi:
in Weekly Volcanic Activity Report, 21 September-27 September 10.1029/JB092iB01p00441.
2016, S. K. Sennert (Editor), Vol. 41, Smithsonian Institution and Nugraha, A. D., P. Supendi, S. Widiyantoro, D. Daryono, and S.
U.S. Geol. Surv, 8 pp., doi: 10.5479/si.GVP.BGVN201608-264030. Wiyono (2018). Hypocenter relocation of earthquake swarm
Green, R. G., T. Greenfield, and R. S. White (2015). Triggered earth- around Jailolo volcano, North Molucca, Indonesia using the
quakes suppressed by an evolving stress shadow from a propagat- BMKG network data: Time periods of September 27-October
ing dyke, Nature Geosci. 8, 629–632, doi: 10.1038/ngeo2491. 10, 2017, AIP Conf. Proc., Vol. 1987, 020093, doi: 10.1063/
Greenfield, T., R. S. White, T. Winder, and T. Ágústsdóttir (2018). 1.5047378.
Seismicity of the Askja and Bárðarbunga volcanic systems of Ramdhan, M., and A. D. Nugraha (2013). Study of seismicity around
Iceland, 2009-2015, J. Volcanol. Geoth. Res. 391, 106432, doi: Toba area based on relocation hypocenter result from BMKG cata-
10.1016/j.jvolgeores.2018.08.010. logue, AIP Conf. Proc., Vol. 1554, 242, doi: 10.1063/1.4820330.
Hamilton, W. B. (1977). Subduction in the Indonesian region, in Robiana, R., P. A. Minarno, S. Hidayati, S. Supartoyo, and F. Nurfalah
Island Arcs, Deep Sea Trenches and Back-Arc Basins, M. (2018). Gempa Lombok Tanggal 29 Juli 2018 dan Dampaknya—
Talwani and W. C. Pitman III (Editors), Vol. 1, Maurice Ewing Kajian Rangkaian Gempa Lombok Provinsi Nusa Tenggara Barat,
Series, Am. Geophys. Union, 15–31. Pusat Penelitian dan Pengembangan Perumahan dan Pemukiman,
Hamilton, W. B. (1979). Tectonics of the Indonesian region, Tech. Badan Penelitian dan Pengembangan, Kementerian Pekerjaan
Report 1078, U.S. Govt. Print. Off., Washington, D.C. Umum dan Perumahan Rakyat (in Indonesian).
Hanks, T. C., and H. Kanamori (1979). A moment magnitude scale, Silver, E. A., D. Reed, R. McCaffrey, and Y. Joyodiwiryo (1983).
J. Geophys. Res. 84, 2348–2350. Backarc thrusting in the Eastern Sunda Arc, Indonesia: A conse-
Hanus, V., and J. Vanek (1979). Morphology and volcanism of the quence of arc-continent collision, J. Geophys. Res. 88, 7429–7448.
Wadati-Benioff zone in the Tonga-Kermadec system of recent Stork, A. L., J. P. Verdon, and J. M. Kendall (2014). The robustness of
subduction, New Zeal. J. Geol. Geophys. 22, 6, 659–671. seismic moment and magnitudes estimated using spectral analysis,
Havskov, J., and L. Ottemöller (2010). Routine Data Processing in Geophys. Prosp. 62, 862–878.
Earthquake Seismology, Springer, Dordrecht, The Netherlands, 9 pp. Supartoyo, S., Surono, and E. Tofani (2014). Katalog Gempa Merusak
Hayes, G. P., G. L. Moore, D. E. Portner, M. Hearne, H. Flamme, M. Indonesia Tahun 1612–2014 (Edisi Kelima), Pusat Vulkanologi
Furtney, and G. M. Smoczyk (2018). Slab2, a comprehensive sub- dan Mitigasi Bencana Geologi, Badan Geologi, (in Indonesian).
duction zone geometry model, Science 362, 58–61, doi: 10.1126/ Supendi, P., A. D. Nugraha, N. T. Puspito, S. Widiyantoro, and D.
science.aat4723. Daryono (2018). Identification of active faults in West Java,

10 Seismological Research Letters www.srl-online.org • Volume XX • Number XX • – 2020

Downloaded from https://pubs.geoscienceworld.org/ssa/srl/article-pdf/doi/10.1785/0220190348/5049242/srl-2019348.1.pdf


by Univ Bergen Library user
Indonesia, based on earthquake hypocenter determination, reloca- Wessel, P., and W. H. F. Smith (1998). New, improved version of
tion, and focal mechanism analysis, Geosci. Lett. 5, 31. Generic Mapping Tools released, Eos Trans. AGU 79, 579.
Supendi, P., A. D. Nugraha, and S. Widiyantoro (2018). Hypocenter Wichmann, A. (1918). The earthquakes of the Indian Archipelago
relocation of the aftershocks of the Poso, Sulawesi (Mw 6.6, until the year 1857, Natuurkunde 2, Sect. Deel 20, No. 4 and
May 29, 2017) event using the BMKG network data, AIP Conf. Deel 22, No. 5.
Proc., Vol. 1987, 020076. Widiyantoro, S., and Fauzi (2005). Note on seismicity of The Bali
Susilo, S., A. Wijanarto, S. Wibowo, and H. Z. Abidin (2018). GPS/ Convergen region in the eastern Sunda Arc-Indonesia, Aus. J.
GNSS analysis on Lombok earthquakes: Co-seismic deformation, Earth Sci. 52, 379–383.
Experimental Findings. doi: 10.13140/RG.2.2.34871.78242. Widiyantoro, S., J. D. Pesicek, and C. H. Thurber (2011). Subducting
Vidal, C. M., N. Métrich, J. Komorowski, I. Pratomo, A. Michel, N. slab structure below the eastern Sunda arc inferred from non-lin-
Kartadinata, V. Robert, and F. Lavigne (2016). The 1257 Samalas ear tomography imaging, in The SE Asian Gateway: History and
eruption (Lombok, Indonesia): The single greatest stratospheric Tectonics of the Australian-Asia Collision (Special Publication),
gas release of the Common Era, Sci. Rep. 6, 34868. Geological Society of London, 139–155.
Waldhauser, F. (2001). HypoDD—A program to compute double- Zhang, P., D. B. Slemmons, and F. Mao (1991). Geometric pattern,
difference hypocenter locations, U.S. Geol. Surv. Open-File Rept. rupture termination and fault segmentation of the Dixie Valley-
2001-113. Pleasant Valley active normal fault system, Nevada, U.S.A, J.
Waldhauser, F., and W. L. Ellsworth (2000). A double-difference Struct. Geol. 13, no. 2, 165–176.
earthquake location algorithm: Method and application to the Zubaidah, T., M. Korte, M. Mandea, and M. Hamoudi (2014). New
Northern Hayward Fault, California, Bull. Seismol. Soc. Am. 90, insight into regional tectonics of the Sunda-Banca Arcs region
1353–1368, doi: 10.1785/0120000006. from integrated magnetic and gravity modelling, J. Asia Earth
Wang, C., X. Wang, W. Xiu, B. Zhang, G. Zhang, and P. Liu (2020). Sci. 80, 172–184.
Characteristics of the seismogenic faults in the 2018 Lombok,
Indonesia, earthquake sequence as revealed by inversion of InSAR
measurements, Seismol. Res. Let. 91, 733–744, doi: 10.1785/ Manuscript received 14 November 2019
0220190002. Published online 27 May 2020

Volume XX • Number XX • – 2020 • www.srl-online.org Seismological Research Letters 11

Downloaded from https://pubs.geoscienceworld.org/ssa/srl/article-pdf/doi/10.1785/0220190348/5049242/srl-2019348.1.pdf


by Univ Bergen Library user

Das könnte Ihnen auch gefallen