See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/227285730

Geomagnetic field anomalies over the Lombok Island region: An attempt to

understand the local tectonic changes

Article  in  International Journal of Earth Sciences · July 2010

DOI: 10.1007/s00531-009-0450-4

CITATIONS READS

13 790

5 authors, including:

Teti Zubaidah Monika Korte

University of Mataram Helmholtz-Zentrum Potsdam - Deutsches GeoForschungsZentrum GFZ
39 PUBLICATIONS   93 CITATIONS    137 PUBLICATIONS   4,204 CITATIONS   

SEE PROFILE SEE PROFILE

Mioara Mandea Bulkis Kanata

Centre National d’Etudes Spatiales University of Mataram
290 PUBLICATIONS   6,983 CITATIONS    40 PUBLICATIONS   84 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Global long-term geomagnetic field reconstruction View project

ForM@Ter - French Solid Earth' services and data pole View project

All content following this page was uploaded by Monika Korte on 20 May 2014.

The user has requested enhancement of the downloaded file.

 Originally published as:

Zubaidah, T., Korte, M., Mandea, M., Quesnel, Y., Kanata, B., Arumdati, N. (2010):
Geomagnetic field anomalies over the Lombok Island region: an attempt to understand the
local tectonic changes. - International Journal of Earth Sciences, 99, 5, 1123-1132

DOI: 10.1007/s00531-009-0450-4
Geomagnetic Field Anomalies over the
Lombok Island Region: An Attempt to
Understand the Local Tectonic Changes

T. Zubaidah
1- Helmholtz Centre Potsdam, Deutsches GeoForschungsZentrum (GFZ), Sektion
2.3 Erdmagnetfeld, Potsdam, Germany
2- Jurusan Teknik Elektro, Fakultas Teknik Universitas Mataram, Indonesia

M. Korte
Helmholtz Centre Potsdam, Deutsches GeoForschungsZentrum (GFZ), Sektion
2.3 Erdmagnetfeld, Potsdam, Germany

M. Mandea
Helmholtz Centre Potsdam, Deutsches GeoForschungsZentrum (GFZ), Sektion
2.3 Erdmagnetfeld, Potsdam, Germany

Y. Quesnel
Helmholtz Centre Potsdam, Deutsches GeoForschungsZentrum (GFZ), Sektion
2.3 Erdmagnetfeld, Potsdam, Germany
Now at
Centre Européen de Recherche et d’Enseignement des Géosciences de
l’Environnement (CEREGE), France

B. Kanata
Jurusan Teknik Elektro, Fakultas Teknik Universitas Mataram, Indonesia

Correspondences should be addressed to:

Teti Zubaidah
Helmholtz Centre Potsdam, Deutsches GeoForschungsZentrum (GFZ), Sektion
2.3 Erdmagnetfeld, Raum F-420, Telegrafenberg, 14473 Potsdam, Germany
Tel.: +49-331-288-1233, Fax.: +49-331-288-1235, E-mail: [email protected]

"The original publication is available at www.springerlink.com".

Citation: Zubaidah, T., Korte, M., Mandea, M., Quesnel, Y., Kanata, B. (2009 online first)
Geomagnetic field anomalies over the Lombok Island region: an attempt to understand the local
tectonic changes. International Journal of Earth Sciences, doi: 10.1007/s00531-009-0450-4.

1
Abstract During the last years, several investigations on the earthquakes and related
tectonic structures along the Java trench have been conducted. In this study, we focus on the
Lombok Island – West Nusa Tenggara (Indonesia), which lies between the centres of the highest
intensity of contiguous negative-positive geomagnetic anomalies in this region. The geological and
tectonic structures, however, are not known in detail for this island, whereas a better knowledge of
these conditions could lead to better estimate the potential earthquake risks and thus a possible
improvement of the existing early warning system.
We have performed a ground-based geomagnetic survey at 56 stations in the Lombok Island
during October–November 2006 for a detailed mapping and interpretation of geomagnetic
anomalies related to the geological and tectonic characteristics. The 2D and 3D magnetic maps
show a general geomagnetic anomaly pattern in the Lombok Island which consists of repeated
contiguous negative-positive anomalies. Two forward models have been proposed for a profile
connecting a strongest apparent dipolar structure. The first model assumes a uniformly magnetized
sphere as the source of the anomaly, and could be interpreted as a specific local structure
composed by a quite large magnetic body. The second model considers several rocks with
different susceptibilities and magnetizations, and could be interpreted as a discontinuity in the
geological structures. This model agrees with the local geological surface conditions and the
known large scale regional tectonic structure. Therefore, it is used to interpret our results in terms
of tectonic characteristics, which suggests the potential existence of a new tectonic element (e.g. a
local normal fault) in this region.

Key words: Geomagnetic field anomaly, Tectonic earthquake, Java Trench,

Lombok Island

1. Introduction
After the 2004 Sumatra earthquake, needs and efforts to assess seismic hazards in
the Indonesian region have strongly increased. Several geophysical methods have
been proposed to study this region, one of them being the potential field analysis.
The relation between the geomagnetic field – in which we are interested – and
earthquakes is not obvious, but several recent studies have suggested that
magnetic anomalies could help to better understand the geological and tectonic
conditions related to seismic activity in the subduction zones.
Purucker and Ishihara (2005) have reported that the present Java subduction zone
is evident as a band of negative anomalies situated south and west of Sumatra and
Java Islands. Blakely et al. (2005) have suggested that serpentinized magnetic
mantle may be common in forearc settings and thus magnetic anomalies may be
useful in mapping hydrated mantle in convergent margins. Some studies have
indicated that the intense seismic activity could be associated with the subduction
in the southern and eastern Indonesian regions. Hinschberger et al. (2005) have
proposed a new kinematic model of eastern Indonesia based on a synthesis of
geophysical, geological and geochemical studies: the eastern Indonesian region
clearly appears as a very active area where the deformation associated to the
AUS/PSP/SEA (Australian/Philippine Sea/South-East Asia) triple junction zone is
widely distributed and rapidly evoluting. Špičák et al. (2007) have proposed the
beginning of a new subduction cycle along the Java Trench, in which the position
of Java and Lombok Troughs may be interpreted as a “fossil” trench for the onset
of the recently terminating subduction cycle.
In a different direction, piezomagnetic effects have also been considered in
relation to earthquakes. Nishida et al. (2007) have explored such effects to analyze
precursory and co-seismic signals. While they could not detect significant signals

2
for the two recent Japanese earthquakes (the 2003 Tokachi-oki of M 8.0 and the
2004 Kushiro-oki of M 7.1, in Hokkaido), they predict effects of several nT for
strong earthquakes expected along the Kurile Trench, a region which is also
characterized by large amplitudes of geomagnetic anomalies.
Finally, let us note that Balasis and Mandea (2007) have applied wavelet analysis
to investigate if electromagnetic disturbances related to the recent great
earthquakes could be detected by satellite magnetometers. They have concluded
that only a statistical study based on large earthquakes recorded during the
CHAMP magnetic mission could bring an answer to such a crucial question. The
need for more magnetic measurements in this region is clearly supported by all
studies indicated before.
In the following, we focus on the available data over the Lombok Island, the
region of interest, from ground to satellite platforms. Let us note that the existence
of a large scale contiguous negative-positive high intensity geomagnetic anomaly
pattern along the Java trench has been reported in several occasions. Figure 1
shows the total component of the magnetic anomaly over Indonesian region using
the magnetic field model MF5 (Maus, et al. 2006) at 5 km altitude. Figure 2
shows a more detailed magnetic anomaly map over the islands along the Java
Trench, including the Lombok Island. This map has been generated from cleaned
and levelled marine magnetic data available from GEODAS for the interval 1950
- 2004 (Quesnel et al. 2009), combined with marine magnetic data along the
Sunda-Banda Arc transition of BGR (Müller and Neben 2006) and aeromagnetic
data of AIST and CCOP (Ishihara and Kisimaoto 2002). Figure 1 indicates that
the Lombok Island is located between the centres of the highest negative-positive
magnetic anomalies in this region, but the detailed map of Fig. 2 can not
absolutely support this view, as ground and marine magnetic data are not
completely available. Although some preliminary studies were conducted
(Zubaidah et al. 2005), no high resolution geomagnetic data are available for the
Lombok Island. Due to the lack of state-of-the-art instruments, the quality of some
of the first magnetic measurements is questionable for more interpretations.
Our aim is to obtain a comprehensive view of the local conditions of the region by
applying some geophysical methods systematically, in order to map and thereafter
to interpret in detail the geomagnetic anomalies related to the geological and
tectonic structures. Here, we only report on the first results of regional
geomagnetic investigations, from new measurements in the field to preliminary
interpretations.

2. Geomagnetic survey and data processing

New geomagnetic field measurements on the Lombok Island were done during
October–November 2006, by conducting a ground-based magnetic survey of total
field intensity at 56 stations, with an average distance between adjacent stations
around 2.5 km, as shown in Fig. 3. This survey included measurements at 36 old
stations situated in the northern area, which were previously occupied (Zubaidah
et al. 2005). Several stations had to be re-located to magnetically more quiet
neighbouring places, in order to ensure a higher quality of these measurements.
The purpose of the repeated survey is to investigate possible local geomagnetic
field anomaly changes, by comparing the recently measured values with the old

3
ones. We also conducted an additional survey at 20 new stations in the southern
area to explore deeper the presumed negative anomaly region. The total
geomagnetic field intensity was measured every 30 seconds during 30–60
minutes, using an ENVI PRO Proton Magnetic System (Scintrex), which
measures total magnetic fields in the range of 23,000–100,000 nT, with an
accuracy of ±1 nT. The measurements noise has been minimized by eliminating
those data considered as outliers, based on the percentage of the noise level as
recorded by the instrument.
The measured values in a geomagnetic measurement consist of the sum of
different field contributions. The so-called internal parts are the core field, also
known as the main field, and the lithospheric field, produced by magnetized
crustal rocks. The external parts have sources in the ionosphere and
magnetosphere. Variations of the external parts also induce currents in the
lithosphere and upper mantle resulting in induced secondary fields. Since the
external and quickly varying induced fields can reach similar amplitudes as the
desired lithospheric signal, they have to be minimized (removed) from
measurements. Moreover, the influence of the secular variation, i.e. the slow
change of the core magnetic field, has to be taken into account by reducing all
measurements to a common epoch.
Based on the assumption that transient variations of geomagnetic field are
identical at both the observed stations and the reference station, a simple
formulation (Newitt et al. 1997) can be applied to eliminate external field
contributions from the measured values (i.e. consider only the internal
contributors to obtain geomagnetic mean values of the observed stations), as
follows:
O(t) – O = C(t) – C,

hence

O = O(t) – [C(t) – C],

where O is the geomagnetic mean value of the observed station; O(t), the
instantaneous geomagnetic value measured in the observed station; C(t), the
instantaneous geomagnetic value measured in the reference station; and C, the
geomagnetic mean value of the reference station.
Following the above formulation, continuous field readings from a nearby
reference station are necessary for this data processing. Two Base Stations have
been set up consecutively for this purpose, one for the southern area (BS-1) and
one for the northern area (BS-5). Unfortunately, the second Base Station was
destroyed on the second day of the second part of our survey. Therefore, for the
northern area, we had to choose and to use the minute magnetic data provided by
one of the three neighbouring geomagnetic observatories – Tondano (TND),
Kupang (KPG) and Kakadu (KDU) – as the reference station for data reduction
processing.
The evaluation of the suitability of the data provided by the three neighbouring
geomagnetic observatories has been done by comparing them with corresponding
variations recorded continuously during 16 days in the Penyu Island (PNY), a tiny
and magnetically quiet island located very close (about 1.5 km) to the Lombok
Island. An Overhauser magnetometer (GSM-19 v7.0, GEM System), measuring

4
total magnetic fields ranged 15,000–120,000 nT with an accuracy of ±0.1 nT, was
installed on PNY as well as on BS-1 and BS-5. This instrument was verified by
comparing its reading with the NGK (Niemegk geomagnetic observatory,
Germany) standard Overhauser magnetometer. Moreover, the Proton
magnetometer (ENVI PRO) used at the stations was calibrated to this Overhauser
magnetometer, via cross-reading at the Calibration Stations (Calib). The locations
of three neighbouring observatories, the Penyu Island, the Calibration Stations,
and two Base Stations are shown in Fig. 1 and Fig. 3.
Data provided by TND turned out not to be suitable, because of significantly
larger diurnal variation (about 36 nT between day and night time), probably due to
its location closer to the magnetic equator. Data provided by KPG, which is the
nearest observatory (about 850 km East of the Lombok Island), are actually of the
best quality, but unfortunately, because of some technical problems they are not
available for some days of our magnetic survey. Consequently, we chose the KDU
data, which are in good agreement with PNY, to eliminate the external field
contributions on the observed stations data. All measurements have been reduced
to epoch 2006.84, which is November 1st 2006, a date near the middle of the
survey time-span.
After the mean values for each station have been calculated, the data were
classified by using the standard deviation (StDev) values, following the criteria:
good quality (StDev < 2 nT), intermediate (2 nT ≤ StDev < 5 nT), and low quality
(StDev ≥ 5 nT) data. The next step in the data processing was to choose more
reliable data, when there were redundant measurements at the same station on
different days. In this case, the observations that were made during more
magnetically quiet days, i.e. on the day when the Kp index (http://www.gfz-
potsdam.de/pb2/pb23/GeoMag/niemegk/kp_index/index.html) and K indices for
the Australian observatories (http://www.ips.gov.au) are smaller than 3, have been
kept.
By using only good quality data and subtracting the 10th generation IGRF values
(Maus et al. 2005) to eliminate the core field, a 2D geomagnetic anomaly map has
been generated using Oasis montaj 6.4 (Geosoft software). To obtain it, the
Kriging method has been applied, using linear power model with a blanking
radius of 0.0527° (equal to 5.85 km) to interpolate the available data and to show
the empty spaces for the parts without data, as shown in Fig. 3. This figure shows
a strong dipolar magnetic anomaly in the southern part of the survey area, with the
minimum point located on (-8.76°N, 116.03°E) and the maximum point on (-
8.73°N, 116.09°E). The new values of the geomagnetic anomaly are lower
compared to the 2004–2005 survey results with the differences ranging from
-239.26 to -2941.80 nT. These differences are larger than reasonable decreases of
an induced anomaly due to the main field in this region, characterized by a secular
variation of some -13 nT/year. However, we have to stress once more that the
previous measurements have to be taken with care because of the large
uncertainties in the observations. Therefore, these differences are not discussed
further here, but will be investigated in a future work when survey data from fall
2007 to spring 2008 will be available.
Furthermore, by generating a 3D view as shown in Fig. 4, the horizontal gradient
as well as geomagnetic anomaly pattern in the Lombok Island could be described
as repeated contiguous negative-positive anomalies: a low anomaly in the north, a
high positive anomaly in the middle, and a negative anomaly in the south. If the

5
negative anomaly extends southward, then it matches the above mentioned
general geomagnetic anomaly pattern along the Java Trench (see Fig. 2). This
ensures its reliability and promises a smooth transition to the global geomagnetic
data (see Fig. 1).

3. Models and interpretations

By extracting a profile, connecting the strongest apparent dipolar magnetic
anomaly from the negative part on (-8.79°N, 115.98°E) to the positive part on
(-8.70°N, 116.16°E), as illustrated by a blue straight line on Fig. 3 and by a black
curve on Fig. 4, two forward models have been constructed.
The first model only assumes one uniformly magnetized sphere, equivalent to a
dipole (Blakely 1995). In this modelling, location and magnetization parameters
have been adjusted (individually and together) using a discrete exploration
procedure to predict the magnetic anomaly along the profile. Seven parameters
have been investigated: the horizontal coordinates (x, y), the depth (z), the radius
(r), the magnetization intensity (M), inclination (I) and declination (D).
Parameters ranges are indicated in Table 1. We have arbitrarily constrained the
source to be along the profile (i.e. -8.8°N ≤ y ≤ -8.7°N; with sampling interval
(denoted thereafter as si) of 0.1°) and within the longitudinal extent of the
anomaly (116.00°E ≤ x ≤ 116.13°E; with si of 0.01°). Then, the radius of the
sphere is considered exclusively smaller than the depth of its center (r < z), and z
is limited to a maximum 20 km, because the resolution of the used data prevents
to map very deep crustal sources. Nevertheless, characterizing the depth with high
accuracy is needed. Therefore, si of 0.1 km between 0 and 20 km is used for the
depth, and si of 0.5 km for the radius. Concerning magnetization parameters, the
intensity is investigated using si of 1 Am-¹ between 0 and 20 Am-¹. This allows us
to cover a wide range of rock magnetized susceptibility values. Magnetization
directions are first forced to be aligned with the main magnetic field directions in
the corresponding area for the survey epoch, i.e. Iif = -34.75°and Dif = 1.70°
(http://ngdc.noaa.gov/geomagmodels/IGRFWMM.jsp). However, it appears that
assuming an induced magnetic source does not show a good fit of the magnetic
anomaly, whatever the remaining parameters are. Therefore, we have also used a
discrete exploration of inclination (-90° ≤ I ≤ 90°; si: 2°) and declination (-180° ≤
D ≤ 180°; si: 15°) values.
Figure 5 shows the best fit model obtained via this approach, and its parameters
are indicated in Table 1. This magnetic body is located in the upper-middle crust
and has an intense magnetization, with inclination and declination far from the
induced directions. With this geometry, the source of the anomaly would have to
have a very strong remanent magnetization. This result is in a good agreement
with the previous interpretation (Zubaidah et al. 2005). However, the new model
suggests a shift of the structure location to the southwest direction and a relatively
smaller size of the body.
The second model was obtained using the GM-SYS Profile Modeling (an
extension to Oasis montaj, Geosoft software), which assumes rocks and
sediments, as several blocks having different susceptibilities and remanent
magnetizations, as the sources of anomalies. In this modelling method, the general
shape of the total anomaly response signal is determined primarily by locations
6
and shapes of each block, whereas detailed curvature and its amplitude are
influenced by the susceptibilities and the remanence parameters (magnetization
intensity, inclination, and declination) of each block.
To start our model, the observed magnetic anomaly signals have been directly
imported from the concerned magnetic grid by digitizing the desired profile on the
2D anomaly map of Fig. 3, taking 225 points to get about 100 m intervals between
adjacent points. The Earth’s surface shape has been derived from the topography
grid of SRTM30 Plus v4, obtained from the Geosoft DAP server
(http://dap.geosoft.com/geodap/home/default.aspx). The structures of surficial
rocks as well as their subsurface continuations have been determined from the
local geological map (Mangga et al. 1994), whereas deeper structures have been
estimated based on the seismic and geoacoustic investigations over the closest
area (the 116° corridor of Kopp and Flueh 2007). To resolve uncertainties of
geological structures across the coastal area, the geological profile (green lines of
Fig. 3 and Fig. 7) has been shifted 1.5 km south-eastward (landward) from the
anomaly profile (blue lines of Fig. 3 and Fig. 7).
To get proper calculations, several initial parameters of the ambient core magnetic
field have to be given, including its magnitude (H), inclination (FI), and
declination (FD); additionally the profile azimuth and the relative strike angle
have also to be set. All initial magnetic parameters are determined from the 10th
IGRF for the location of the south Base Station
(http://ngdc.noaa.gov/geomagmodels/IGRFWMM.jsp), while profile azimuth
could be determined from the direction of the extracted profile counted clockwise
from the true North. The strike angle is set to 70° relative to the extracted profile,
considering the known nearest surface lineament on the geological map, which
shows a south-eastward trend. The red lines of Fig. 3 and Fig. 7 show the intended
strike line. All the initial parameter inputs of the second model are indicated in
Table 2.
Firstly, an intrusive igneous rock (labelled as Tmi on Fig. 6 as well as on the
geological map of Fig. 7), due to its extremely high susceptibility, has been
assumed as the main source of anomalies. We have started our model by taking a
simple subsurface continuation of this block (directed straight downward to the
basement rocks), also by simply determining its inclination and declination
initially in line with the ambient magnetic field. Based on the local geological
map, it has been considered that this intrusion is composed by Dacites and
Basalts. To produce such a huge anomaly, the initial susceptibility should be close
to the maximum value of basaltic rocks (ranged 250–180,000 x 10-6 SI (Table 1 of
Hunt et al. 1995)), while the remanent magnetization has to be relatively low and
in the permissible range of Koenigsberger ratio (Qn) of intrusive rocks (ranged
0.1–20 (Table 6 of Hunt et al. 1995)). We have then tried to adjust the inclination
and declination values, in combination with magnetization intensity, to get the
desirable response. Changes of the declination only affect the anomaly amplitudes
and have no effects on the shape of the response signal. Therefore we keep the
initial value of declination. Changes in the inclination of the magnetization only
affect the response signal on the right side of the block. In other words, the
negative part of the anomaly signal could only be fit by shifting the block to the
left (as seen in the dark blue curve of Fig. 6), and getting a significant decrease of
the RMS error (from initially 349.18 nT down to about 150 nT).

7
To improve the fit to the anomaly signal, contributions of other nearby rocks have
to be taken into account. Volcanic rocks typically have strong remanent
magnetization. After looking back at the geological map, we have modelled two
blocks of volcanic rocks on both sides of the intrusive rocks, labelled as TompR
and TompL on Fig. 6, representative of surrounding volcanic rocks (labelled as
Tomp on the geological map of Fig. 7). For these blocks, we take rather low
values of susceptibility (i.e. 1,000 x 10-6 SI) and relatively high values of
remanent magnetization intensity, but still keep the relationship between both
values in the permissible range of Koenigsberger ratio (Qn) of volcanic rocks
(ranged 30–50 (Table 6 of Hunt et al. 1995)). Their inclinations then have to be
adjusted to a value which provides the lowest error; let us note that both
inclinations must be equal, regarding the same age of their geological formations.
With the above described approaches, the total response shape (black curve of
Fig. 6) already agrees with the observed curvature (black points of Fig. 6). At this
step, the RMS error has been reduced to about 120 nT, without reaching yet our
suitable limitation of misfit. We expect that the RMS error would be less than 5%
of the peak to peak anomaly (i.e. about 60 nT), which substantially reduces the
range of possible solutions, and can provide a reasonably reliable model.
Then, adjustments of the lowest ends of both volcanic blocks (their lowest
boundaries to the lower sedimentary rocks) lead to significantly reduce the errors.
Further decreasing of errors could be reached by introducing lateral extensions of
the intrusive rock, representing the sills on the left and right sides (labelled as
Tmi-extL and Tmi-extR on Fig. 6). The possible occurrence of such sills is
supported by the geological map of Fig. 7, regarding the occurrence of many
small intrusions in the surroundings.
After applying those approaches, we actually have reduced the RMS error to
about 80 nT, close to the expectation, but still have considerably bothersome
ripples at both peaks of negative and positive anomalies. Although they are below
the possible resolution given by our measurement station density, we aim for a
smoother response of the negative anomaly. Therefore we have considered the
contribution of a surficial sedimentary rock (labelled as Tomk on Fig. 6 as well as
on the geological map of Fig. 7), which had been underestimated before. This
block, as described by the geological map, could be considered as Sandstone with
mean susceptibility value of about 10,000 x 10-6 SI (Table 1 of Hunt et al. 1995).
Its inclination and declination is assumed to be the same as those of volcanic
rocks (TompL and TompR), due to the same age of their geological formations. By
adjusting its remanent magnetization intensity in the permissible range of
Koenigsberger ratio (Qn) of average sedimentary rocks (ranged 0.02–10 (Table 6
of Hunt et al. 1995)), we have achieved a smooth negative anomaly peak. Table 3
indicates magnetic properties of all rocks used for the second model. Finally,
carefully adjusting the detail shapes of intrusive rock, as well as its right side
bounding to the volcanic rock, leads to much better smoothness of the positive
anomaly peak, hence reducing the RMS error to the very low value of about 35.5
nT (see red curve of Fig. 6).
Let us suggest a possible interpretation of the second model in terms of tectonics.
The intrusive rock formations on the Lombok Island might be caused by the
subduction processes on the Java Trench, about 315 km south of the block
considered for the studied geomagnetic anomaly (Kopp and Flueh 2007). On this
convergent margin, the Australian plate subducts beneath the Eurasian plate,
resulting in compressional faults with directions normal to the Java Trench (Kopp
8
et al. 2006). Moreover, from the best documented tectonic regime for the Lombok
Island and circumference regions (segment 3 in Fig. 7 of Špičák et al. 2007), all
19 earthquake events in 1977 show normal faulting (no strike-slip) with very
homogeneous position of nodal planes parallel to the trench. In the second model,
we can infer a possible tectonic interpretation from the differences of the depth
extensions of the volcanic rocks, with the left block reaching about 0.5 km deeper
than the right one. The volcanic rocks (i.e. the oldest formation on the Lombok
Island, which formed in the Late Oligocene to Early Miocene) can be assumed to
be a direct product of the Java Trench subduction, and therefore initially a single
continuous structure. Discontinuities in the local geological structures occurred
later, by intruded igneous rock in the Middle Miocene, as a result of a continued
subduction on this region. Here, a subsurface normal faulting could be assumed,
regarding the relationship between volcanic rocks on the left side (TompL of Fig.
6) and the volcanic rocks on the right side (TompR of Fig. 6), corresponding to the
hanging wall and the footwall (i.e. TompL moves downward relative to TompR).
Finally, we summarize that the second model suggests the possible existence of a
subsurface normal fault in this region, which might be considered as a potential
trigger for local tectonic earthquakes. This interpretation agrees well with the
known regional geological and tectonic structures. As shown in the geological
map of Fig. 7, the boundary between the negative and positive part of
geomagnetic anomaly (red lines of Fig. 3 and Fig. 7), also considered as the strike
direction of the suspected fault line in our model, is parallel to the known nearest
surface lineament.

4. Discussions and Conclusions

The general geomagnetic anomaly pattern in the Lombok Island consists of
repeated contiguous negative-positive anomalies. A strong dipolar magnetic
anomaly in the south could be interpreted in two different ways: as a specific local
structure composed by a quite large magnetic body or as a discontinuity in the
geological structure (e.g. potentially local fault). The second proposed model and
its interpretation are supported by the agreement with the known large scale
regional tectonic structure and local geological surface conditions. This model
seems to be the more reasonably explanation for the observed magnetic anomaly,
and suggests a possible association with geological and tectonic features in this
area.
We have presented the first detailed images of the magnetic field over a part of the
Lombok Island as an initial step to obtain a better understanding of the geological
and tectonic structures. Our preliminary modelling results show the capability of
magnetic ground survey data to underline new tectonic elements (e.g. the local
faults), which, however, should be further confirmed by other studies. Although
MT (magneto-telluric) measurements are preferable to investigate electrical
conductivity of Earth’s rocks and sediments, it would be very difficult to conduct
such measurements appropriately, due to topographical constraints over the
Lombok Island and very dense population at some places. DC resistivity
measurements are, therefore, expected to provide complementary information.
Moreover, because the significant magnetic anomaly in this island is an intriguing
one, recently (during fall 2007 and spring 2008) new scalar geomagnetic

9
measurements have been conducted over a broader area of the Lombok Island.
Our aim is to cover not only the subduction zone in the South (the Java Trench),
but also the volcanic area (Mount Rinjani) and the zone of possible subduction
extension and reactivation in the North (the Flores Thrust). Figure 2 shows the
locations of the main concerned tectonic settings of this region. At the time of
writing, these measurements are in the processing stage. Let us note that we
would like to connect our results to the existing eastern Asian anomaly map and
extend our investigated area. Furthermore, we plan to install some continuous
geomagnetic measuring instruments permanently on the Lombok Island. This
observatory, planned to record the total field intensity as well as three components
of the geomagnetic field, will be a very valuable complement for the next
magnetic surveys.

Acknowledgements We would like to address our appreciations to the reviewers and editors
for very constructive comments and suggestions. We thanks to Muhammad Husni and Hendar
Gunawan (BMG, Indonesia) for providing the Tondano geomagnetic observatory data; K. Yumoto
and Shuji Abe (SERC, Kyushu University, Japan) for providing the MAGDAS Kupang data. C.
Subarya, Hery Hardjono and Awang H. Satyana have provided us with fruitful discussions about
tectonics of Indonesia, Naila Mohamed Babiker about seismicity, and Mohamed Hamoudi about
magnetic modelling. Special thanks for Seiya Uyeda and K. I. Oyama for their interests and
comments on our first public speaking during the International Workshop on Seismo-
Electromagnetic Phenomena (IWSEP) 2007 in Bandung. For our successful field research, we also
gratefully thanks to all family and friends, Sekotong people, the members of the research team of
Electrical Engineering Dept. of Mataram University, Institut Teknologi Bandung (ITB), and
Institut Teknologi Sepuluh Nopember (ITS) Surabaya, especially for the best co-operation of
Widya Utama and D.D. Warnana.

References
Balasis G, Mandea M (2007) Can electromagnetic disturbances related to the recent great
earthquakes be detected by satellite magnetometers? Tectonophysics 431:173-195.
Blakely RJ (1995) Potential theory in gravity and magnetic applications. Cambridge University
Press, Cambridge.
Blakely RJ, Brocher TM, Wells RE (2005) Subduction-zone magnetic anomalies and implications
for hydrated forearc mantle. Geology 33(6):445-448.
Hinschberger F, Malod JA, Rehault JP, Villeneuve M, Royer JV, Burhanuddin S (2005) Late
Cenozoic geodynamic evolution of eastern Indonesia. Tectonophysics 404:91-118.
Hunt CP, Moskowitz BM, Banerjee SK (1995) Magnetic properties of rocks and minerals. In:
Ahrens TJ (ed) Rock physics & phase relations, Handbook of physical constants. American
Geophysical Union, Washington, pp 189-204.
Ishihara T, Kisimaoto K (2002) Magnetic Anomaly Map of East Asia 1:4,000,000, CD-ROM
Version (2nd Edition), Digital Geoscience Map P-3, Geological Survey of Japan (AIST) and
Coordinating Committee for Coastal and Offshore Geoscience Programmes in East and
Southeast Asia (CCOP).
Kopp H, Flueh ER, Petersen CJ, Weinrebe W, Wittwer A, Meramex Scientists (2006) The Java
margin revisited: Evidence for subduction erosion off Java. Earth Planet Sci Lett 242:130-142.
Kopp H, Flueh ER (2007) Seismic and geoacoustic investigations along the Sunda-Banda Arc
transition. FS Sonne Fahrtbericht/Cruise Report SO 190 SINDBAD, Report Nr. 8, IFM-
GEOMAR, Leibniz-Institut für Meerewissenschaften an der Universität Kiel.
Mangga SA, Atmawinata S, Hermanto B, Amin TC (1994) Geologi lembar Lombok,
Nusatenggara, Lembar 1807, Pusat Penelitian dan Pengembangan Geologi, Indonesia.
Maus S, Macmillan S, Chernova T, Choi S, Dater D, Golovkov V, Lesur V, Lowes F, Lühr H, Mai
W, McLean S, Olsen N, Rother M, Sabaka T, Thomson A, Zvereva T (2005) The 10th
generation international geomagnetic reference field. Phys Earth Planet Inter 151:320-322.

10
Maus S, Rother M, Hemant K, Stolle C, Lühr H, Kuvshinov A, Olsen N (2006) Earth’s
lithospheric magnetic field determined to spherical harmonic degree 90 from CHAMP satellite
measurements. Geophys J Int 164:319-330.
Müller C, Neben S (2006) Research Cruise SO190 Leg 1, Seismic and geoacoustic investigations
along the Sunda-Banda Arc transition, with RV Sonne. Cruise report and preliminary results,
Bundesanstalt für Geowissenschaften und Rohstoffe (BGR), Hannover.
Newitt LR, Barton CE, Bitterly J (1997) IAGA Guide for Magnetic Repeat Station Surveys.
Nishida Y, Utsugi M, Mogi T (2007) Tectonomagnetic study in the eastern part of Hokkaido, NE
Japan (II): Magnetic fields related with the 2003 Tokachi-oki earthquake and the 2004
Kushiro-oki earthquake. EPS 59:1181–1186.
Purucker M, Ishihara T (2005) Magnetic images of the Sumatra region crust. EOS 86(10):101-102.
Quesnel Y, Catalan M, Ishihara T (2009) A new global marine magnetic anomaly data set. J.
Geophys. Res. Doi:10.1029/2008JB006144.
Špičák A, Hanuš V, Vaněk J (2007) Earthquake occurrence along the Java trench in front of the
onset of the Wadati-Benioff zone: Beginning of a new subduction cycle? Tectonics
26:TC1005.
Zubaidah T, Kanata B, Utama W, Warnana DD, Arumdati N (2005) Laporan akhir penelitian -
PRSD MIPA tahun 2005. Laporan penelitian, Jurusan Elektro Fakultas Teknik-Universitas
Mataram.

Table 1. Parameters used for the first model

Parameter Range Best fit
Coordinate (x) 116.00°E ≤ x ≤ 116.13°E; si: 0.1° 116.07°E
Coordinate (y) -8.8°N ≤ y ≤ -8.7°N; si: 0.01° -8.74°N
Depth (z) 0 ≤ z ≤ 20 km; si: 0.1 km 6.5 km
Radius (r) 0 ≤ r ≤ 20 km; si: 0.5 km 3.5 km
Magnetization intensity (M) 0 ≤ M ≤ 20 Am-1; si: 1 Am-1 14 Am-1
Inclination (I) -90° ≤ I ≤ 90°; si: 2° 0°
Declination (D) -180° ≤ D ≤ 180°; si: 15° -113°

Table 2. Initial parameters used for the second model

Parameter Initial value
Earth’s magnetic field:
Magnitude [Am-1] 36.095 (equal to 45358.4 nT)
Inclination [deg] -34.75
Declination [deg] 1.70
Profile azimuth [deg] 68.2
Relative strike angle [deg] 70

Table 3. Magnetic properties of rocks used for the second model

Block’s name Susceptibility Remanence Magnetization
[10-6 SI] Intensity Inclination Declination
[Am-1] [deg] [deg]
Intrusive igneous rock (Tmi) 110,000 0.4 52 1.7
Volcanic rock (TompL) 1,000 1.2 17 1.7
Volcanic rock (TompR) 1,000 1.2 17 1.7
Intrusive igneous rock (Tmi-extL) 110,000 0.4 52 1.7
Intrusive igneous rock (Tmi-extR) 110,000 0.4 52 1.7
Sedimentary rock (Tomk) 10,000 1.0 17 1.7

11
Figure 1. Map of the total component of the large-scale lithospheric field for Indonesian region obtained from the MF5 model (Maus, et al. 2006) at 5 km altitude. Plate
boundaries and subduction zones are indicated as red lines. The Lombok Island (pointed out by a black arrow) is located between the highest intensity contiguous negative-
positive geomagnetic anomalies in this region. Black circles are the locations of the Penyu Island (PNY) and three neighbouring geomagnetic observatories, i.e. Tondano (TND),
Kupang (KPG), and Kakadu (KDU).

12
Figure 2. Detailed map of the total field anomaly for southern part of Indonesian region: a contiguous negative-positive anomaly pattern clearly exists along the Java
Trench. The position of the Lombok Island is pointed out by a black arrow, while the black triangle represents the location of Mount Rinjani. The subduction zones are indicated
as black solid lines, while the dashed one indicates its presumed extension. This map is generated from cleaned and levelled marine magnetic data available from GEODAS for
the interval 1950–2004 (Quesnel et al. 2009), combined with marine magnetic data along the Sunda-Banda Arc transition of BGR (Müller and Neben 2006) and aeromagnetic
data of AIST and CCOP (Ishihara and Kisimaoto 2002). The white areas over oceans or islands, including that over the Lombok Island, indicate that no data are available.

13
Figure 3. Map of the geomagnetic total field anomaly interpolated over the surveyed area on the Lombok Island. The two base station locations are shown as green
inverted triangles, while the Calibration Stations is depicted as a red inverted triangle. The exact location of the Penyu Island is marked with PNY. The geomagnetic survey
stations are also shown, providing good quality (black points), intermediate (blue triangles) and low quality (red crosses) data (see text for more details). Only the data with good
quality are used. The blue line represents the trace of the magnetic profile studied in the following, which connects two peaks of a strongest apparent dipolar magnetic anomaly.
The green line represents the trace of the geological profile used in the second model, while the red one represents the presumed strike direction.

14
Figure 4. 3D view of the geomagnetic total field anomaly map for the same area as in Fig. 3, underlining the repeated contiguous negative-positive anomalies. The
intensities of the geomagnetic anomalies are contoured in the form of elevations and leveled with colour scale exactly as in Fig. 3. The black curve represents the trace of the
magnetic profile studied in the following, which connects two peaks of a strongest apparent dipolar magnetic anomaly.

15
Figure 5. Comparison between the first forward model (thin line) that assumes a uniformly magnetized sphere as the source of the anomaly and the profile of observations
(thick line).

16
Figure 6. Comparison between the second forward modelled profile (black line on top panel) and the profile of observations (black dots on top panel). The individual
responses of the blocks could be seen as the colour legend on the left top of top panel, while the errors are represented by the red line. This model assumes rocks and sediments
(as several blocks having different susceptibilities and remanent magnetizations) as the sources of anomalies, obtained by using the GM-SYS Profile Modeling (an extension to
Oasis montaj, Geosoft software).

17
View publication stats

Figure 7. The geological map of corresponding area (Mangga, et al., 1994.) shows the geological formations which have contribution to the geomagnetic anomaly:
Intrusive igneous rock (Tmi), Volcanic rock (Tomp), and Sedimentary rock (Tomk). The blue line indicates the location of the extracted profile of Fig. 3 and Fig. 4, while the
green line represents the trace of the geological profile used in the second model. The exact positions of two peaks of anomaly are shown as cross circled symbols, labelled as
Min (for the peak of the negative anomaly) and Max (for the positive one). The red line, representing the boundary between the negative and positive parts of the studied
geomagnetic anomaly, considered also as the strike direction of the suspected fault line in the second forward modelling, is parallel to the known nearest surface lineament.

18