Tes PDF

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

Analysis of Preseismic Event Using Seismo-

electromagnetics Data
(Earthquake Case Study in Megathrust Sumatra Region August-October 2016)

M. Fakhrul Islam Masruri, Ramadhan Priadi, and Suaidi Ahadi


Bayu Merdeka Tri Fristiyan Nanda Center of Engineering Seismology, Potential Geophysics
School of Meteorology, Climatology and Geophysics and Time Mark
Meteorological, Climatological and Geophysical Agency Meteorological, Climatological and Geophysical Agency
South Tangerang, Indonesia Jakarta, Indonesia
fakhrulmasruri@gmail.com su4idi@gmail.com

Abstract—The megathrust region of Sumatra is an active Atmosphere Ionosphere) coupling [7]. A few days before the
seismic area. Before the earthquake happened, geomagnetic field earthquake came SEAs anomalies (Seismo-electromagnetic
anomaly was recorded at the geomagnetic recording station anomalies) [6].
located in Sumatera. In the lithospheric layers that gain stress
and strain before an earthquake generate electromagnetic Prior to the earthquake the rocks experienced an
emissions that can be recorded at a geomagnetic station at a accumulated increase in stress which then broke out when it
certain distance. The electromagnetic emission propagation was no longer able to withstand the accumulation of stress [9],
passed the rock until it reaches the ionosphere which is often resulting in a shift in the fault field and generating
called the LAI coupling (Lithosphere, Atmospheric, and
electromagnetic emissions (piezoelectric and piezomagnetic)
Ionosphere). The purpose of this study is to confirm the existence
of geomagnetic field anomalies before the earthquake. This
which can be recorded at a Geomagnetic station at a certain
research uses the earthquake data in Sumatera region with distance [2]. There are many mineral constituents, one of
period between August to October 2016. The earthquake data which is quartz mineral [11]. Induced quartz minerals emit
used were Mukomuko earthquake August 24 (M 5.8), South electromagnetic waves and a certain few of electric ions when
Bengkulu earthquake September 17 (M 5.4), and Padang get pressed [12].
earthquake October 21 (M 4.5). The geomagnetic sensor used is
MAGDAS 9 located at 3 stations in Gunungsitoli station, Sicincin Electromagnetic wave emission is one of the parameters
station, and Liwa station. The method used is the ULF data that can be used as earthquake precursor [1]. Electromagnetic
signal filtering method from the multi-station geomagnetic data wave emissions can only be observed in ULF (ultralow
to obtain an azimuth geomagnetic field anomaly. From the frequency) data in the frequency range 0.01-10 Hz whereas in
research results obtained lead time ranged between 9 to 18 days the extremely low/very low-frequency ELF/VLF data in the 10
before the earthquake. The results from the azimuth direction of Hz-32 Hz anomaly range can’t be observed prior to the
the geomagnetic field anomaly correlated to the epicenter earthquake [10]. ELF/VLF can’t be used as data precursors
location of the earthquake.
because high-frequency electromagnetic waves will experience
Keywords—preseismic; seismo-electromagnetics; geomagnetic; very rapid attenuation when propagating through the
sumatera lithosphere, whereas ULF is only slightly attenuated so that
emissions can still be recorded at observation points on the
surface of the earth [3].
I. INTRODUCTION
Sumatra is an area strongly influenced by the activity of The type of ULF anomaly emission is divided into three
Indo-Australia plate motion that infiltrated into the Eurasian types that are polarization anomalies caused by changes in
plate, besides the island of Sumatra also has a Sumatera fault conductivity in the lithosphere, ULF emissions caused by
system that stretches from the north to the south of the island micro-fracturing, and ULF emissions caused by electro-kinetic
and several megathrusts which has the potential to generate an effects [13]. It is assumed that all earthquakes produce an ULF
earthquake with magnitude big [5]. Sumatra region has several emission anomaly that can be captured by a magnetometer at
geomagnetic stations that can be used as earthquake precursor the observer station. The Magnetometer's ability to detect an
data. Geomagnetic stations located in Sumatra are Tuntungan ULF emission anomaly depends on the potential magnitude
(TUN), Gunungsitoli (GSI), Liwa (LWA), and Sicincin (SCN). and distance of the hypocenter to the geomagnetic station [1].

Earthquakes are a phenomenon sudden energy release by Electromagnetic wave emissions can be analyzed using an
rock fractures. Prior to the earthquake, stress forces in rocks analytical signal at the local station's magnetic data recording.
emit electromagnetic waves and a number of electrical ions The anomalies are obtained from the onsite time recorded by
when crossing the earth's layer or known as LAI (Lithosphere the geomagnetic station [1]. The difference between the signals
generated by geomagnetic activity and seismic activity when TABLE I. SUMATRA’S GEOMAGNETIC STATION LOCATION

microfracture occurs will result in the onset of geomagnetic Geomagnetic Station Longitude Latitude
data recording which can be used as earthquake precursor data
1 Gunungsitoli Station (GSI) 97.57° 1.304°
[8].
2 Sicincin Station (SCN) 100.30° -0.550°

3 Liwa Station (LWA) 104.06° -5.020°

The earthquakes data used in the megathrust Sumatra


during the period of August, September, and October 2016 i.e.,
Mukomuko earthquake August 24 (M 5.8), South Bengkulu
earthquake September 17 (M 5.4), and Padang earthquake
October 21 (M 4.5). Earthquake data obtained from BMKG’s
earthquake data repository. And then table II shows the
parameters of the earthquake data used in this paper.

TABLE II. EARTHQUAKES DATA USED

Location OT M Lon. Lat. Dep.


Southwest of Aug 24, 2016
1 5.8 100.19˚ -2.90˚ 25km
Mukomuko 13:48:45 UTC
Southwest of Sep 17, 2016
2 5.4 102.49˚ -4.79˚ 19km
South Bengkulu 02:50:53 UTC
Fig. 1. Representative power spectra of typical geomagnetic activity (solid Southwest of Oct 21, 2016
3 4.5 100.16˚ -1.83˚ 10km
curve) and corresponding broadband seismic activity (dashed curve) recorded Padang 01:45:02 UTC
[8].

Figure 1 shows the different recording signals due to Figure 2 is a map showing the location of geomagnetic
geomagnetic activity and seismic activity. Very visible if the stations, and the epicenter of three earthquake in Sumatra.
signal due to seismic activity has a larger peak when Geomagnetic stations are marked with triangular symbols,
microfracture occurs in the rock. This is the underlying if while epicenter of an earthquake is marked by a star symbol.
before an earthquake magnetic anomaly happens.

Although earthquakes are still not predicted today because


of the conditions for predicting things we need space
dimensions and time dimensions but we can start to make
precursor earth earthquakes as the first step of earthquake
mitigation.

II. METHODS

A. Research Area
The research area is located in the megathrust region of
Sumatra with the extent of research is 6.69 north to -6.56 south
and 94.7 east to 107 east, Indonesia. Selection of research area
is adjusted to the distance of earthquake occurrence and
availability of geomagnetic observation station. This research
uses data recording geomagnetic activity from record station in
Sumatera areas such as Gunungsitoli station (GSI), Sicincin
station (SCN) and Liwa station (LWA). The coordinate
location of the geomagnetic station in Sumatra is shown in
Fig. 2. Map of research area.
Table I.
B. Instruments
The geomagnetic activity recorder used in the geomagnetic
station in Sumatra is using the Magnetic Data Acquisition (2)
System (MAGDAS-9) magnetometer manufactured by Space
Environment Research Center (SERC) Kyushu University, Then we get daily value of component H and Z as follows:
Japan.
(3)

(4)
Then the polarization ratio can be formulated as:

(5)
Then change the power ratio back to time domain. The Z/H
Fig. 3. Main parts of MAGDAS-9 magnetometer: (1) 3-axial flux-gate sensor; polarization can show the geomagnetic field anomaly which
(2) pre-amplifier; (3) data logger; and (4) antenna of GPS-receiver. (source: can be used as an earthquake precursor parameter if it has a
ikfia.ysn.ru) value that exceeds its standard deviation value. The
geomagnetic field anomaly is confirmed using the recording
TABLE III. MAGDAS-9 SPECIFICATION (SOURCE: IKFIA.YSN.RU) data from other observation stations (multi-stations method)
MAGDAS-9 Magnetometer
and then DST index to ensure that the recorded anomaly is true
an anomaly of seismic activity and is not derived from anomaly
Magnetic field components measurement H, D, Z geomagnetic form of a magnetic storm. Here is a classification
Accuracy 0.1 nT of magnetic storms based on DST index shown by Table IV.
Range ±70000 nT TABLE IV. DST INDEX CLASSIFICATION
Sampling rate Up to 10 Hz
Criteria of Magnetic Storm Value Range
Weak -30 nT > Dst > -50 nT
Magnetic field components measured by MAGDAS-9
include horizontal component (H), declination (D), and vertical Moderate -50 nT > Dst > -100 nT
component (Z). The principle of operation of magnetometer is Strong -100 nT > Dst > -200 nT
using the principle of flux-gate sensor. A flux-gate Severe -200 nT > Dst > -350 nT
magnetometer can measure the magnetic field in a specific
Great -350 nT > Dst
direction quite precisely. It takes three independent flux-gate
coils to measure the total field by adding the three independent
vectors. After that an analysis of the direction of the anomaly
(azimuth) using Single Station Transfer Fourier method
C. Research Methods (SSTF). The direction (azimuth) anomaly can be indicated as a
Starting from a raw data recording geomagnetic activity representation of the earthquake preparation zone. This SSTF
obtained from MAGDAS sensor then filtered with band-pass method is a function of Fourier complex component numbers.
filter at frequency 0.012 Hz. After filtering then used Fast This method can solve the equations of the correlation of
Fourier Transforms (FFT) to convert data from time domain to geomagnetic field components, i.e., components X, Y, and Z
frequency domain. After that calculate ratio value of power [4].
spectra density Z/H component at the frequency of 0.012 Hz.
0.012 Hz frequency is chosen because based on graph (6)
representation of power spectra in Figure 1, frequency of 0.012
Hz is categorized as Pc3 which is frequency of geomagnetic The values of constants A(ω) and B(ω) are obtained by
activity so that suitable to be used to detected geomagnetic using linear inversion. Then can be used to determine the
anomaly [8]. amplitude and the angle of the direction of the geomagnetic
anomaly.
Power ratio polarization method is comparing the vertical
and horizontal values of the geomagnetic fields (Z/H) at the
0.012 Hz frequency recorded in the earth magnetic station. The
ratio Z/H is obtained from the following equations: (7)

(1) (8)
III. RESULTS AND DISCUSSION there was an earthquake precisely in Southwest of Mukomuko,
on the August 24, 2016, with magnitude of 5.8. In this case, the
A. Mukomuko Earthquake August 24, 2016 (M 5.8) earthquake precursor parameters are the onset time on August
9, 2016, and lead time for 15 days from the occurrence of
anomalies until the earthquake.
While for anomaly on August 7, 2016 at GSI then August
8, 2016 in SCN cannot be considered an anomaly because it is
less accurate, it is because only one station recorded the
anomaly and it is estimated that the anomaly is caused by very
local factors such as human activity around the sensor.

Fig. 5. Azimuth of geomagnetic field anomaly on August 9, 2016 at three


observation stations in Sumatra correlated with epicenter of Mukomuko
earthquake.

Using SSTF method calculation obtained by azimuth from


geomagnetic field anomaly on August 9, 2016 at 177° from
GSI station, 192° from SCN station and 290° from LWA
station. Based on plotting azimuth on the map shown in Figure
5, azimuth anomaly at SCN and LWA stations is estimated to
be correlated with an earthquake epicenter located in the
Fig. 4. The top graph is the DST index, then below it is the graph of Z/H ratio Southwest of Mukomuko. But azimuth anomaly from GSI
polarization at frequency 0.012 Hz of the three geomagnetic stations that have slightly deviated. This happens because the earthquake is an
been limited by each standard deviation. (In this case for Mukomuko
earthquake of August 24, 2016) area and not a point so there may be an anomaly azimuth
deviation obtained from each station. Thus, the azimuth of
anomalies is considered as a representation of the earthquake
Based on the results of the data processing shown in Figure epicenter, so that the recorded anomaly is a precursor for the
4, there are geomagnetic field anomalies recorded on three earthquake of August 24, 2016.
geomagnetic stations, GSI, SCN and LWA stations. The
anomalies of the geomagnetic field recorded simultaneously at
the three stations were recorded on August 9, 2016, at 1:00
UTC. And at that time the DST index value is in the range of -
7 nT, so it can be said at that time there is no magnetic storm
disturbance. Each of the Z/H ratio amplitudes recorded in the
three stations is 4.885 in GSI, 5.356 in SCN, and 4.367 in
LWA. 15 days later after the recorded geomagnetic anomaly,
B. South Bengkulu Earthquake September 17, 2016 (M 5.4) an onset time on September 8, 2016, and a lead time for 9 days
from an anomalous air to an earthquake.
While for anomaly on September 7, 2016 at SCN cannot be
considered an anomaly because it is less accurate, it is because
only one station recorded the anomaly and it is estimated that
the anomaly is caused by very local factors such as human
activity around the sensor.

Fig. 7. Azimuth of geomagnetic field anomaly on September 8, 2016 at three


observation stations in Sumatra correlated with epicenter of South Bengkulu
earthquake.

In the figure 7 shown if the direction of anomalies magnetic


corresponding with the direction of epicenter earthquake.
Using SSTF method calculation obtained by azimuth from
geomagnetic field anomaly on September 8, 2016 at 149° from
GSI station, 195° from SCN station and 308° from LWA
station. Based on plotting azimuth on the map shown in Figure
7, azimuth anomaly at GSI stations is estimated to be
correlated with an earthquake epicenter located in the
Southwest of South Bengkulu. But azimuth anomaly from SCN
Fig. 6. The top graph is the DST index, then below it is the graph of Z/H ratio and LWA slightly deviated. This happens because the
polarization at frequency 0.012 Hz of the three geomagnetic stations that have earthquake is an area and not a point so there may be an
been limited by each standard deviation. (In this case for South Bengkulu
earthquake of September 17, 2016) anomaly azimuth deviation obtained from each station. From
the azimuth results of magnetic field anomalies can be
modified as a representation of the epicenter the earthquake, so
Based on the results of the data processing shown in figure that the recorded anomaly is a precursor for the earthquake of
6, there are geomagnetic field anomalies recorded at three September 17, 2016.
geomagnetic stations, GSI stations, SCN and LWA. An
anomalies geomagnetic field recorded on September 8, 2016, at Analysis of azimuth anomalies was performed using the
21:00 UTC. The value of DST index is in the range of -12 nT SSTF method considered as a representation of the earthquake
so it is stated if there is no magnetic storm disturbance. Each preparation zones. This method is capable of generating
Z/H amplitude ratio recorded at the three stations is 2.95 in the azimuth from anomalies because it has information about
GSI, 4.431 in the SCN, and 3.623 in the LWA. 9 days later underground electrical conductivity.
after the recorded geomagnetic anomaly, an earthquake
occurred in South Southwest Bengkulu, on September 17,
2016, with a magnitude of 5.4. In the Bengkulu earthquake had
C. Padang Earthquake October 21, 2016 (M 4.5) 21, 2016, with a magnitude of 4.5. In this case, the earthquake
precursor parameters are onset of time on October 3, 2016, and
lead time for 18 days from an anomalous to an earthquake.
While for anomaly on October 4, 2016 at GSI and SCN
cannot be considered an anomaly because each anomaly shows
the direction of the divergent azimuth, so possibly that the
anomaly is caused by local factors around sensor.

Fig. 9. Azimuth of geomagnetic field anomaly on October 3, 2016 at three


observation stations in Sumatra correlated with epicenter of Padang earthquake.

Using SSTF method calculation obtained by azimuth from


geomagnetic field anomaly on October 3, 2016 at 191° from
GSI station, 244° from SCN station and 305° from LWA
station. Based on the azimuth value generated then when
represented in the map the azimuth of the LWA station will be
correlated with the earthquake epicenter shown in Figure 9.
While for the anomalies retained at the SCN and GSI stations
slightly deviate from the epicenter of the earthquake but still
have a mutually correlated direction. Each azimuth anomaly of
Fig. 8. The top graph is the DST index, then below it is the graph of Z/H ratio the three stations is estimated to be correlated with an
polarization at frequency 0.012 Hz of the three geomagnetic stations that have
been limited by each standard deviation. (In this case for Padang earthquake of
earthquake epicenter located in the Southwest of Padang. Thus,
October 21, 2016) the azimuth of anomalies is considered as a representation of
the earthquake epicenter area, so that the recorded anomaly is a
precursor for the earthquake of October 21, 2016.
The results of data processing for Padang earthquake 3
October 2016 in Figure 8 shown indicate an anomalies
IV. CONCLUSION
geomagnetic field recorded at three geomagnetic stations, GSI
stations, SCN and LWA. Geomagnetic field anomaly occurred The results of magnetic field processing show if there is a
on October 3, 2016, at 22:00 UTC. When the anomaly occurs difference between anomalies generated by magnetic storms
the DST index is in the range of -25 nT, be an anomalous event and anomalies generated by seismic activity. For the three
occurs not by a magnetic storm but a result of seismic activity. earthquakes, if both magnetic anomalies and magnetic field
Each of the Z/H amplitude ratios recorded at the three stations azimuth have a very close correlation. In the Mukomuko
is 2,088 in the GSI, 13,439 in the SCN, and 2,006 in the LWA. earthquake the Z/H ratio of amplitudes recorded in the three
18 days later after the recorded geomagnetic anomaly, an stations is 4,885 in GSI, 5,356 in SCN, and 4,367 in LWA and
earthquake occurred in the Southwest of Padang, on October 15 days later after the recorded geomagnetic anomaly
earthquake happen. In the Bengkulu earthquake each Z/H the with the Strong Earthquakes in Sumatra, Indonesia, during 2007-2012.
amplitude ratio recorded at the three stations is 2.95 in the GSI, Journal of Mathematical and Fundamental Sciences, 47(1), 84–103.
4,431 in the SCN, and 3,623 in the LWA and 9 days later after [2] Finkelstein, D., Hill, R. D., & Powell, J. R. (1973). The piezoelectric
theory of earthquake lightning. Journal of Geophysical Research, 78(6),
the recorded geomagnetic anomaly earthquake happen. In the 992–993.
Padang earthquake each Z/H the amplitude ratios recorded at [3] Hayakawa, M., Kawate, R., Molchanov, O. A., & Yumoto, K. (1996).
the three stations is 2.088 in the GSI, 13,439 in the SCN, and Results of ultra‐low‐frequency magnetic field measurements during the
2,006 in the LWA and 18 days later after the recorded Guam Earthquake of 8 August 1993. Geophysical Research Letters,
geomagnetic anomaly earthquake happen. So that can be said 23(3), 241–244.
as earthquake precursors are anomaly recorded on multi-station [4] Hattori, K. (2004). ULF geomagnetic changes associated with large
and not an local anomaly. If using the SSTF method earthquakes. Terrestrial Atmospheric and Oceanic Sciences, 15(3), 329-
360.
calculation to determine the azimuth of the anomaly
geomagnetic field obtained the azimuth obtained during [5] Ibrahim, G. (2010). Tektonik dan Mineral di Indonesia. Jakarta,
Puslitbang BMKG.
anomaly almost always corresponds to the direction of the
[6] Ikeya, M., & Takaki, S. (1996). Electromagnetic fault for earthquake
epicenter area of earthquake. lightning. Japanese journal of applied physics, 35(3A), L355.
Although the time interval between the occurrence of [7] Kamogawa, M. (2006). Preseismic lithosphere‐atmosphere‐ionosphere
anomalies and earthquakes remains so varied that it has not coupling. Eos, Transactions American Geophysical Union, 87(40), 417–
424.
been able to provide a clear picture of when the earthquake
[8] Karakelian, D., Klemperer, S. L., Fraser-Smith, A. C., & Beroza, G. C.
occurred but the magnetic anomaly data has a great chance to (2000). A transportable system for monitoring ultralow frequency
be studied more deeply about the correlation to earthquakes. electromagnetic signals associated with earthquakes. Seismological
Although earthquakes are still not predicted by precursors of Research Letters, 71(4), 423–436.
earthquakes but precursors can be used for earthquake [9] Mogi, K. (1963). Magnitude-frequency relation for elastic shocks
mitigation study. We still need more exercise with adequate accompanying fractures of various materials and some related problems
earthquake events for a more accurate statistical analysis. in earthquakes (2nd paper).
[10] Surkov, V., & Hayakawa, M. (2014). Ultra and extremely low frequency
electromagnetic fields. Springer.
ACKNOWLEDGMENT [11] Uyeda, S., Kamogawa, M., & Nagao, T. (2011). Earthquakes,
The author would like to thank Mr. Suaidi Ahadi who has Electromagnetic Signals of. In Extreme Environmental Events (hal.
406–424). Springer.
guided the author in completing this research. Thank you also
to Mr. Syirojudin and all staff of BMKG's Potential [12] Yamada, I., Masuda, K., & Mizutani, H. (1989). Electromagnetic and
acoustic emission associated with rock fracture. Physics of the Earth and
Geophysics Department for guidance, which has provided raw Planetary Interiors, 57(1–2), 157–168.
data along with its processing software. [13] Yumoto, K., Ikemoto, S., Cardinal, M. G., Hayakawa, M., Hattori, K.,
Liu, J. Y., … Widarto, D. (2009). A new ULF wave analysis for Seismo-
REFERENCES Electromagnetics using CPMN/MAGDAS data. Physics and Chemistry
of the Earth, Parts A/B/C, 34(6), 360–366.
[1] Ahadi, S., Puspito, N. T., Ibrahim, G., Saroso, S., Yumoto, K., & Muzli,
M. (2015). Anomalous ULF Emissions and Their Possible Association

You might also like