Abstract Number

T4.1-P18.

Preliminary Study of the CTBTO’s Seismic Station Characteristics

Using Spectral Analysis Method in Indonesia

B. Wibowo, N. Heryandoko, S. Rohadi

Abstract.We analysed recording waveform of six seismographics stations which is part of CTBTO's seismic

network in Indonesia. The analysis using the spectral analysis method conducted to determine the charac-

teristics response of each seismographic station. We analyzed background noise level of sites using Power

Spectral Density (PSD) and Probability of Density Function (PDF). The result of spectral analysis indicate

that PSI station (Parapat, Sumatera) has the lowest background noise level, so its has highest Signal To

Noise Ratio (SNR). This station has best recording of nuclear explosion and earthquake event compare to

recording of other station. This good quality of recording signal because the seismometer located on the

representative bedrock and the site good protected from the ambient or environmental noise. Otherwise,

LEM station (Lembang, Bandung) has the highest background noise level and has lowest SNR. LEM sta-

tion located near the Tangkuban Perahu Mountain that one of active vulcano in Bandung. Activity of the

vulcano may create disturbance noise to the recording signal in Lembang station (LEM). The significance

noise also may because of human activity around this site. We also conduct time frequency analysis by

comparing frequency content between nuclear explosion and the earthquake event. The results indicate

that explosion have more clear and focus of spectogram than that of an earthquake events, this character-

istic can be used to identified nuclear explosion accurately.

SEKOLAH TINGGI METEOROLOGI,

KLIMATOLOGI, DAN GEOFISIKA

S T M K G

Figure 4. Frequency response from different type of

seismometer used by CTBTO’ stations in Indonesia.

Figure 1 & 3. The Location of CTBTO’s Auxiliary Sta-

tions in Indonesia, indicate by red triangle and nuclear

explosion and earthquake (yellow star).

.

Table 1. The name, seismometer type, and geo-

graphical coordinat of CTBTO’s Auxilliary Sta-

tions in Indonesia.

(a) (b)

Figure 2. (a) Parameters from the North Korea’s 2013 Nuclear

Test (http://earthquake.usgs.gov/ earthquakes), and (b) Earth-

quake parameters from the Hokkaido’s Earthquake (geofon).

INTRODUCTION We analyzed background noise level of sites using Power Spec-

tral Density (PSD) and Probability of Density Function

(PDF).

Time Frequency Analysis to describe frequency content of nu-

clear event, as well as earthquake event, recorded by these sta-

tions

Estimating the level of effectiveness the stations regarding de-

tection of nuclear and earthquake events.

Using 2 datasets for this research. Each dataset consisted by 6

vertical component records:

First, we used seismic signal recording from 6 stations with

duration about 7 days from April 1st 2015, 00:00:00 UTC

to April 8th 2015, 00:00:00 UTC

Second, we proccessed nuclear event data recorded by

CTBTO’s stations in Indonesia. The North Korea’s 2013

Nuclear Test was held on February 12th 2013 (Figure 2.a).

We also proccessed an earthquake event that occur within

similar distance and magnitude, comparing to nuclear event.

The Earthquake was occured in Hokkaido, March 25, 2015

(Figure 2.b and Figure 3).

All of the data processing on both dataset used Obspy, a seismic

waveform analysis software that based on Python Program.

The Power Spectral Density, Probability of Density Function, Day vs Night Amplitudo Ratio

(i)

(ii) (iii)

(iv)

(v) (vi)

Figure 5. The Result of Power Spectral Density and Probability of Density Function of (i) SIJI, (ii) PSI, (iii) LEM, (iv) KAPI, (v) JAY, (vi) BATI Station.

Figure 6. The Mean Value of PSD from from 6 Stations Figure 7. The ratio between day and night data from all stations

are plotted in a graph.

Figure 9. The graph be-

tween distance from 6 sta-

tions and its time duration

from the North Korea’s

2013 Nuclear Event.

Time Frequency Analysis (Spectrograms) of The North Korea’s 2013 Nuclear Event

(i) (ii)

(iii) (iv)

(v) (vi)

Figure 8. The Time Frequency Spectrogram and Waveform of The

North Korea’s 2013 Nuclear Event of (i) SIJI, (ii) JAY, (iii) KAPI,

(iv) BATI, (v) PSI and (vi) LEM station, respectively.

Time Frequency Analysis’s (Spectrograms) of The Hokkaido Earthquake

(i) (ii)

(iii)

(iv)

(v) (vi)

Figure 10. The Time Frequency Spectrogram and Waveform of

Hokkaido Eartquake of (i) SIJI, (ii) BATI, (iii) JAY, (iv) PSI, (V)

KAPI, and (vi) LEM Station, respectively.

Figure 11. The graph between

distance from 6 stations and its

time duration from the Hokkai-

do Earthquake.

RESULT AND DISCUSSION

LEM Station has the highest background noise level comparing to other stations, otherwise PSI Station has the

lowest one

The signal from nucle-

ar event clearly ap-

peared at KAPI, BATI,

and PSI stations

KAPI and PSI have

same distance (about

47°) to the epicenter

and BATI Station have

same distance to epi-

center (about 52°) with

LEM station, but

BATI station has high-

er SNR rather than

LEM Station.

2 stations has high SNR re-

garding this event, PSI Sta-

tion (10.IV)) and LEM sta-

tion (10.VI)

all station we can still rec-

ognize the signal from this

event better than nuclear

event as shown before,

caused by situation that at

early in the morning, noise

level caused by human ac-

tivity had not increase yet.

LEM station, which has

higher noise level, we can

still see the signal clearly,

but it rather has shorter fre-

quency range compared

with PSI Station.

The writers thank The State College of

Meteorology, Climatology and Geophysics

(STMKG), The Indonesian Meteorology,

Climatology, and Geophysics Agency

(BMKG), for the chances and support dur-

ing the research and The CTBTO’s Sci-

ence and Technology Conference 2015’s

Commitees for the opportunities to present

our research result in their event.

ACKNOWLEDGEMENT

REFERENCES

Daniel E. McNamara, Raymond P. Buland.

Ambient Noise Levels in Continental

United States. (Bulletin of Seismological

Society of America, Vol.94, August 2004)

Daniel E. McNamara, R. I. Boaz. Seismic

Noise Analysis System Using Power Spec-

tral Density Probability Density Func-

tion: A Stand Alone Software Package.

(Open-File Report 2005-1438, USGS,

2005)

Jon Patterson. Observation and Modelling

of Seismic Background Noise. (Open-File

Report 93-322, USGS, 1993)

Eléonore Stutzmann, Geneviève Roult, Lu-

ciana Astiz. GEOSCOPE Station Noise

Levels. (Bulletin of Seismological Society

of America, June 2000)

Moritz Beyreuther, Robert Barsch, Lion

Krischer, Tobias Megies, Yannik Behr,

Joachim Wassermann. Obspy: A Python

toolbox for seismology, a Data Perspec-

tive. (Ludwig Maximilians University).