Cianjur M5.6 Earthquake Aftershock Survey (CEAS) Using ITB Temporary Seismographs

Western Java, part of the Sunda Arc, is an active tectonic region with high seismicity. The tectonic activity has produced active faults which are located near densely populated areas. On November 21, 2022, a significant earthquake rocked the Cianjur region, West Java, causing severe damage and loss of life. In this study we acquired aftershock data which was generated from the Cianjur M5.6 earthquake. Twenty stations consisting of 19 Smartsolo IGU-16HR 3C 5 Hz Seismometer (short-period type) and 1 Guralp 6TD 30s to 100 Hz Seismometer (broadband type) were deployed to monitor aftershocks from 22 November to 23 December 2022. The seismograph stations were deployed covering the earthquake source area and tectonic features in the Cianjur region. The results show that the aftershock events is clearly recorded at 20 stations with an average difference between the arrival times of the S waves and P waves (Ts-Tp) of 2 seconds. A total of ±500 earthquake events were obtained consisting of ±4000 P wave phases and ±3000 S wave phases with a Vp/Vs ratio value of 1.72 and RMS error less than 0.1 second. The Cianjur M5.6 earthquake was classified as a shallow crustal seismicity with a strike slip fault mechanism, hypocenters were generally distributed at a depth of 3-10 km.


Introduction
West Java is a region with complex tectonic characteristics.Located north of the megathrust subduction zone, where the Indo-Australian oceanic plate subducts beneath the surface of Sundaland, on the southern boundary of the Eurasian Plate [1].Tectonic activity from the subduction of the Indo-Australian Plate under the Eurasian Plate has resulted in several active faults in West Java including the Cimandiri Fault, Lembang Fault and Baribis Fault, which are located close to densely populated areas [2].The existence of these local faults has an impact on high seismic activity in West Java and its surrounding area.
2 On November 21, 2022, a significant earthquake rocked the Cianjur, West Java and surrounding area with the 5.6 magnitude and 10 km focus depth [3].The Meteorological, Climatological and Geophysical Agency (BMKG) stated that the mainshock event had generated quite strong shocks (MMI III-VIII) around the earthquake source area in Cianjur, West Java.The BMKG further stated that this earthquake was included in the mainshock-aftershock type of shallow crustal earthquake, namely the main earthquake which was then followed by a series of aftershocks [4].The National Disaster Management Agency (BNPB) reported that the incident caused damage to buildings and public facilities, especially in the Cianjur area, West Java.Based on information from BNPB on 27 December 2022, the Cianjur M5.6 earthquake damaged more than 56,000 buildings and killed 602 people.There were more than 114,000 people, originating from 16 districts in Cianjur, who were evacuated due to the high intensity of aftershocks in the area.The interesting things about Cianjur earthquake, the structure which caused this earthquake is still being debated considering that the epicenter of the earthquake was quite far from the Cimandiri fault and also Gede volcano as shown in Figure I.1.Where the epicenter is estimated to be at a distance of 20 km north of the Cimandiri Fault Rajamandala Segment, this fault has a left-lateral slip mechanism [5].There was possibility the Cianjur earthquake was produced by unidentified fault.Based on historical records of seismicity, this area experienced devastating earthquakes in 1844, 1879, 1900 [6] and also 1962, M5.5 1982 and M5.4 2000, Cianjur earthquakes [7].Along with the series of earthquake events that occurred in Cianjur, West Java, it certainly has implications for changes in tectonic conditions in the region.Therefore, to provide further information of the recent tectonic dynamics, studies are needed to identify the characteristics of the earthquake mechanism, so further investigation is needed regarding the complexity of the structure and the source characteristic of the Cianjur earthquake.

Data Acquisiton
The data used in this study is aftershock recording data using the ITB temporary seismograph network deployed after the Cianjur M5.6 earthquake.In this research, 20 stations were used with a data recording period of 1 month from November 22 to December 23, 2022.There are 2 types of seismometers used in this study, including 1 unit of broadband seismometer Guralp 6TD 30s to 100 Hz 2 x 1200 V/m/s and 19 units of Smartsolo seismometer (short-period) IGU-16HR 3C 5 Hz.The seismograph was installed in two stages.The first stage was carried out on November 22, 2022, with 6 seismometer stations installed.While the second stage will be carried out on November 26, 2022, 14 seismometer stations will be installed.Data recording time ranges from 26-31 days.The following is a data record table.

Data Processing
The stages of processing the earthquake aftershock data include several stages.

Waveform Identification and Picking P & S Wave
We have identified the aftershock events which were recorded at a minimum of four stations and pick the arrival time of the P and S waves at the event.This process is done manually using Seisgram2K Software.

Velocity Model Parameter Setting
Second, the 1D seismic velocity model is determined according to the characteristics of the study area.
In this case, the ANT velocity model from Rosalia (2022) has been used, which has modeled the seismic velocity structure in each subsurface layer in the western part of Java [8].

Hipocenter Determination
Third, we determine the location of the hypocenter using the non-linear method of the NLLoc program [9].This program implements a probabilistic inversion approach, a direct search method that the algorithm employed by Tarantola and Valette [10].In addition, the hypocenter mapping is determined using GMT program.

Result and Discussion
Events recorded by temporary seismograph stations have been identified.Followed by determining the arrival time of the P and S waves by picking the first motion for each wave.P and S wave picking is carried out at a minimum of four stations that record earthquake events.From the picking results, more than 500 events were recorded and the arrival times of the P and S waves were successfully determined.The process is carried out using the Seisgram2K software.An example of a successfully recorded seismogram is shown in the following figure.Furthermore, the arrival time observation data of the entire event is plotted into a Wadati Diagram to see the linear relationship between the phases.From the plotting results, it can be seen that the gradient of the arrival time distribution trend is 0.7, so that the value of the Vp/Vs ratio is 1.72.It shows the Pwave and S-wave arrival times that we observed fit the linear relationship, and the quality is good enough for earthquake hypocenter determination.The distribution of the difference in the arrival time of the P and S waves on average for all events is dominated by events with a difference in arrival time of less than 2 seconds.This shows that the recorded event location is very close to the seismometer station.Beside that, this can minimize the error value obtained from determining the hypocenters.The results of determining the location of the hypocenter aftershock for the Cianjur M5.6 earthquake 2022 show that the non-linear method (NonLinLoc) has good statistical abilities, seen from the distribution of residual RMS values for each specified event.In general, the residual RMS value is under 1 second with the highest RMS value being under 0.1 second.
From the results of hypocenter determination, each parameter was obtained, such as origin time, azimuth gap, RMS error and location of the epicenter.The location of the earthquake hypocenter was plotted using the GMT program.The distribution of hypocenters is shown in figure below.Based on hypocenter determination using the NonLinLoc program, more than 500 aftershock events were successfully recorded by the ITB temporary seismograph stations.Total phases obtained are ±4000 P-arrival time phases and ±3000 S-arrival time phases.The hypocenters of Cianjur aftershock were also distributed at a depth of between 3-10 km.Aftershock cluster was located at the north of the M5.6 mainshock (shown by a yellow star).It can be seen that the seismicity pattern of the aftershocks for the Cianjur M5.6 earthquake has two distinct segments, including the North-South segment and the West-East segment which are consistent with the focal mechanism parameters with a strike slip fault mechanism.In addition, the North West-South East (NW-SW) segment looks more dominant based on the event distribution.In the previous study, moderate earthquake could affect aftershock evolution, for example the seismogenic fault of Alxa M5.8 earthquake associated with a NEE-striking blind fault that becomes steeply dipping toward the south [11].
The Cianjur aftershock map shows that this earthquake is an earthquake originating from a fault that has not been identified before (blind fault), not related to the Cimandiri fault segment.There is a possibility that this earthquake caused fault conjugation, but this finding needs to be confirmed by several advanced method to gain comprehensive understanding.Futhermore, the findings show that this earthquake did not cause volcanic activity from Gede volcano.Gede volcano is monitored visually and instrumentally from the Volcano Observation Center in Cianjur.The latest data shows that Gede volcano erupted in 1957 with the characteristics of an explosive eruption in the form of thick gray to black ash eruptions, with an eruption column height of 3 km above the peak, but the activity level of Gede volcano at this time is Level I (Normal).The latest developments in Gede volcano activity following the Cianjur earthquake from 21 November 2022 to 27 November 2022 show that there has been no increase in the number of volcanic earthquakes since the Cianjur earthquake occurred on 21 November 2022 to 27 November 2022.Only 2 deep volcanic earthquakes were recorded with an amplitude of 21-45 mm and an earthquake lasting 5-8 seconds [12].The location of Gede volcano, which is relatively close to the center of the Cianjur earthquake, is feared to affect the magmatic activity of Gede volcano.Linde and Sacks (1998) state that volcanoes can be triggered by eruptions up to a distance of 100 km from the epicenter of an earthquake with a magnitude < M 8.0 and up to a distance of 50 km for earthquakes with a magnitude ≥M8.0 [13].Manga and Brodsky (2006) estimate that only 0.4% of explosive volcanic eruptions occur a few days after a major earthquake [14].Whereas Sawi and Manga (2018) define that there are volcanoes that are sensitive (ever triggered by an eruption) and insensitive (never been triggered by an eruption).Sensitive volcanoes tend to be more active than those that are not sensitive and erupt more frequently, only occur in subduction zones, and are dominated by andesite and basaltic andesite eruptions.Eruptions triggered by earthquakes do not show a relationship between magnitude and distance.Meanwhile, there is a 5-12% increase in eruptive activity in the 2 months to 2 years after the main earthquake [15].

Conclusion
Based on data acquisition, it shows good results with clear waveforms recorded at 20 stations.The difference between the arrival time of the P and S waves (Tp-Ts) is ≤ 2 seconds.A total of ±500 earthquake events were obtained consisting of ±4000 P wave phases and ±3000 S wave phases with a Vp/Vs ratio value of 1.72 and RMS error less than 0.1 second.From hypocenter determination shows that the aftershock of Cianjur M5.6 earthquake was dominated by shallow crustal seismicity with a depth of 3-10 km, it was produced by a strike slip mechanism.For the case of Gede volcano, there was no significant increase in volcanic seismic activity following the Cianjur M5.6 earthquake.However, the future research is necessary to consider the possibility of volcanic and tectonic interaction especially in West Java region, Indonesia.

Figure 4 .
Figure 4. Determination of the P and S wave arrival time for one of the events on Desember 3, 2022.The green line shows the arrival time of P waves and the blue line shows the arrival time of the S waves.In determining the arrival time of P and S waves, P waves can be easily observed on the vertical component (Z) seismogram, while S waves can be observed on the horizontal component (N and E) seismograms.The quality of picking results can be controlled by using Wadati Diagrams.The diagram depicts the arrival time of the P wave (Tp) as the X-axis and the difference between the arrival time of the S wave and the arrival time (Ts-Tp) as the Y-axis to determine the Vp/Vs ratio.The following is an example of the results of the Wadati Diagram plotting in one of the picked events.

Figure 5 .
Figure 5. Wadati Diagram of P and S wave picking results in one of the earthquake events

Figure 6 .
Figure 6.Diagram of the overall earthquake events.The Vp/Vs ratio value is 1.72.

Figure 7 .
Figure 7. Distribution of the difference value between the arrival times of the P and S waves on average for all events

Figure 8 .
Figure 8. Distribution of the RMS residual error for all events

Table 1 .
Aftershock Data Recording