Peak Ground Velocity (PGV) estimation of Mw 5.6 Cianjur earthquake November 11, 2022 using Global Positioning System (GPS)

The primary purpose of GPS is to be used for positioning, navigation, and timing (PNT). In the last decade, research on GPS for earthquake hazards has increased rapidly, such as for tectonic deformation, GPS seismology, modeling earthquake source parameters, estimation of magnitude, etc. In this study, following the Mw 5.6 Cianjur earthquake on November 11, 2022, we estimate the peak ground velocity from the CJUR GPS station, which is located about 15 km from the epicenter. We compute GPS velocity from the high-rate GPS receiver independent exchange (RINEX) data and broadcast ephemeris. Using a narrow line combination of L1 and L2 observations, we invert velocity epoch by epoch. The result shows the CJUR GPS station has a peak ground velocity of about 11.5 cm/s and equal to VI MMI scale. The peak ground velocity from the GPS station could be important for deriving the earthquake shakemap. By increasing the number of GPS stations in Indonesia, it is very important to support earthquake hazard monitoring.


Introduction
On November 21 s t, 2022 at 13:21:10 Western Indonesia Time (06:21:10 UTC) the Mw 5.6 Cianjur earthquake occurred northwest of Cimandiri Fault (Figure 1).The epicenter is located at latitude = 6.87 o S, longitude = 107.01o E, and depth = 11 km (https://inatews.bmkg.go.id).The mainshock of this event felt in Cianjur city up to V-VI MMI (Modified Mercalli Intensity), IV-V MMI in Garut and Sukabumi, III MMI in Cimahi, Lembang, Bandung city, Cikalong Wetan, Rangkasbitung, Bogor and Bayah, II-III MMI in Tangerang Selatan, Jakarta and Depok (Figure 2) [1].The Meteorology, Climatology and Geophysics Agency (BMKG) reported 424 aftershocks recorded until December 17 t h, 2023.Based on the National Disaster Management Authority (BNPB) on December 12, 2023, this event caused 334 people dead, and 35.601 damaged buildings.The utilization of the high-rate GNSS data are widely used to estimate displacement due to the earthquake and potential used to support earthquake early warning system (EEWS) [2].In the recent studies show the availability of high-rate GNSS data to determined earthquake magnitude rapidly.However, research on the ground velocity waveform using high-rate GNSS data to earthquake studies or monitoring is still lacking..The recent study [3], show the consistency betwwen velocity from GNSS observations with the ground motion models and macroseismic intensity observations.The residual standard deviation of ground motion model for Peak Ground Velocity (PGV) from GNSS is about 0.58 [3].
In this research, we applied the "variometric approach" proposed by [4,5] to capture velocity waveform from high-rate GNSS due to the Mw 5.6 Cianjur earthquake.Due to the lack of GNSS stations, we only use one GNSS station that located close to the epicenter.The velocity waveform can be used as complementary constraints to ShakeMaps and to determine the earthquake magnitude using PGV scaling law [6].

Methods
In this research, we use the continuous GPS high-rate data provided by BIG.The nearest cGPS site is CJUR station, which is located about 5 km to the southeast of the epicenter.CJUR station was built at the top of the building and equipped with Septentrio SPT POLARX5 GNSS receiver with the LEIAR25 GNSS antenna.The station recorded GNSS data at 1 Hz intervals and send the data in real-time to the BIG processing center using Virtual Private Network. Figure 3 shows the monumentation of CJUR GNSS site (srgi.big.go.id).The "variometric approach" is performed a single time difference on the GNSS phase observation and orbital position [3].In this methods, we taking a difference between the current time observation and the prior time observation depend on the interval of observation [3].This method was implemented on Satellite Navigation-Derived Instantaneous Velocities (SNIVEL) software package [7].Using SNIVEL we derived the velocity waveform high-rate GNSS data with 1 Hz intervals.We used a broadcast ephemeris and high-rate Receiver Independent Exchange (RINEX) GPS data to derived the velocity waveform.The SNIVEL software needs the L1 and L2 phase observables GPS signal.With the narrow lane combination from the L1 and L2 signal, velocity waveform was derived on an epoch by epoch observation.We processed 5 minutes data starting from 06:21:00 UTC.

Result and discussion
Figure 4 shows the velocity waveform in the north-south, east-west, and vertical component starting from 06:21:00 -06:26:00 and the Ground Velocity (GV) waveform.Within 5 minutes length of the data, the high-rate GPS data captured the ground velocity due to the Cianjur earthquake.The maximum velocity which recorded by CJUR GNSS site is about 11.5 cm/s (see Figure 4).The maximum value reached about 4s from the origin time of the earthquake.Regarding the ShakeMap documentatiom ( [8]), the ground velocity 11.5 cm/s equal to VI MMI scale.This MMI value is quite different with the color scale in the Figure 1.Our suggestion, this discrepancy due to the calculation of the shakemap in the Figure 1 used model and update with accelerometer data and not included in situ observation.From this result, we show that the high-rate GPS data could be complementary constraints to ShakeMap.

Conclusion
This research demonstrates the ability to use a high-rate GPS data to determine the ground velocity waveform by post processing method.The result shows the velocity in the horizontal component of CJUR GNSS site are significant, while the vertical velocity

Figure 2 .
Figure 2. Location of CJUR GNSS site and the epicenter of Cianjur earthquake (BMKG and USGS)

Figure 4 .
Figure 4. Velocity waveform of CJUR GNSS site.a) is the velocity waveform in the nourth-south component.b) is velocity waveform in the east-west component.c) is is velocity waveform in the vertical component.d) is Ground velocity (GV).Red line indicate the earthquake origin time in GPS time.