Search

On 21 November 2022, a MW 5.6. earthquake hit Cianjur area in West Java, Indonesia. The small magnitude earthquake claimed 329 people’s lives, and about 114,414 people were taking shelter in refugee camps as of 1 December 2022 according to the National Disaster Mitigation Agency (BNPB) of Indonesia. We produced maps of coseismic deformation and surface disturbance (a.k.a. Damage Proxy Map, DPM) using Synthetic Aperture Radar (InSAR) data acquired by the ALOS-2 satellite about 11 hours after the earthquake. We combine and analyze ground observations, coseismic deformation and surface disturbance maps derived from ALOS-2 and Sentinel-1 SAR data, seismic waveforms of the mainshock and aftershocks, high-rate GNSS data, and tiltmeter data to characterize the source parameters of the earthquake and damage caused by the strong ground motion, landslides, and liquefaction, and study the potential impact of the event on the geohazards of the area.

Accurate and precise location of earthquake sequence is critical to better understand seismotectonics, such as better delineation fault geometry and understanding of the rupture of the earthquakes. However, nearfield seismic observations are usually rare for such study. Here we study a unique dense nodal array data acquired by the deployment after the 31/08/2019 Mw5.5 crustal earthquake that is located ~50km to the west of Sagaing fault near Mandalay beneath ShweBo Central Myanmar Basin (CMB). The network, composed of 20 nodal stations with station spacing of ~5km, was deployed ~ 2 weeks after the mainshock and continuously recording for ~ 40 days. High quality waveforms containing clear P and S phase arrivals, and an interesting P-to-S phase converted at the basement of CMB were recorded for aftershocks. We applied a machine learning based automatic phase detection software (Earthquake-Transformer) to the dataset and detected 1143 events that were recorded by at least 3 stations. Double difference relocation of these aftershocks reveals a near E-W trending fault with a dimension of ~10km along strike and located between 7 to 12 km in depth. The strike of aftershock lineation is highly consistent with the focal mechanism derived from regional waveform inversion, indicating a left lateral strike-slip fault beneath CMB. Mainshock epicenter refined by a path calibration technique is located to the western edge of the seismicity, suggesting an eastward rupture directivity of the mainshock. Taking advantage of the P-basin-S converted phase at the basement of CMB, we constrained the thickness of the basin to be 5 ± 0.7 km. Strong strength of the P-basin-S phase requires sharp velocity change between the basin and bedrock. It is possible that the earthquake sequence is a result of small block rotation that has been taking place beneath the CMB due to the convergence of India plate. Another possible explanation is a conjugate fault system associated with 2012 Mw 6.8 Thabeikkyin earthquake sequence which ruptured close to Sagaing fault.

Lembang Fault is one the major fault in the Western region of Java, Indonesia, located just 10-15 km north of Bandung (capital city of West-Java). Although this fault has little to no significant historical of big earthquake records, the fault shows significant geomorphic evidence of its tectonic activity. Previous study of this geomorphic markers combine with paleoseismological trenching data and the geodetic study based on GPS campaign method suggest that this 29 km length- fault has a dominantly sinistral sense of movement with relatively very low slip rate. Even if the rate is small, the fault could potentially generate a M6.5 to M7 earthquake, strong and considerably very damaging to nearby Bandung city populous area. Starting April 2022, the National Agency for Research and Innovation (BRIN) of Indonesia and Earth Observatory of Singapore (EOS) collaboratively install four continuous seismo-geodetic stations in the vicinity of Lembang Fault to monitor seismic activity and active fault movement of the Lembang Fault and to get movement information when the Lembang Fault earthquake occurs. The collaboration also aimed to strengthen data-based and geoscience-based Earthquake Disaster Risk Reduction efforts, and become a model observatory that bridges from earth science to local government policies and complete public education from upstream to downstream. In the first several months since the installation, we have collected seismic data from teleseismic earthquakes recorded by our stations as well as a few other seismic stations around Lembang Fault. Waveforms from the earthquakes recorded have been analysed, we performed a joint inversion of teleseismic receiver functions and surface waves (H/V ratio). The preliminary results, based on the forward and the inversion method of receiver function, suggest the crustal thickness underneath the station is about 25-30 km. In addition, there are some variability in back azimuth 90-180 degree which may be caused by waves traveling through the nearby active volcano.

Accurate and precise location and focal mechanism of aftershocks is a fundamental topic in seismology. However, nearfield seismic observations are usually not available for high-resolution source studies, or even when they are available high frequency waveform analyses are rarely conducted to extract more information. Here we study a unique dense nodal array data acquired by the deployment in the source region of the 2019 Mw5.5 strike-slip earthquake in Central Myanmar Basin (CMB). The network, composed of 20 nodal stations with station spacing of ~5km, was deployed ~2 weeks after the mainshock for ~ 40 days. We applied a machine learning based algorithm (Earthquake-Transformer) to detect 667 events from the dataset. Double difference relocation reveals that these events are distributed between 7 to 16 km in depth with a near E-W trending horizontal distribution, which is consistent with the left-lateral fault plane solution of the mainshock. On the vertical component of most of the stations, we observed a strong phase between P and S arrival times. This is an S-to-P converted phase from a sharp velocity boundary between the basin and the bedrock. The best 1D velocity model constrained by 3-component waveform modelling suggests a sedimentary layer thickness of ~3.5km beneath the stations. To determine the focal mechanism of aftershocks, we conducted high-frequency (up to a few Hz) waveform inversions that result in high quality waveform fits hence robust focal mechanisms of ~ 100 aftershocks with Mw1-2. In these focal mechanisms, ~50% are strike-slip events, ~40% are thrust events and ~10% are normal events, all corresponding to NE-SW oriented compressive stress. The thrust events have strikes oriented mostly in NW-SE direction and have dip angles of ~ 45°. Highly diverse aftershock focal mechanisms suggest the fault system is likely immature. The thrust and normal events indicate that the mainshock rupture had branched into or activated nearby secondary faults, which allow the strike-slip fault to propagate and develop a more complex fault system.

We estimated the velocity structure beneath four three-component broad-band seismic stations and three short period stations in the Gede Volcano region, West Java, Indonesia by the receiver function (RF) technique jointly with H/V (Horizontal/Vertical) amplitude ratio of Rayleigh waves to constrain the 1D velocity structure around each station, and in the back-azimuth of the source considered, ultimately providing a 3D understanding of the velocity structure of the volcanic plumbing system. The stations are part of the collaborative seismological network between Center for Volcanology and Geological Hazard Mitigation (CVGHM) and Earth Observatory of Singapore (EOS). Our preliminary result suggest the depth of the Moho is about 30 km. Based on the back azimuth of the earthquakes, the receiver functions inversion and H/V amplitude radio from the earthquakes coming from the NE-SE direction exhibit strong negative signals between direct P and Pms phase correspond to a low velocity layer in the crust, predominating all broadband station stations. This is consistent with the result of previous study of travel time tomography. We observed very few earthquakes from SW-NW direction where the receiver functions show differently from the first group due to the waves travel to magmatic body. Combining results from this study and the tomography we aim is to further find the magmatic body under the volcano at shallow as well as deeper depths. Since we have 10-year data, we will be able to see the yearly evolution or changes of the magmatic system under the volcano.

Over the past decade since the start of collaboration between Earth Observatory of Singapore (EOS) and Centre for Volcanology and Geohazard Mitigation (CVGHM), Gede Volcano has experienced frequent earthquake swarms, re ecting dynamic volcanic and tectonic processes. This volcano is located in the West Java Province 60-km from Indonesian capital Jakarta. We located the earthquake using seismic stations on Gede and Salak Volcanoes operated since 2011. We recorded earthquake swarms in 2012, 2015, 2016, 2019 and 2021 with hypocentres located under Gede Volcano and episode duration range from one day to a few weeks. We examine earthquakes from different earthquake swarm episodes where several tens of earthquakes are closely clustered in space and time. To further explore the evolution of the swarm, we carefully and consistently pick P and S wave arrivals owing to the highly repetitive and similar nature of waveforms and relocate these similar earthquakes. The hypocentres were clustered 1 km to the East of Gede active crater with depth range 2-3 km below the highest volcano topography. When we compared locations of different episodes, the swarms in 2012, 2019 and 2021 showed highly similar locations suggesting the nondestructed earthquake source. This could be magma pathway or a pre-existing fracture on the volcano. We hypothesize that the swarm was triggered by hydrothermal uid into a preexisting fault system, prompting release of accumulated stress. The ongoing work explore source mechanism of the earthquakes utilizing full wave inversion.

Bangladesh is a densely populated country located on the north-eastern side of the India-Eurasia collision zone. The eastern half of the country is underlain by the seismically active Chittagong-Myanmar Fold and Thrust Belt, which forms the accretionary prism associated with the eastward-subducting Indian Plate, and is bordered to the north by the Dauki thrust fault, which bounds the southern margin of the actively rising Shillong Plateau. Two great earthquakes occurred here in 1897 (~Mw8.0) and 1762 (~Mw8.5). The seismic hazard in this region is poorly constrained due to questions about slip partitioning between the frontal and splay faults of the fold and thrust belt, whether faults will eventually slip at seismic speeds, and how shaking will be amplified and/or attenuated by the thick sedimentary cover. Seismic monitoring of the region can help to address these questions. In 2016, a collaborative effort between the Earth Observatory of Singapore and Dhaka University led to the installation of a network of 22 Lennartz 1-Hz (Short-period) and 6 Trillium Compact 120-s (Broadband) seismometer stations along the north-eastern (Sylhet region) and south-eastern (Chittagong region) parts of Bangladesh. The aim of this network is to monitor seismicity and to develop a better understanding of the subsurface structure in the region. This network has been actively running for over 4 years. The stations are spaced about 15 to 30 km apart, and are installed in either outdoor shallow vaults, or in above ground indoor locations on solid foundations, depending on local conditions. The stations are visited every 3 to 5 months for data collection, as well as repair and maintenance work. Here we discuss challenges faced and overcome during the deployment associated with environmental conditions (soil, plant growth, monsoon flooding, soil erosion, and cyclones) as well as arising from human and wildlife interference, political instability, and the COVID pandemic. We present various solutions that were implemented to mitigate or eradicate these problems. We have derived a preliminary catalog of earthquake locations consisting of >1100 events, and also detected >130 teleseismic events; event locations are currently being refined. Our data will provide a better understanding of the seismic activity and subsurface geology of the region.

Infrasound, sound below the range of human hearing, is becoming recognized as a valuable addition to local volcano monitoring networks. Independent of cloud cover and time of day, infrasound provides a source of information for a wide range of surface activity. However, the presence of liquid water at the source has a significant impact on the style and amplitude of explosion signals produced, and hence on the interpretation. Understanding the impact of water on these signals is essential to understand the range of appropriate infrasound applications. In April, 2020, an eruptive sequence at Anak Krakatau, Indonesia, was captured on local infrasound sensors. This sequence included the drying out of the lake that had been present above the vent since the collapse of the island in December, 2018. This case study offers an opportunity to examine in detail the changes in close range signal characteristics within a system that exhibits both wet and dry conditions. Initial results indicate that the signal characteristics are significantly influenced by the presence of water. This project aims to understand how these changes influence the results of different processing methods used in infrasound monitoring. As infrasound measurements can be used to not only identify eruptions, but estimate eruption source parameters like plume height, it is important to understand how the presence of water might change the outcome of different monitoring methods. Therefore this has implications for monitoring networks including volcanoes with crater lakes, and the design of data processing methods.

Mount Mayon in Albay province (Philippines) is an openly-degassing basaltic-andesitic stratovolcano, located on the northern edge of the northwest-trending OAS graben. Its latest eruptions were in Aug-Sept 2006 and Dec 2009. Mayon’s current status is PHIVOLCS’ level 1 with low seismicity dominated mostly local and regional tectonic earthquakes and continuous emission of SO2 from its summit crater. A research collaboration between the Earth Observatory of Singapore-NTU and the Philippine Institute of Volcanology and Seismology (PHIVOLCS) was initiated in 2009, aimed at developing a multi-disciplinary monitoring network around Mayon. Published version

The CALIPSO collaborative volcano monitoring system on the Caribbean island of Montserrat includes observations of strain at depths ∼200 m using Sacks-Evertson strainmeters. Strain data for the March 2004 explosion of the Soufrière Hills Volcano are characterized by large, roughly equal but opposite polarity changes at the two near sites and much smaller changes at a more distant site. The strain amplitudes eliminate a spherical pressure (Mogi-type) source as the sole contributor. The initial changes are followed by smaller recoveries, but with differing relative recovery magnitudes. This dissimilarity requires a minimum of two pressure sources, which we model as a deep spherical pressure source and a shallow dike. The spherical source is fixed at the location derived from data for the massive dome collapse in July 2003. We solve for the best fitting dike plus sphere source combination. The dike geometry is consistent with earlier interpretations of dikes based on GPS data and other lines of evidence. Published version

We use well-documented time histories of episodic GPS surface deformation and efflux of compressible magma to resolve apparent magma budget anomalies at Soufrière Hills volcano (SHV) on Montserrat, WI. We focus on data from 2003 to 2007, for an inflation succeeded by an episode of eruption-plus-deflation. We examine Mogi-type and vertical prolate ellipsoidal chamber geometries to accommodate both mineralogical constraints indicating a relatively shallow pre-eruption storage, and geodetic constraints inferring a deeper mean-pressure source. An exsolved phase involving several gas species greatly increases andesite magma compressibility to depths >10 km (i.e., for water content >4 wt%, crystallinity ∼40%), and this property supports the concept that much of the magma transferred into or out of the crustal reservoir could be accommodated by compression or decompression of stored reservoir magma (i.e., the “magma-sponge”). Our results suggest quasi-steady deep, mainly mafic magma influx of the order of 2 m3s−1, and we conclude that magma released in eruptive episodes is approximately balanced by cumulative deep influx during the eruptive episode and the preceding inflation. Our magma-sponge model predicts that between 2003 and 2007 there was no evident depletion of magma reservoir volume at SHV, which comprises tens of km3 with radial dimensions of order ∼1–2 km, in turn implying a long-lived eruption. Published version

Vulcanian explosions with plumes to 12 km occurred at Soufrière Hills volcano (SHV) between July 2008 and January 2009. We report strainmeter and barometric data, featuring quasi-linear strain changes that correlate with explosive evacuation of the conduit at rates of ∼0.9−2 × 107 kg s−1. July and January explosion-generated strains were similar, ∼20 nanostrain at ∼5 km, and interpreted as contractions of a quasi-cylindrical conduit, with release of magmastatic pressure, and exsolution-generated overpressure of order 10 MPa. The 3 December 2008 event was distinctive with larger signals (∼140–200 nanostrain at 5–6 km) indicating that a rapid pressurization preceded and triggered the explosion. Modeling suggests a dike with ENE trend, implying that feeder dikes at SHV had diverse attitudes at different times during the eruption. All explosions were associated with acoustic pulses and remarkable atmospheric gravity waves. Published version