Borehole Seismic Network Research Articles

Spatio-temporal variations of shallow seismic velocity changes in Salton Sea Geothermal Field, California in response to large regional earthquakes and long-term geothermal activities

We measure spatio-temporal variations of seismic velocity changes in Salton Sea Geothermal Field, California based on cross correlations of daily seismic traces recorded by a borehole seismic network from December 2007 to January 2014. We find clear co-seismic velocity reductions during the 2010 M 7.2 ​El Mayor–Cucapah, Mexico earthquake at ∼100 ​km further south, followed by long-term recoveries. The co-seismic reductions are larger with longer post-seismic recoveries in higher frequency bands, indicating that material damage and healing process mostly occurred in the shallow depth. In addition, the co-seismic velocity reductions are larger for ray paths outside the active fluid injection/extraction regions. The ray paths inside injection/extraction regions are associated with smaller co-seismic reductions, but subtle long-term velocity increases. We also build 3D transient water flow models based on monthly injection/extraction rates, and find correlations between several water flow parameters and co-seismic velocity reductions. We interpret the relative lack of co-seismic velocity changes within the geothermal region as unclogging of fracture network due to persistent fluid flows of geothermal production. The long-term velocity increase is likely associated with the ground water depletion and subsidence due to net production.

Amplification Behaviour of Compressional Waves in Unconsolidated Sediments

Similar to horizontal earthquake motions, vertical motions are amplified depedent on the local site conditions which can be critical for the safety of certain structures. Production of natural gas in Groningen, the Netherlands, results in reservoir compaction causing low magnitude, shallow earthquakes which are recorded with a borehole seismic network. These recordings form an excellent data set to understand how shallow unconsolidated subsurface geology influences the amplification behaviour of compressional waves (P-waves). First, we present borehole and single-station techniques (amplification factors, empirical transfer functions (ETF) and V/H spectral ratio implementations) to quantify vertical amplification. We show that vertical-wave incidence is a reasonable assumption. All techniques are capable of emphasising the sites with strong amplification of vertical ground motion during an earthquake. Subsequently, we compare ETF with single-station methods with the aim to develop proxies for vertical site-response using spectral ratios. In a second step, we link vertical site-response with shallow subsurface conditions, like the P-wave velocity and peat content. To better understand the amplification mechanisms, we analytically simulate P-wave propagation. In the simulations, we compute synthetic transfer functions using realistic subsurface conditions and make a comparison with the ETF. The simulations support the hypothesis that thin layers of shallow gas, originating from the Holocene peat, result in wave amplification. We observe strong vertical site-response in particular in the eastern part of Groningen, with industrial facilities and pipeline infrastructure in the region. Here, if high vertical amplifications are persistent at large earthquake magnitudes, appreciable levels of vertical loading may be expected. This study demonstrates that vertical motions should be assessed separately from horizontal motions, given that the amplification behaviour of P-waves is affected by distinctive mechanisms.

Orientation Corrections of a Borehole Seismometer Network in Taiwan Using Teleseismic Earthquakes

Abstract The Central Weather Bureau, Taiwan, has deployed a borehole seismic network of 62 stations nationwide in Taiwan since 2008. Each borehole station includes a force-balance accelerometer on the surface, and a combination of an accelerometer and a broadband seismometer in the deep hole. The noise analyses have shown significant improvement in the performance of the borehole instruments compared with the instruments deployed on the surface. We propose a method to obtain robust orientations, as well as the uncertainty for both the borehole accelerometer and broadband seismometer using long-period waves (>2 s) after the S-wave arrival time of three mid-large teleseismic events and one local event. Overall, we successfully calibrated the orientations of the borehole sensors and found that the significant azimuth change for some borehole seismometers was always due to sensor maintenance. Based on the results of this study, we conclude that the long-period waveforms generated by a teleseismic event, wavelength of which is four times longer than the borehole depth, are useful for borehole orientation calibration. The correlation coefficients of the three components for both surface and borehole sensors larger than 0.9 are appropriate. The magnitude and epicenter distance of the event and on-site noise level are the factors that affect the quality of calibration. In addition, we conclude that the method of orientation calibration we propose is helpful for data quality control, including data timing and sensor failure. With the supplement of this azimuth estimation information, we expect that these high-quality borehole data in Taiwan will improve seismological research and will benefit the society.

Simultaneous estimation of site response and source parameters of reservoir-triggered earthquakes using data from the borehole seismic network in the Koyna–Warna seismic zone, Maharashtra, India

Simultaneous estimation of site response and source parameters of reservoir-triggered earthquakes using data from the borehole seismic network in the Koyna–Warna seismic zone, Maharashtra, India

Relative Locations of an Earthquake Sequence Recorded during June 2017 on the Koyna–Warna Borehole Seismic Network of Western India

ABSTRACTIn the Koyna–Warna region, western India, an enormous number of microearthquakes was detected automatically on borehole records. Most of these events could not be identified on the surface network by a routine approach based on visual inspection primarily due to signal attenuation and the presence of noise. In this work, we implemented an automatic detection workflow to analyze the time series of an earthquake sequence that has clear foreshock and aftershock activity associated with an Mw 4.0 earthquake that occurred on 3 June 2017. Further, we applied a nested grid-search algorithm to constrain the absolute earthquake locations. For about one month of data, a total of ∼1500 earthquakes were detected based on the automatic detection process, out of which ∼1000 earthquakes were locatable. All event detections, P-wave and S-wave phase readings were manually inspected and refined to ensure their quality. Previously, only about 435 events were well located based on the visual inspection approach for the same time period. Also, we analyzed repeated earthquakes based on waveform similarity leading to an improvement in the relocations of earthquakes of the aforementioned earthquake sequence. The relocated seismicity aligns parallel to a deep-reaching lineament derived from recent investigations using airborne light detection and ranging measurements.

Recent Seismicity Patterns and Microearthquake Activity on an Active Intraplate Fault System at Koyna-Warna, Western India

ABSTRACT Recent seismic activity of the Koyna region, western India is reported in this article. The Koyna region is considered to be one of the premier sites of reservoir triggered seismicity worldwide and it has been quite active since the initial impoundment of Shivaji Sagar Lake, north of Koyna Dam, during 1962 and nearby Warna Reservoir in 1985. Recently, a borehole seismic network consisting of 6 seismometers is deployed below the Deccan Traps to monitor the earthquakes in the study region, which recorded a large number of microearthquakes. A dense network of 23 surface broadband seismometers was also operative during the study period. A total of 2478 earthquakes of ML -0.8 to 3.7 recorded by the borehole seismic network, occurred during January 2016 to May 2017. Seismicity patterns in the recent time, which includes a few new zones of intense seismicity clusters in the vicinity of the borehole locations, are identified. Reservoir water levels are found to be strongly associated with the seismicity patterns. The seismicity is mostly concentrated in the vicinity of the Donachiwada fault and Warna during January – May, whereas it spreads during June – December. ‘b’ values have been estimated and found to be varying from 0.74 to 0.93. Seismicity continues to be present in the identified block where the pilot borehole has been drilled.

Seismicity related to the hydraulic stimulation of GRT1, Rittershoffen, France

The Rittershoffen deep geothermal reservoir, in Northeastern France, is well characterized and has been extensively studied by a multidisciplinary approach. A hydraulic stimulation for the development of the geothermal reservoir was performed in June 2013. This injection of fluid led to seismic activity which was closely monitored by a dedicated set of seismic stations. The seismic sequence started during the injection but showed an unusual long quiet period of 4 d after shut-in before the occurrence of a second swarm of events. Here we take the opportunity of this well-monitored activity to gain insight into the geomechanical factors favouring the development of induced earthquakes. We apply a template matching approach and a relative relocation procedure to obtain a precise estimate of the geometries of the activated structures. Our approach shows that the induced events during the injection took place on two parallel planar structures. It shows that details of the seismicity generally obtained from borehole seismic network are achievable from surface network when an appropriate analysis is performed. The development of this induced seismicity is in good agreement with the known stress field and failure criterion proposed for the reservoir. In particular, the orientation of the activated structure, associated focal mechanisms and the overpressure needed to initiate the seismic activity are all in line with the geomechanical model of the area. The swarm of delayed events, 4 d after shut-in, can be well explained by considering an aseismic slip on the imaged fault and the related static stress transfer. We therefore suggest that the ability to monitor local slow aseismic movements at depth, in conjunction with precise tracking of the seismicity, is of primary importance to understand induced earthquake activity.

Location of the pilot borehole for investigations of reservoir triggered seismicity at Koyna, India

Artificial water reservoir triggered earthquakes are now known to have occurred at over 120 sites globally. The part played by the reservoirs in triggering is not exactly known due to lack of near field observations of triggered earthquakes. Koyna, located near the west coast of India, where triggered earthquakes have been occurring since 1962 provides an excellent site for near field observations of the target M≥2 earthquakes. A 6 borehole seismic network has been deployed recently in the Koyna region at depths of 981–1522m to improve the hypocenter locations. During May–December 2015, a total of 1039 earthquakes of ML≥0.5 were located using the borehole seismic network. The region is also monitored through a dense network of 23 surface broad-band stations. Our analysis indicates a significant improvement in the estimation of absolute locations of earthquakes with errors of the order of ±300m, combining both the networks. Based on seismicity, and logistics, a block of 2×2km2 area has been chosen for drilling the first pilot borehole of ~3km depth, where M≥2 earthquakes have been occurring frequently since 2005.

Borehole Seismic Network at Koyna, India

The Koyna region located near the west coast of India has been continuously having artificial water‐reservoir‐triggered earthquakes since the impoundment of Koyna Dam in 1962. To improve the accuracy of locations of these triggered earthquakes, a unique network of eight borehole seismometers surrounding the seismicity was designed. Five of these have been installed at depths varying from 1134 to 1522 m, well below the Deccan basalt cover. As it is difficult to install the borehole seismometers with desired orientation of the horizontal components, amplitudes of P waves were used to decipher the true orientations after installation. Analysis of data for the period March–June 2015 demonstrated that there is a remarkable increase in the number of earthquakes located with the help of borehole seismometers. Using the existing network of 21 broadband surface seismometers, 184 earthquakes of M L≥0.5 were located; however, with the borehole seismometers, the number of the located events increased to 888 such events. The data also indicate an increase in seismic activity corresponding to the period of increase in the rate of emptying of the Koyna and Warna reservoirs.

Relationships among seismic velocity, metamorphism, and seismic and aseismic fault slip in the Salton Sea Geothermal Field region

AbstractThe Salton Sea Geothermal Field is one of the most geothermally and seismically active areas in California and presents an opportunity to study the effect of high‐temperature metamorphism on the properties of seismogenic faults. The area includes numerous active tectonic faults that have recently been imaged with active source seismic reflection and refraction. We utilize the active source surveys, along with the abundant microseismicity data from a dense borehole seismic network, to image the 3‐D variations in seismic velocity in the upper 5 km of the crust. There are strong velocity variations, up to ~30%, that correlate spatially with the distribution of shallow heat flow patterns. The combination of hydrothermal circulation and high‐temperature contact metamorphism has significantly altered the shallow sandstone sedimentary layers within the geothermal field to denser, more feldspathic, rock with higher P wave velocity, as is seen in the numerous exploration wells within the field. This alteration appears to have a first‐order effect on the frictional stability of shallow faults. In 2005, a large earthquake swarm and deformation event occurred. Analysis of interferometric synthetic aperture radar data and earthquake relocations indicates that the shallow aseismic fault creep that occurred in 2005 was localized on the Kalin fault system that lies just outside the region of high‐temperature metamorphism. In contrast, the earthquake swarm, which includes all of the M > 4 earthquakes to have occurred within the Salton Sea Geothermal Field in the last 15 years, ruptured the Main Central Fault (MCF) system that is localized in the heart of the geothermal anomaly. The background microseismicity induced by the geothermal operations is also concentrated in the high‐temperature regions in the vicinity of operational wells. However, while this microseismicity occurs over a few kilometer scale region, much of it is clustered in earthquake swarms that last from hours to a few days and are localized near the MCF system.

Low-frequency earthquakes in Shikoku, Japan, and their relationship to episodic tremor and slip

Non-volcanic seismic tremor was discovered in the Nankai trough subduction zone in southwest Japan and subsequently identified in the Cascadia subduction zone. In both locations, tremor is observed to coincide temporally with large, slow slip events on the plate interface downdip of the seismogenic zone. The relationship between tremor and aseismic slip remains uncertain, however, largely owing to difficulty in constraining the source depth of tremor. In southwest Japan, a high quality borehole seismic network allows identification of coherent S-wave (and sometimes P-wave) arrivals within the tremor, whose sources are classified as low-frequency earthquakes. As low-frequency earthquakes comprise at least a portion of tremor, understanding their mechanism is critical to understanding tremor as a whole. Here, we provide strong evidence that these earthquakes occur on the plate interface, coincident with the inferred zone of slow slip. The locations and characteristics of these events suggest that they are generated by shear slip during otherwise aseismic transients, rather than by fluid flow. High pore-fluid pressure in the immediate vicinity, as implied by our estimates of seismic P- and S-wave speeds, may act to promote this transient mode of failure. Low-frequency earthquakes could potentially contribute to seismic hazard forecasting by providing a new means to monitor slow slip at depth.

Preliminary Report on the 28 September 2004, M 6.0 Parkfield, California Earthquake

The Mw 6.0 Parkfield earthquake struck central California at 17:15:14 UTC on 28 September 2004. The epicenter was located 11 km southeast of the rural town of Parkfield, adjacent to Gold Hill and on the San Andreas Fault (Figure 1). The California Integrated Seismic Network (CISN) reported that the hypocenter was located at 35.819°N, 120.364°W at a depth of 8.8 km. From the distribution of aftershocks and from models of seismograms, strain changes, and geodetic displacements from the earthquake, it appears that the rupture propagated to the northwest along the San Andreas Fault from its hypocenter beneath Gold Hill to Middle Mountain. Because of the earthquake's moderate size and the low population density, only minimal damage was reported, but strong ground motions of approximately 1 g were recorded at a few isolated points (Shakal et al. , 2005, this issue). The 2004 Parkfield earthquake is the sixth of a series of historical events located on this reach of the San Andreas Fault; similar-sized earthquakes occurred in 1881, 1901, 1922, 1934, and 1966 (Bakun and McEvilly, 1984). Because of the frequency of moderate-sized earthquakes in the same locality, the Parkfield Prediction Experiment (Bakun and Lindh, 1985) was initiated in the mid-1980's, with cooperation between the U.S. Geological Survey (USGS), the California Geological Survey (CGS), and various universities to instrument the area densely to capture a future earthquake. In addition to enlarging the seismic network with more short-period, broadband, and strong-motion instruments, other instruments were installed that are unique to this experiment. These instruments include borehole strainmeters, Global Positioning System (GPS) receivers, a borehole seismic network, creepmeters, a telluric array, magnetometers, and fluid-pressure sensors (Roeloffs and Langbein, 1994). The goal of the experiment was to capture the complete earthquake cycle: the interseismic period, the actual earthquake, and the postseismic response. The …

Joint inversion of fault zone head waves and direct P arrivals for crustal structure near major faults

A newly recognized class of seismic phases, namely fault zone head waves arising from refraction at a transfault velocity contrast, provides additional constraints to direct P waves in inversions of near‐fault arrival time data for earthquake location and velocity structure. Incorporation of fault zone head waves in the inversion process increases the available data and broadens the spatial coverage of ray paths between sources and receivers. Equally important, the explicit recognition of fault zone head waves eliminates a source of error resulting from misidentifying head wave first arrivals as direct P waves. A simple algorithm for the joint inversion of fault zone head waves and direct P phases is illustrated with observed and synthetic data. Application to a small data set recorded by the borehole seismic network at Parkfield, California, provides separate seismic velocity depth profiles for the crustal blocks on the two sides of the San Andreas fault. For depths less than 3 km the obtained velocity contrast is 10–20%. For greater depths it decreases to 3–7%. Numerical tests with synthetic data show that for sources with known locations and for well located local earthquakes (randomized mislocation errors of up to 600 m) proper use of fault zone head waves can significantly improve the accuracy of the fitted velocity structure. The results of this work motivate the inclusion of fault zone head waves in state‐of‐the‐art joint structure‐hypocenter inversions using high quality near‐fault data.

San Andreas Fault Zone Head Waves Near Parkfield, California

Microearthquake seismograms from the borehole seismic network on the San Andreas fault near Parkfield, California, provide three lines of evidence that first P arrivals are "head" waves refracted along the cross-fault material contrast. First, the travel time difference between these arrivals and secondary phases identified as direct P waves scales linearly with the source-receiver distance. Second, these arrivals have the emergent wave character associated in theory and practice with refracted head waves instead of the sharp first breaks associated with direct P arrivals. Third, the first motion polarities of the emergent arrivals are reversed from those of the direct P waves as predicted by the theory of fault zone head waves for slip on the San Andreas fault. The presence of fault zone head waves in local seismic network data may help account for scatter in earthquake locations and source mechanisms. The fault zone head waves indicate that the velocity contrast across the San Andreas fault near Parkfield is approximately 4 percent. Further studies of these waves may provide a way of assessing changes in the physical state of the fault system.

Fault zone trapped seismic waves

AbstractTwo instances of fault zone trapped seismic waves have been observed. At an active normal fault in crystalline rock near Oroville in northern California, trapped waves were excited with a surface source and recorded at five near-fault borehole depths with an oriented three-component borehole seismic sonde. At Parkfield, California, a borehole seismometer at Middle Mountain recorded at least two instances of the fundamental and first higher mode seismic waves of the San Andreas fault zone. At Oroville recorded particle motions indicate the presence of both Love and Rayleigh normal modes. The Love-wave dispersion relation derived for an idealized wave guide with velocity structure determined by body-wave travel-time inversion yields seismograms of the fundamental mode that are consistent with the observed Love-wave amplitude and frequency. Applying a similar velocity model to the Parkfield observations, we obtain a good fit to the trapped wavefield amplitude, frequency, dispersion, and mode time separation for an asymmetric San Andreas fault zone structure—a high-velocity half-space to the southwest, a low-velocity fault zone, a transition zone containing the borehole seismometer, and an intermediate velocity half-space to the northeast. In the Parkfield borehole seismic data set, the locations (depth and offset normal to fault plane) of natural seismic events which do or do not excite trapped waves are roughly consistent with our model of the low velocity zone. We conclude that it is feasible to obtain near-surface borehole records of fault zone trapped waves. Because trapped modes are excited only by events close to the fault zone proper—thereby fixing these events in space relative to the fault—a wider investigation of possible trapped mode waveforms recorded by a borehole seismic network could lead to a much refined body-wave/tomographic velocity model of the fault and to a weighting of events as a function of offset from the fault plane. It is an open question whether near-surface sensors exist in a stable enough seismic environment to use trapped modes as an earth monitoring device.