Figure 1 - uploaded by Gareth O'Brien
Content may be subject to copyright.
Topography of Cotopaxi volcano, Ecuador, and locations of stations (gray circles and black solid triangles) used in our simulations. Black solid triangles represent the locations of five real stations at Cotopaxi.
Source publication
We performed numerical simulations of seismic waveforms with frequencies up to 10 Hz in heterogeneous media with topography to investigate the effects of topography and structural heterogeneity on seismic scattering. We used the simulated waveforms to test the source location method assuming isotropic radiation of S waves for long-period events and...
Contexts in source publication
Context 1
... used the topography of Cotopaxi volcano, Ecuador, and distributed stations on its surface at intervals of 750 m. We also included five stations that mimic the positions of real seismic stations on the vol- cano [Kumagai et al., 2010] (Figure 1). We assumed a source position beneath the summit at a depth of 4 km above sea level, similar to the source location estimated for a VLP/LP event observed at Cotopaxi volcano [Kumagai et al., 2010]. ...
Context 2
... Figure 10a plots the mean square (MS) velocity amplitudes bandpassed between 5 and 10 Hz at a lapse time of 2.5 s with a time window of 0.2 s as a function of the distance from the source to the individual stations for the heterogeneous model. In Figure 10a, the S wavefront is seen at the distance between 4500 and 5000 m and the amplitudes decays as the distance decreases. ...
Context 3
... Figure 10a plots the mean square (MS) velocity amplitudes bandpassed between 5 and 10 Hz at a lapse time of 2.5 s with a time window of 0.2 s as a function of the distance from the source to the individual stations for the heterogeneous model. In Figure 10a, the S wavefront is seen at the distance between 4500 and 5000 m and the amplitudes decays as the distance decreases. This feature is similar to that predicted from the single scattering model [e.g., Sato and Fehler, 1998]. ...
Context 4
... esti- mated the coda energy densities in the four frequency bands for the models with different a values as well as the topog- raphy model. The estimated coda energy densities were plotted as a function of ka in Figure 10b. The coda energy densities become smaller than that for the topography model as ka increases in each frequency band (Figure 10b). ...
Context 5
... estimated coda energy densities were plotted as a function of ka in Figure 10b. The coda energy densities become smaller than that for the topography model as ka increases in each frequency band (Figure 10b). This feature also suggests that scattering due to topography is suppressed by structural heterogeneity in the range ka ) 1. We note that the difference in the coda energy density levels for the individual frequency bands originated in the source excitation. ...
Context 6
... Next, we used a shear faulting mechanism in our model. As shown in Figure 11, the amplitude distribution for the heterogeneous model in the 5-10 Hz band is clearly distorted from the four-lobe pattern of the half-space model. The residuals for the shear faulting models with various correlation distances in the four frequency bands as a function of ka showed a trend similar to that in Figure 9, although results are not shown here. ...
Context 7
... Figure 12 shows the simulated vertical velocity waveforms at the five real station locations for the vertical crack source in the heterogeneous model. To test isotropic radiation of S waves, we used these waveforms to locate the source using the amplitude source location method. ...
Context 8
... used a 5-10 Hz band-pass filter and a 10 s time window to determine the source location as described previously. The normalized residual distribution and best fit location ( Figure 13) show that the source location was not correctly determined. Neither could we correctly locate the source using the seismograms of the topography model. ...
Context 9
... We then used the simulated seismograms at the five real station locations from an isotropic source in the het- erogeneous model (Figure 14). These simulated seismo- grams show clear P wave onsets followed by scattered waves. ...
Context 10
... we used a 10 s time window in the 5-10 Hz band, the source location was not properly determined. If we assume isotropic radiation of P waves instead of S waves in the source location method, the source is more accurately determined (Figure 15a). This result indicates that the simu- lated wavefield is dominated by P waves. ...
Context 11
... result indicates that the simu- lated wavefield is dominated by P waves. When we excluded the onset of P waves by using a time window starting from 0.5 s after the origin time, the source location determined by assuming isotropic radiation of S waves was closer to the correct position (Figure 15b), indicating that the scat- tered waves are dominated by S waves. ...
Context 12
... The distortion of RMS amplitude patterns of the simulated vertical waveforms increases as a decreases (Figure 7), f increases (Figure 8), and travel distance in- creases ( Figures 5 and 11). These features can be consis- tently explained by the ka-kL diagram of Aki and Richards [1980]. ...
Context 13
... distance from the source to station BREF (the closest of the five real station locations to the source) is roughly 2000 m, so kL is approximately 50. As seen in Figures 5c and 11b, the distortion of the RMS amplitude patterns is greater where distances from the source are larger than that of sta- tion BREF. On the other hand, close to the source, the amplitude distributions are more affected by the source mechanism. ...
Context 14
... Our simulated wavefield from an isotropic source ( Figure 14) was dominated by P waves, as indicated by the result of the source location determination (Figure 15a). However, if we used the simulated waveforms after the P wave onsets, the source location determined was close to its correct position (Figure 15b), which suggests that for the heterogeneous model the S waves were converted from P waves. ...
Context 15
... Our simulated wavefield from an isotropic source ( Figure 14) was dominated by P waves, as indicated by the result of the source location determination (Figure 15a). However, if we used the simulated waveforms after the P wave onsets, the source location determined was close to its correct position (Figure 15b), which suggests that for the heterogeneous model the S waves were converted from P waves. ...
Context 16
... Our simulated wavefield from an isotropic source ( Figure 14) was dominated by P waves, as indicated by the result of the source location determination (Figure 15a). However, if we used the simulated waveforms after the P wave onsets, the source location determined was close to its correct position (Figure 15b), which suggests that for the heterogeneous model the S waves were converted from P waves. P-S conversions were observed in volcanic regions by Matsumoto and Hasegawa [1991] using seismograms from a marine air gun source and by Yamamoto and Sato [2010] using dense seismic network data from an active seismic experiment. ...
Context 17
... previously mentioned, seismograms from shots recorded at volcanoes are characterized by small or missing P wave onsets with long coda waves [Wegler and Lühr, 2001;Wegler, 2003]. Our simulated waveforms from Figure 15a but with a time window starting from 0.5 s after the origin time to exclude the onset P waves and assuming isotropic radiation of S waves for the source location method. ...
Similar publications
Mathematical models and numerical solution procedures for predicting the trajectory, oscillation, possible
rotation, and mass and size time-evolution of cylindrical wind-driven firebrands are described and discussed. Two test
problems and the results, used for validating the mathematical models, are presented. In one, experimental measurements
of n...
Pétrequin was a French surgeon born in the town of Villeurbanne near the city of Lyon. He lived in an era when surgery was evolving towards a more complex and modern specialty. His magnificent career was shared between surgical pathology and history of surgery. His two most significant works were the "Traité d'anatomie médico-chirurgicale et topogr...
The joint probability distribution of streamwise particle hop distance, lateral particle hop distance, and travel time constrains the relationships between topographic change and sediment transport at the granular scale. Previous studies have investigated the ensemble characteristics of particle motions over plane bed topography; however, it is unc...
Plain Language Summary
Seismologists frequently use traveltimes of seismic waves to understand how the earth's internal structure changes at various depths. This study presents pictorial examples of how traveltimes of seismic waves sense the depth variations inside the earth. We focus on waves which reflect from specific depths and they are often...
The difficulties of measuring bedload transport in gravel bed rivers have given rise to the morphological method wherein sediment transport is inferred from changes in riverbed elevation and estimates of the distance traveled by sediment, its path length. Because current methods for estimating path length are time and labor intensive, we present a...
Citations
... Numerical simulations have concentrated on the effect of topography on seismic wavefields at local-intermediate distances or at teleseismic distances in 2-D, thereby significantly reducing the propagation complexity and possibly ignoring important 3-D effects. Topographic effects on local wavefields were also studied in the context of crustal earthquakes [e.g., 53], volcano seismology [e.g., 26,44], and ground motions for seismic hazard assessment [e.g., 3,10,21,22,24,29,33,34,43,60]. These studies demonstrate that topographic scattering is significant at short distances provided the length scale of the topography is comparable to the seismic wavelength, although the effects are more variable in the presence of steep slopes. ...
Over the last decade there has been an international effort to find methods to recover and digitize recordings from historical earthquakes and explosions that occurred during the 1950’s through to the 1980’s. Making these recordings accessible in digital format offers opportunities to study what signatures are encoded in the data, and to apply state-of-the-art techniques and methods to historical data. In this study we employ unsupervised machine learning to cluster historical teleseismic waveforms from nuclear explosions conducted at the former USSR Degelen test site, in Kazakhstan, recorded at seismic arrays in the UK (EKA), Canada (YKA), Australia (WRA) and India (GBA). In particular, we use two unsupervised algorithms to cluster waveforms using shape-based clustering: kernel k-means and k-Shape. The algorithms clearly split waveforms into distinct clusters that are spatially related, even when waveform differences are subtle, and we show with local and teleseismic numerical simulations that the clusters are related to the topography. The topography at the Degelen test site has characteristic wavelengths of 2-4 km and local simulations highlight that the seismic wavefield is trapped in reverberating mountain peaks. The location of the explosion is crucial in determining which section of the mountain range reverberates, influencing the outgoing wavefield. Teleseismic waveform simulations confirm that it is this superposition of energy leaving the reverberating peaks that results in the observed teleseismic waveform differences we observe.
... The CCF-based SSA uses the direct S-wave phase to obtain the travel time differences, while the ASL requires that the radiation pattern of S-wave is sufficiently masked at high frequencies. Kumagai et al. [2010] and Kumagai et al. [2011] suggest that isotropic radiation assumption is valid for the frequencies larger than 5 Hz at which seismic scattering process masks the radiation pattern. Nevertheless, we are able to extract the source information using the direct S-wave phase of VTs at Izu-Oshima at such high frequencies. ...
[Japanese Title] 地震波の相互相関解析に基づく火山性微動と地震の検知と震源決定
Detection and source location of volcanic tremor and earthquakes enable us to better understand the physical processes inside a volcano. Volcanic tremor and earthquakes are seismic events that are associated with the movements and pressurization of magma and volcanic gas within the volcanic edifice. Volcano-tectonic earthquakes (VTs), which are generated by fault motions, are well detected and located easily because of their clear onsets of P and S waves. The other events such as volcanic tremors and low-frequency events are often not detected or accurately located because of unclear arrivals of P and S waves and great variations in the amplitudes and durations of the waveforms. Previous studies have proposed alternative methods such as amplitude source location method using the spatial distribution of seismic amplitudes, but they generally require a considerable amount of initial information. Recently, the cross-correlation analyses have been used for locating and detecting volcanic tremors. The analyses retrieve, without initial information, the similarity of seismic signals at station pairs which contain information about the coherency and relative travel time that can be used for detection and location. However, previous studies applied those seismic correlation-based methods in a sparse seismic network over a large area or using a network with a high number of seismic stations. Also, the source locations were often determined in 2-D, so that the depth of seismic events is not located, which is one of the most important information to capture magma migration from the depths.
In the present study, therefore, we improve the method for monitoring seismic signals at active volcanoes where several seismic stations are deployed around the active craters and on the flanks. Also, we extend the method to locate the source in 3-D by taking into account a three-dimensional velocity structure. Previous studies applied the correlation-based method to volcanic tremors observed at active volcanoes, but no evaluation was done for the accuracy of the source location because of unknown real locations. To overcome such ambiguity, the present study estimates the errors of source location by analyzing VTs that are well located. We then apply the improved location method to real volcanic tremors and track the changes in the source locations with time. However, it is difficult to evaluate the reliability of the source location and quantitatively distinguish the source locations of tremors from noise. To address this problem, we perform detection by using methods that measure signal coherency evaluated from seismic cross-correlation analyses. While the detections methods are able to pick up coherent signals coming from volcanic tremors and earthquakes, the present study also uses the detection results to evaluate the reliability of the source location. To discriminate the detection and source location that might correspond to different types of seismic events, we extend the detection method by developing a classification method based on the characteristics of the detection for each type of event.
In Chapter 2, we develop a method called the cross-correlation function-based source scanning algorithm (CCF-based SSA) to obtain the 3-D location of volcanic tremors and earthquakes in a small seismic network. The method computes stacked CCFs from a set of time windows to enhance the coherent part of the seismic signals. First, we evaluate the method using VTs observed at Izu-Oshima volcano, Japan. We perform analysis at 4–16 Hz, where the dominant frequencies of direct S-wave are present. The estimated locations are well-matched with the hypocenters determined from the onsets of P and S waves. The misfits between them are 2 km or less for the VTs occurring inside the network coverage. Secondly, we simulate tremors by superimposing plural VTs to test the applicability of CCF-based SSA method to locate volcanic tremor. The sources of simulated tremor are located by calculating CCFs using 10 s time windows with a 90% overlap. The location errors are estimated to be 1 km or less for the data length of 2 min. Higher location accuracy is achieved when pseudo-sources (i.e., VTs) are set in a smaller source region and/or when we use longer data. The advantage of the CCF-based SSA is that the obtained location depends only on the seismic velocity model. Evaluation using VTs and simulated tremors suggest that the CCF-based SSA can be applied to real volcanic tremors and used to monitor changes in the source locations. The location errors of the CCF-based SSA are slightly larger than those from the amplitude source location method, but location offsets, which is computed as the average of the misfits, are found to be smaller.
In Chapter 3, we apply the CCF-based SSA to real volcanic tremors observed at Sakurajima volcano, Japan, by a network consisting of six seismic stations. The analysis is performed at 1–4 Hz, which is the frequency band where volcanic tremors are dominant. We track the source locations every 10 min by computing the CCFs using 20 s subwindows with 75% overlap and stacked them over a 10 min window. We analyze volcanic tremor episodes occurring on May 6–8, May 11–12, June 2–5, and August 22–25, 2017. Volcanic tremor sources are located under the active craters from the ground surface to 6 km depth, although the source locations are widely distributed under the active craters when the signal-to-noise ratio of the observed waveforms is low. The vertical extension of the source from 0 to 6 km is confirmed by the fact that the peaks of CCFs approach zero lag times as the estimated source becomes deeper. Because of the seismic station coverage, the maximum depth of source location that we can estimate is less than 7 km depth.
In Chapter 4, we perform the detection of volcanic tremors and B-type earthquakes, a type of volcanic earthquakes that is dominant at 1–5 Hz, at Sakurajima volcano using seismic correlation-based methods. The single-station correlation coefficient method and the network covariance matrix analysis are applied to six months of seismic records. The computed correlation coefficient and spectral width at 1–4 Hz, which are the measures of signal coherency, are used to detect volcanic tremors and B-type earthquakes by using an averaging window length of 10 minutes. Consistent with the seismic activity reported by Japan Meteorological Agency (JMA), we detect volcanic tremors that mainly occurred during May–June and August–September 2017. Volcanic earthquakes and short-duration tremors are better detected with the network covariance matrix analysis. For both detection methods, using shorter time windows are better to detect short-duration events, but longer windows improve the detection stability. We extract the characteristics of the detected volcanic tremors and B-type earthquakes during April–July 2017 based on JMA catalog, and develop a method using spectral width with weight functions for classifying volcanic tremors and B-type earthquakes during April–September 2017. The weight functions are prepared for each type of seismic event based on the signal coherency. As a result, the method is able to systematically classify volcanic tremors and B-type earthquakes.
In Chapter 5, we discuss volcanic tremors and B-type earthquakes at Sakurajima volcano. Correlation coefficient and spectral width are used to evaluate reliable source locations of volcanic tremors, because spatially scattered source locations are obtained from the tremors with low coherency or short-duration signals. Reliable source locations that are showing high coherency are distributed from a depth of about 4 km to the ground surface beneath the active craters. The horizontal distribution of the sources in SW-NE direction around the active craters is consistent with the tensile crack inferred from analyses of ground deformation in a previous study. We compare the results with data of tilt and infrasound, JMA catalog, and previous studies to understand the volcanic and eruptive processes at Sakurajima volcano, which are related to magma and gas movement in the volcanic conduit and/or ejection of volcanic materials during eruptive periods. Observed infrasound signals and stable infrasound spectra during volcanic tremor indicate that tremor activities are associated with eruptions, and the fluctuation of volcanic flow at the vent is the same regardless of the amplitudes. Seismic spectra show some changes with time which seems to be correlated with the estimated source depth. For example, the volcanic tremor located at 3–6 km depth before an explosive eruption shows wider spectral peaks around 1.25 Hz. Different seismic spectra may correspond with different generation depth or source mechanism of volcanic tremors.
We also show that not only volcano-seismic events, but also meteorological effects such as precipitation and typhoons are recognized in the temporal changes of the results obtained by cross-correlation based method.
In this thesis, we introduced a method for the detection and location of volcanic tremors and earthquakes. In the application of volcano monitoring, the technique can be used to detect and monitor the changes in source locations in real-time, with no additional information other than a seismic velocity model. A comparison of the results with various data such as infrasound and ground deformation may provide more insights into the source process of volcanic tremors and/or eruption mechanisms. Therefore, our method may contribute to the field of volcano monitoring and the understanding of the volcanic processes.
... For this purpose, we select a subset of 127 known volcanic earthquakes that show clear signals with an SNR >10 at all 12 stations, with respect to a 5 s noise signal before the origin time. The seismic spectra of volcanic tremors and earthquakes (Fig. 1d) show broad frequency contents at >1 Hz. Kumagai et al. (2011) and Morioka et al. (2017) show that seismic scattering at volcanoes sufficiently masks the seismic wave radiation pattern at >5 Hz and fulfills the isotropic radiation assumption. Takemura et al. (2009) examined the S-wave radiation pattern from tectonic earthquakes and found significant distortions at frequencies >2 Hz, eventually resembling isotropic radiation at >5 Hz. ...
Volcanic tremors and earthquakes must be monitored to gain insights into volcanic activity. Localization of their sources is often challenging because of the unclear onset of seismic waves, particularly when the volcanic activity increases before and during an eruption. Existing alternative techniques to locate the seismic sources are based on the information on the spatial amplitude distribution or the travel-time difference of seismic waves. Exploring the idea of combining both information for source location determination, we propose a new location method that uses the amplitude and travel-time difference information obtained from the unnormalized cross correlations of seismic data. Evaluation using volcanic earthquakes that occurred in 2020 at Tokachidake volcano, Japan, reveals an improvement in location accuracy compared to existing methods using individual information. Analysis of an episode of volcanic tremors and earthquakes accompanying a rapid tilt change event on 14 September 2020 reveals that during the inflation of the crater area, reliable seismic source locations with an error of ≤1 km become more concentrated at around 0.6 km beneath the 62-2 crater, in which the most recent eruptive activity had occurred. Such changes in source locations are associated with the movement of volcanic gas and hot water from the hydrothermal system below. Our proposed method is useful for locating and monitoring seismic source locations corresponding to volcanic fluid movements.
... According to Sato et al. (2012), the pattern and strength of scattering can be described by the power spectral density (PSD) of topographies. The topographic PSD of the region shows a power-law decay for large wavenumbers ( Figure S3 in the Supporting Information S1), which can be explained by a von Karmantype PSD function (Kumagai et al., 2011;Takemura et al., 2015), suggesting that the surface topography here is rich in short-wavelength topographic constituents (Sato et al., 2012). ...
With dense nodal seismic arrays growing rapidly, receiver function (RF) imaging has revealed unprecedented small‐scale features. However, whether these fine signals represent real intra‐crustal structures needs to be verified. In this study, we take the RF profiles recorded by a dense array in Ruoergai basin in northeastern Tibet as an example and demonstrate that continuous dipping phases in receiver functions (RFs) could be generated from the interference between the incoming P‐wave and dramatic topographic variations. Full‐waveform modeling illustrates that when an incident P‐wave encounters a topographic peak, it can induce scattered P‐ and Rayleigh waves that propagate in backward and forward directions, consistent with the observations. The scattered Rayleigh waves exhibit a cosine radiation pattern in R‐components, which explains the azimuthal‐dependent observation of dipping phases in RF profiles. We then propose a two‐step separation method to suppress the continuous topographic scattered phases in RF imaging, which involves a time‐domain adaptive matched filtering and a curvelet‐based separation. The resulting common conversion point (CCP) images show that the method can effectively suppress the continuous scattered phases induced by topography. CCP images exhibit flat intra‐crustal layers, suggesting a decoupled deformation style in the upper and lower crust in NE Tibet. The upper crustal layer may represent the basement of the décollement layer, and the intermittent lower crustal interfaces are likely to indicate the partial melt within the crustal channel flow. The separation method is applicable for dense nodal array RF imaging to reveal reliable, detailed intra‐crustal structures in tectonically active regions that developed dramatic topography.
... The assumption for isotropic S-wave radiation is valid for signals with frequencies higher than 5 Hz (Takemura et al. 2009;Kumagai et al. 2011). Hence, we focus on ground velocity in the band of 5-20 Hz, which is almost identical to one of the dominant frequency ranges in observed seismic records (Fig. 4). ...
Kusatsu-Shirane volcano hosts numerous thermal springs, fumaroles, and the crater lake of Yugama. Hence, it has been a particular study field for hydrothermal systems and phreatic eruptions. On 23 January 2018, a phreatic eruption occurred at the Motoshirane cone of Kusatsu-Shirane, where no considerable volcanic activity had been reported in observational and historical records. To understand the eruption process of this unique event, we analyzed seismic, tilt, and infrasound records. The onset of surface activity accompanied by infrasound signal was preceded by volcanic tremor and inflation of the volcano for ~ 2 min. Tremor signals with a frequency band of 5–20 Hz remarkably coincide with the rapid inflation. We apply an amplitude source location method to seismic signals in the 5–20 Hz band to estimate tremor source locations. Our analysis locates tremor sources at 1 km north of Motoshirane and at a depth of 0.5–1 km from the surface. Inferred source locations correspond to a conductive layer of impermeable cap-rock estimated by magnetotelluric investigations. An upper portion of the seismogenic region suggests hydrothermal activity hosted beneath the cap-rock. Examined seismic signals in the 5–20 Hz band are typically excited by volcano-tectonic events with faulting mechanism. Based on the above characteristics and background, we interpret that excitation of examined volcanic tremor reflects small shear fractures induced by sudden hydrothermal fluid injection to the cap-rock layer. The horizontal distance of 1 km between inferred tremor sources and Motoshirane implies lateral migration of the hydrothermal fluid, although direct evidence is not available. Kusatsu-Shirane has exhibited unrest at the Yugama lake since 2014. However, the inferred tremor source locations do not overlap active seismicity beneath Yugama. Therefore, our result suggests that the 2018 eruption was triggered by hydrothermal fluid injection through a different pathway from that has driven unrest activities at Yugama.
... In the best case, a few instruments are deployed in borehole, thus sensing the strain at only a few points (Currenti and Bonaccorso, 2019). Numerical investigations have clearly shown that seismic velocities, gradient displacements and strain dramatically changes at sharp boundaries and in the presence of steep topography (Kumagai et al., 2011;Jousset et al., 2004;Cao and Mavroeidis, 2019). This poses challenges in accurately interpreting strain observations when relying on only a few measurement points. ...
We demonstrate the capability of distributed acoustic sensing (DAS) to record volcano-related dynamic strain at Etna (Italy). In summer 2019, we gathered DAS measurements from a 1.5 km long fibre in a shallow trench and seismic records from a conventional dense array comprised of 26 broadband sensors that was deployed in Piano delle Concazze close to the summit area. Etna activity during the acquisition period gives the extraordinary opportunity to record dynamic strain changes (∼ 10−8 strain) in correspondence with volcanic events. To validate the DAS strain measurements, we explore array-derived methods to estimate strain changes from the seismic signals and to compare with strain DAS signals. A general good agreement is found between array-derived strain and DAS measurements along the fibre optic cable. Short wavelength discrepancies correspond with fault zones, showing the potential of DAS for mapping local perturbations of the strain field and thus site effect due to small-scale heterogeneities in volcanic settings.
... The pattern of energy ratios from the horizontal signal components i = N, E is comparable to the vertical pattern, indicating that the wave propagation along the topography dominates over source-characteristic radiation patterns. This topographic path effect (e.g., Kuehnert et al., 2020;Kumagai et al., 2011) is similar to the distortion of radiation patterns by the scattering of the wave field by smallscale soil heterogeneities, validating the assumption of an isotropic radiation at high frequencies above around 5 Hz (e.g., Kumagai et al., 2010;Takemura et al., 2009 ...
Rockfalls generate seismic signals that can be used to detect and monitor rockfall activity. Event locations can be estimated on the basis of arrival times, amplitudes, or polarization of these seismic signals. However, surface topography variations can significantly influence seismic wave propagation and hence compromise results. Here, we specifically use the signature of topography on the seismic signal to better constrain the source location. Seismic impulse responses are predicted using Spectral Element based simulation of three‐dimensional wave propagation in realistic geological media. Subsequently, rockfalls are located by minimizing the misfit between simulated and observed inter‐station energy ratios. The method is tested on rockfalls at Dolomieu crater, Piton de la Fournaise volcano, Reunion Island. Both single boulder impacts and distributed granular flows are successfully located, tracking the complete rockfall trajectories by analyzing the signals in sliding time windows. Results from the highest frequency band (here 13–17 Hz) yield the best spatial resolution, making it possible to distinguish detachment positions less than 100 m apart. By taking into account surface topography, both vertical and horizontal signal components can be used. Limitations and the noise robustness of the location method are assessed using synthetic signals. Precise representation of the topography controls the location resolution, which is not significantly affected by the assumed impact direction. Tests on the network geometry reveal best resolution when the seismometers triangulate the source. We conclude that this method can improve the monitoring of rockfall activity in real time once a simulated database for the region of interest is created.
... Besides the low-velocity and scattering layer, the local structure, for example, surface and/or Moho topographies, could also contribute to the development of seismic scattering (e.g., Blanchette-Guertin et al., 2012;Onodera et al., 2018). For the Earth seismology, the effect of local topographies on the seismic scattering is paramount (e.g., Kumagai et al., 2011;Sato et al., 2012;Takemura et al., 2015). Since the dynamic range of topographies on extraterrestrial planetary bodies, for example, Moon and Mars, is larger than that of the Earth, topographical effects can be a key factor in interpreting seismic scattering in the field of planetary seismology. ...
One of the most critical issues associated with the analysis of lunar seismic data is the intense scattering, which prevents precise seismic phase identifications, thereby resulting in poor constraints on the internal structure of the Moon. Although some studies estimated subsurface scattering properties from analyses of the Apollo seismic data, the properties have large uncertainties and are still open issues to be resolved to improve the inner structure model of the Moon. While the previous studies tried to constrain the scattering features within the lunar crust mainly from data analysis, this study estimated them from a numerical approach. We constrained the scattering properties near Apollo 12 landing site by conducting seismic wave propagation simulations under various parameter settings and comparing the synthetics with the data. As a result, we succeeded in reproducing seismic signals excited by the Apollo artificial impacts. This led to a constraint on the scattering properties, such as typical scale and thickness of heterogeneity, around the Apollo 12 landing site. The derived structure suggests that the intense scattering structure exists down to 20 km at the northern portion of the region of the Apollo 12 landing site, and to 10 km at the southern region from the landing site. In addition, our model requires a smaller P‐ and S‐wave velocity ratio (1.2–1.4) compared with those conventionally considered (>1.73). This implies a dry and porous environment consistent with laboratory measurements of terrestrial samples and reasonable with the generalized lunar environment.
... In the best case, few instruments are deployed in borehole, thus sensing the strain in only few points (Currenti and Bonaccorso, 2019). Numerical investigations have clearly shown that seismic velocities, gradient displacements and strain dramatically changes at sharp boundaries and in presence of steep topography (Kumagai et al., 2011;Jousset et al., 2004;Cao and Mavroeidis, 2019). This poses challenges in interpreting accurately strain observations relying 25 on only few measurements points. ...
We demonstrate the capability of Distributed Acoustic Sensing (DAS) in recording volcano related dynamic strain at Etna (Italy). In summer 2019, we gathered DAS measurements from a 1.5 km long fibre in a shallow trench and seismic records from a conventional dense array comprising 26 broadband sensors deployed in Piano delle Concazze close to the summit area. The multifaceted style of Etna activity during the acquisition period gives the extraordinary opportunity to record and detect tiny strain changes (few 10−8 strain) in correspondence with volcanic events. To validate the DAS strain measures, we explored array-derived methods to estimate strain changes from the seismic signals and to compare with strain DAS signals. A general good agreement was found between array-derived strain and DAS measures along the fibre optic cable. Short wavelength discrepancies correspond with fault zones, showing the potential of DAS in mapping local perturbations of the strain field, and thus site effect, due to small-scale heterogeneities in volcanic settings.
... The subscript i corresponds at the i-th station in the network. The assumption for isotropic S-wave radiation is effective for signals with frequencies higher than 5 Hz (Takemura et al., 2009;Kumagai et al., 2011). Hence, we focus on ground velocity in the band of 5-20 ...
Kusatsu-Shirane volcano has been a particular study field for hydrothermal system and phreatic eruptions with plenty of thermal springs, fumaroles, and a crater lake of Yugama. On 23 January 2018, a phreatic eruption occurred at the Motoshirane cone of Kusatsu-Shirane, where no considerable volcanic activity had been reported in observational and historical records. To understand the eruption process of such a unique event, we examined observed seismic, tilt, and infrasound records. The onset of surface activity accompanied by infrasound signal was preceded by volcanic tremor and inflation of the volcano for 2 minutes. Tremor signals with a frequency of 5–20 Hz remarkably coincide with the rapid inflation. We apply an amplitude source location method to seismic signals in the 5–20 Hz band to estimate tremor source locations. Our analysis locates tremor sources at 1 km north of Motoshirane and at a depth of 0.5–1 km from the surface. Inferred source locations correspond a conductive layer of impermeable cap-rock estimated by magnetotelluric investigations, and an upper portion of the seismogenic region, suggesting hydrothermal activity hosted beneath the cap-rock. Examined seismic signals in the 5–20 Hz band are typically excited by volcano-tectonic events with faulting mechanism. Based on the above characteristics and background, we interpret that excitation of examined volcanic tremor reflects small shear fractures induced by sudden hydrothermal fluid injection to the cap-rock layer. The horizontal distance of 1 km between inferred tremor sources and Motoshirane implies lateral migration of the hydrothermal fluid, although we have not obtained direct evidence. Kusatsu-Shirane has a series of unrest at the Yugama lake since 2014. However, inferred tremor source locations do not overwrap active seismicity beneath Yugama. Therefore, our result suggests that the 2018 eruption was triggered by hydrothermal fluid injection through an independent pathway that has driven unrest activities at Yugama.
