Yuta Maeda’s research while affiliated with Nagoya University and other places

The 2014 phreatic eruption of Mt. Ontake was preceded by increased volcano-tectonic (VT) seismicity, but all other types of precursors were obscure. A subsurface process leading to eruption with few precursors has yet to be identified. This study performed numerical simulations to reproduce the precursory time sequence of the eruption. High-temperature water from a cooling magma reservoir was injected into a modeled porous medium filled with cold water. This resulted in an immediate pressure increase and a delayed temperature increase in the shallow parts. The pressure increased immediately because the injected fluid pushed the existing fluid away from the porous medium. A sharp boundary between low (initial)- and high (injected)-temperature regions, known as a thermal front, developed and migrated upward slowly, leading to a delayed temperature increase. The lag time between the pressure and temperature changes is likely responsible for the precursory VT seismicity. If the injected water temperature was less than the critical point, the water was liquid throughout the model region until the thermal front reached a shallow part, where the pressure was low enough for the heated water to vaporize. This vaporization was sudden and large, likely leading to a large eruption with few precursors, similar to the time sequence in 2014. If the injected water temperature was higher, the water volume increased gradually from liquid to supercritical and then to the gas state. This scenario likely led to a small eruption with numerous precursors, similar to the previous eruption of Mt. Ontake that occurred in 2007. Graphical Abstract

The 2014 phreatic eruption of Mt. Ontake was preceded by increased volcano-tectonic (VT) seismicity, but all other types of precursors were obscure. A subsurface process leading to eruption with few precursors has yet to be identified. This study performed numerical simulations to reproduce the precursory time sequence of the eruption. High-temperature water from a cooling magma reservoir was injected into a modeled porous medium filled with cold water. This resulted in an immediate pressure increase and a delayed temperature increase in the shallow parts. The pressure increased immediately because the injected fluid pushed the existing fluid away from the porous medium. A sharp boundary between low (initial)- and high (injected)-temperature regions, known as a thermal front, developed and migrated upward slowly, leading to a delayed temperature increase. The lag time between the pressure and temperature changes is likely responsible for the precursory VT seismicity. If the injected water temperature was less than the critical point, the water was liquid throughout the model region until the thermal front reached a shallow part, where the pressure was low enough for the heated water to vaporize. This vaporization was sudden and large, likely leading to a large eruption with few precursors, similar to the time sequence in 2014. If the injected water temperature was higher, the water volume increased gradually from liquid to supercritical and then to the gas state. This scenario likely led to a small eruption with numerous precursors, similar to the previous eruption of Mt. Ontake that occurred in 2007.

The Aira caldera, located in southern Kyushu, Japan, originally formed 100 ka, and its current shape reflects the more recent 30 ka caldera-forming eruptions (hereafter, called the AT eruptions). This study aimed to delineate the detailed two-dimensional (2D) seismic velocity structure of the Aira caldera down to approximately 15 km, by means of the travel-time tomography analysis of the seismic profile across the caldera acquired in 2017 and 2018. A substantial structural difference in thickness in the subsurface low-velocity areas in the Aira caldera between the eastern and western sides, suggest that the Aira caldera comprises at least two calderas, identified as the AT and Wakamiko calderas. The most interesting feature of the caldera structure is the existence of a substantial high-velocity zone (HVZ) with a velocity of more than 6.8 km/s at depths of about 6–11 km beneath the central area of the AT caldera. Because no high ratio of P- to S-wave velocity zones in the depth range were detected from the previous three-dimensional velocity model beneath the AT caldera region, we infer that the HVZ is not an active magma reservoir but comprises a solidified and cool remnant. In addition, a poorly resolved low-velocity zone around 15 km in depth suggests the existence of a deep active magma reservoir. By superimposing the distribution of the known pressure sources derived from the observed ground inflation and the volcanic earthquake distribution onto the 2D velocity model, the magma transportation path in the crust was imaged. This image suggested that the HVZ plays an important role in magma transportation in the upper crust. Moreover, we estimated that the AT magma reservoir in the 30 ka Aira caldera-forming eruptions has the total volume of 490 km ³ DRE and is distributed in a depth range of 4–11 km. Graphical Abstract

Seismic exploration was conducted along a profile running through the Aira caldera located in southern Kyushu, Japan. The caldera was formed by an ignimbrite eruption approximately 30 ka BP, namely, the “AT eruption,” which produced the Ito ignimbrite and widespread Aira-Tanzawa ash. This analysis aimed to clarify the detailed P -wave velocity structure beneath the caldera. Accordingly, 829 inland seismic stations and 42 ocean bottom seismographs were deployed along the 195 km-long seismic profile to record seismic waves generated by numerous controlled seismic sources. A detailed velocity structure of the active Aira caldera was successfully obtained to depths of 20 km through travel-time tomography. A substantial structural difference was observed in the thicknesses of the low-velocity zones between the eastern and western sides in the shallowest region of the Aira caldera, suggesting that the Aira caldera is composed of at least two calderas: the AT caldera associated with the AT eruption, and the Wakamiko caldera associated with the post-AT eruption. Perhaps the most interesting feature of the caldera structure is the existence of a substantially high-velocity zone at depths of 6–11 km beneath the center area of the AT caldera, which can be interpreted as the cooled and solidified magma reservoir formed during or after the AT eruption. In addition, a low-velocity region with approximately 15 km depths indicated a deep magma reservoir. Based on these novel and past research results, a new magma supply model in the Aira caldera was proposed. Further, the spatial distribution of the magma reservoir associated with the AT eruption 30 ka BP was estimated, while the future possibility of larger eruptions in this caldera was discussed.

Autocorrelation functions (ACFs) of vertically incident seismic waves are used to image subsurface reflectors. However, the reflection responses derived from ACFs usually contain many false signals. We present a method to quantify the errors in ACFs and extract true reflectors with high reliability. We estimated the errors for each earthquake at each station as follows. We calculated the amplitude of the observed waveform within the noise window and generated 1000 random noise traces that have this amplitude. By subtracting the random noise traces from the observed waveform, we created 1000 candidate earthquake waveforms. We computed the ACF for each of the 1000 waveforms and calculated the ensemble average and standard deviation of the 1000 different ACF amplitudes at each lag time. Then, we applied weighted stacking to the ACFs of many earthquakes to obtain the reflection response at the station. We calculated the standard deviation of the weighted stack to estimate errors in the reflection response. We evaluated the method by applying it to seismic data from the metropolitan area of Japan. The subsurface structure of the study area has been studied extensively and consists of a strong velocity discontinuity between sedimentary and basement layers. Following our method, the discontinuity was imaged as a clear reflector with an amplitude that was substantially greater than three times the standard deviation, which corresponds to statistical significance at the 99% confidence level. At other depths where reflectors are not expected to be present, the amplitudes of the peaks were less than or close to three times the standard deviation. The signal of the discontinuity was clearly visible at frequencies below 10 Hz and was less prominent at higher frequencies. Graphical Abstract

We conducted waveform inversions of an ultra-long-period (~ 240-s) event associated with the phreatic eruption of Mount Kusatsu–Shirane on January 23, 2018. We used broadband seismic and tilt records from three stations surrounding the eruption site. The horizontal components of the broadband seismic records were severely contaminated by tilt motions. We applied a waveform inversion algorithm to account for both the translational and tilt motions. To reduce the number of free parameters, we assumed a tensile crack source and conducted grid searches for the centroid location and orientation of the crack. The results showed a rapid inflation of 10 ⁵ m ³ of the crack, followed by a slow deflation starting 8–11 s prior to the onset of the eruption. The source location and crack orientation were not uniquely determined. The most likely source is a north–south-opening sub-vertical crack near the eruptive craters. This ultra-long-period event may represent volcanic fluid migration from depth to the surface through a vertical crack during the eruption. Graphical abstract

We conducted waveform inversions of an ultra-long-period (~240-s) event associated with the phreatic eruption of Mount Kusatsu-Shirane on January 23, 2018. We used broadband seismic and tilt records from three stations surrounding the eruption site. The horizontal components of the broadband seismic records were severely contaminated by tilt motions. We applied a waveform inversion algorithm to account for both the translational and tilt motions. To reduce the number of free parameters, we assumed a tensile crack source and conducted grid searches for the centroid location and orientation of the crack. The results showed a rapid inflation of 10 ⁵ m ³ of the crack, followed by a slow deflation starting 8–11 s prior to the onset of the eruption. The source location and crack orientation were not uniquely determined. The most likely source is a north–south-opening sub-vertical crack near the eruptive craters. This ultra-long-period event may represent volcanic fluid migration from depth to the surface through a vertical crack during the eruption.

We propose a new algorithm, focusing on spatial amplitude patterns, to automatically detect volcano seismic events from continuous waveforms. Candidate seismic events are detected based on signal-to-noise ratios. The algorithm then utilizes supervised machine learning to classify the existing candidate events into true and false categories. The input learning data are the ratios of the number of time samples with amplitudes greater than the background noise level at 1 s intervals (large amplitude ratios) given at every station site, and a manual classification table in which ‘true'' or ‘false'' flags are assigned to candidate events. A two-step approach is implemented in our procedure. First, using the large amplitude ratios at all stations, a neural network model representing a continuous spatial distribution of large amplitude probabilities is investigated at 1 s intervals. Second, several features are extracted from these spatial distributions, and a relation between the features and classification to true and false events is learned by a support vector machine. This two-step approach is essential to account for temporal loss of data, or station installation, movement, or removal. We evaluated the algorithm using data from Mt. Ontake, Japan, during the first ten days of a dense observation trial in the summit region (November 1–10, 2017). Results showed a classification accuracy of more than 97 per cent.

Seismic scattering and attenuation at volcanoes, thought to be strongest on the Earth, can be used to map volcanic feeding systems. We systematically analyzed high‐frequency (5–10 Hz) seismograms of volcano‐tectonic earthquakes at Galeras volcano (Colombia) and active sources at Kirishima, Unzen, Bandai, and Iwate volcanoes (Japan) to investigate their scattering and attenuation characteristics. The envelope widths estimated from these seismograms were compared with those calculated by Monte Carlo envelope simulations for 1D layered models parameterized by the scattering mean free path and the quality factor of medium attenuation for S waves. Our results indicated a surficial, highly heterogeneous, and attenuative layer up to around 1 km thickness at all studied volcanoes. The strongest heterogeneities at volcanoes thus exist in a thin surface layer, likely comprising unconsolidated and/or highly fractured materials. Using the space‐weighting function for diffusive wavefields, we mapped the residuals between observed envelope widths and those calculated with our estimated 1D models at Kirishima, Unzen, Bandai, and Iwate. These maps showed spatial distributions of the envelope‐width residuals were unique to each volcano and correlated with P wave velocity tomographic images. Areas of positive residuals correspond to low‐velocity anomalies and thus to heterogeneous, strongly scattering rocks, whereas areas of negative residuals correspond to high‐velocity anomalies and thus less heterogeneous volcanic or basement rocks. Our results demonstrate that envelope widths can improve the characterization of scattering and attenuation structures beneath volcanoes.

Long-period (LP) seismic events at active volcanoes are thought to be generated by oscillations of fluid-filled resonators. The resonator geometry and fluid properties of LP sources have been estimated by comparing observed frequencies and quality factors (Q) with those calculated by numerical simulations with a crack model. A method to estimate all the parameters of crack geometry and fluid properties using an analytical formula for crack resonance frequencies has recently been proposed, but this method requires long computational times to compare observed and simulated Q values, especially for LP events with large Q. To resolve this problem, we used numerical simulations to systematically investigate the empirical relation between Q and crack model parameters. We found that Q can be calculated with an empirical formula expressed by the crack width-to-length ratio and the ratio of P-wave velocity in the solid medium to sound speed in the fluid. We applied this formula to LP events at Kusatsu-Shirane volcano, Japan, between August 1992 and January 1993 and at Galeras volcano, Colombia, in January 1993. Assuming misty gas as the fluid in the crack at Kusatsu-Shirane and dusty gas as the fluid at Galeras, the empirical formula provided more detailed estimates of the parameters than those obtained previously using the Q values estimated in numerical simulations. We then applied the empirical formula to LP events with large Q values observed at Galeras between December 2006 and January 2007. When we assumed dusty gas as the fluid in the crack, we found decreasing trends in both crack volume and the gas-weight fraction of water vapour in the crack. We also found that the dust volume was proportionally related to the product of crack aperture and crack length or width. These trends and relations were similar to those in January 1993, suggesting that the LP events at Galeras between 2006 and 2007 were triggered by the explosive fragmentation of intruded magma and the production of a dusty gas, as was previously inferred for the LP events in January 1993. Welding of ash in the dusty gas and dense magma remaining in the conduit after fragmentation led to a decrease in the source crack size prior to the next LP event. These results demonstrate that our empirical formula for Q can be used to estimate the source properties of LP events with large Q values without requiring long computational times. Use of the formula may thus contribute to improved monitoring of fluid states and understanding of LP triggering processes beneath many volcanoes.