The 144th (2021 SPRING) SEGJ CONFERENCE

June 9, Room No.1

Session	[Groundwater][Civil Engineering 1]
Chair	Yosuke Teranishi(JGI)

June 9	9:30 - 9:50　　　 Recently, new inversion methods based on Deep Learning (DL) have been proposed for the inversion of electromagnetic (EM) exploration data. DL is a method of machine learning that has been attracting attention in various fields, for example, in the field of natural language processing and image recognition. In this study, we set the advantages of EM inversion using deep learning as (1) inversion can be performed without an initial model, (2) high-speed calculation of a large amount of observed data. We have developed and applied the inversion method using deep neural network (DNN) to the frequency domain airborne electromagnetic (AEM) system. After confirming the validity of the developed DL method with synthetic data, we have analyzed the field data and compared the results with the existing inversion method (Gauss-Newton method). The results of the DL method show good agreement with those of the existing inversion method. Calculation time of the DL method was found to be more than 10 times faster than the existing method. From the above, we can conclude that there is a good possibility that the inverse analysis using deep learning can be applied to AEM data as a fast and stable inversion method.	1) 1D deep learning inversion of frequency domain airborne electromagnetic method
June 9		Yuya Matsumoto(Graduate School of Waseda Univ.),*Takumi Ueda(Waseda Univ.)
June 9	9:50 - 10:10　　　 This study aims to explore underground wet zones using an airborne electromagnetic drone (Drone-AEM). In this method, the time of wave used for analysis is shorter than 1 ms. In this study, we aim to analyze the response of electromagnetic waves in the frequency domain greater than 1.0 kHz. Additionally, we confirmed that the resistivity of wet soil or wet rock exhibits a frequency-dependent effect with high-frequency electrical waves in a laboratory. We also attempted to analyze the underground wet zone using Drone-AEM data. To analyze an underground wet zone, we used the Drone-AEM data obtained at a steep slope around the slope collapse in the Iya area of Tanabe city, Wakayama prefecture, Japan. The Results indicate that the groundwater level estimated from existing data, such as boring data, was suitably consistent with the wet zone analyzed by this method. We thus confirmed that this analysis is an effective method for surveying underground wet zones.	2) Analysis of underground wet zones using an airborne electromagnetic drone and frequency-dependent effect
June 9		*Akira Jomori・Atsuyoshi Jomori・Tetsuya Toyama(Neo Science),Kentaro Kanayama・Katsushi Kawato(Nippon Engineering Consultants),Atsuhiko Kinoshita(Ministry of Land,Infrastructure,Transport and Tourism[MLIT])
June 9	10:10 - 10:30　　　 It is well known that the electromagnetic wave velocity is related to the volumetric water content of the soil in the field of ground penetrating radar. The quantitative investigation of the relation between attenuation rate and water content is not sufficiently discussed. In this study, we focused on the attenuation rate spectrum by wideband measurement of electromagnetic waves and clarified the relation between the water content and the attenuation spectrum. We conducted a wideband measurement by the vector network analyzer and the antennas with separated bandwidths. We used the large water tank filled with silica sand and changed the saturated groundwater level. It was confirmed that the attenuation rate spectrum changed when the water content of silica sand increased due to the rise of groundwater level.	3) Study for water content estimation by wideband electromagnetic wave measurement
June 9		*Keiichi Suzuki・Taro Kusagaya・Eiichiro Nishiyama(Kawasaki Geological Engineering),Tadanori Goto(University of Hyogo),Jun Ando・Takehiro Ohara(Kawasaki Geological Engineering)
June 9	10:30 - 10:50　　　 Recently， the ice on the moon has attracted considerable attention as an energy source． This ice is expected to exist around the moon’s polar． However， its distribution and amount remain unclear． Therefore， to reveal these factors， we adopted surface wave method． We applied surface wave method to frozen regolith simulants with several water content patterns and observed that higher water content resulted in a faster phase velocity． Thus， surface wave method can estimate the distribution and storage amount of ice even on the lunar surface． In the future， we would like to experiment with different forms of ice and develop exploration devices and systems workable on the earth and the moon．	4) Examination for the effects of ice on S wave velocity of regolith simulant.
June 9		*Tomoki UCHIGAKI・Takeshi TSUJI(Kyushu University),Naohiro UYAMA・Hiroshi KANAMORI(Shimizu Corporation)
June 9	10:50 - 11:10　　　 Recently, lunar development has been accelerated, and the acquisition of the ice on the moon has attracted attention. However, the distribution and amount of the ice remain unclear. Understanding the relationship between ice saturation and elastic wave velocity enables us to estimate its amount. Therefore, to reveal this relationship, we applied numerical simulation to digital image data of regolith simulant. We adopted two rock physics models to assume the adhesion condition of the ice: Matrix-supporting and Contact-cement models. Each model showed different trends of velocity change caused by increased ice saturation. The relationships between ice saturation and elastic wave velocity indicate modeling of the ice using Contact-cement model is better because it is qualitatively consistent with the results from laboratory experiments. This research reveals that modeling ice is important in estimating the ice saturation from elastic velocity.	5) Relationship between ice saturation and elastic velocity of lunar regolith simulant based on digital rock physics: Insight into lunar ice exploration
June 9		*Kunimasu Takaaki・Sawayama Kazuki・Ikeda Tatsunori・Tsuji Takeshi(Kyushu University)
June 9	11:10 - 11:30　　　 The authors have been developing a non-destructive testing method, named as "non-lift test", for evaluating residual tensile force of ground anchors. We have already found that the natural resonance frequency of the free length of an anchor can be observed at the extra length of the anchor by using swept vibration. On the other hand, it was relatively hard to find a resonance point for the wedge fixing type anchor composed of multiple steel strands. This time, we applied the non-lift test to a wedge fixing type anchor and found a clear resonance point.	6) Non-lift test for residual tension of ground anchors composed of multiple steel strands
June 9		*Hideki Saito(OYO),Mitsuru Yamazaki(Central Nippon Highway Engineering Nagoya),Atsushi Yashima(Gifu Univ.),Kazuki Nawa(Nagoya City),Kyosuke Kojima(JR Tokai),Takanori Ogahara(OYO)


Session	[Resource Exploration 1][Oil and Natural Gas][CO₂]
Chair	Haruhiko Suzuki(OYO Corp.)

June 9	13:00 - 13:20　　　 Oil slick detection for oil seep identification is automated with the help of machine leaning object detection technique. Although the performance of the trained detector is limited due to the reliability and quality of the training data, the result of the detection seems promising. Further, post-detection automatic clustering makes it possible to extract oil seep candidates.	7) Oil slick detection by machine learning for oil seep identification
June 9		*Shogo Narahara(JGI, Inc.),Kenichi Akama(JAPEX Co., Ltd.),Nobuhito Shimada(JAPEX Co., Ltd. (formerly))
June 9	13:20 - 13:40　　　 Authors constituted a research consortium and advance the Marine CSEM technological development based on the know-how cultivated at university and research institute. A Marine CSEM demonstration survey using our MCSEM system was conducted offshore Japan in September 2020. In the survey, we confirmed an enhancement of the SN ratio by improving the transmission source power, and we introduced appropriate noise suppression processing techniques. Our results indicate that we constructed an effective data preprocessing method for Marine CSEM surveys.	8) Technological development for the domestic insourcing of Marine CSEM survey -A strategy for de-noising in 2020 Marine CSEM demonstration survey-
June 9		*Yosuke Teranishi・Taku Okamoto・Motonori Higashinaka(JGI),Katsumi Takai(JAPEX),Takafumi Kasaya(JAMSTEC),Tada-nori Goto(University of Hyogo)
June 9	13:40 - 14:00　　　 We have developed Python and MATLAB program that can calculate the electromagnetic response of arbitrary transmitter and receiver locations on the ground, subsurface, underwater and in the air. This program can handle ten types of EM transmitters including VMD/HMD, VED/HED, loop, grounded-wire, and arbitrary-shaped line source. That is mainly expected to simulate a monitoring of injected CO2 for CCS. It enables us to analyze the EM responses of horizontal layered model with arbitrary 3D configurations of transmitters and receivers. The developed code has four options of calculation for semi-infinite integrals of the Bessel function and three methods of frequency to time domain transformation. In addition, we can use an arbitrary transmitter waveform for transient EM response that is suited for the actual survey situation. We verified the effectiveness of the program in the simulation on CCS monitoring, geothermal exploration, and marine/airborne electromagnetic methods.	9) Development of electromagnetic response calculation program in horizontally layered earth with arbitrary transmitter/receiver deployment and transmitting current waveform.
June 9		*Yasuyuki Kuwai・Masato Kawai・Haruhito Takuma・Takumi Okada・Takumi Ueda (Waseda Univ.),Takahiro Nakajima (RITE)
June 9	14:00 - 14:20　　　 In CO2 geological storage, monitoring of the injected CO2 is crucial for predicting a risk of CO2 leakage from the reservoirs, increasing efficiency of CO2 storage, and reducing a risk of fluid injection-induced seismicity. In conventional time-lapse seismic monitoring, the interval of the seismic surveys is usually long because of its high cost, and it is difficult to identify unexpected changes in reservoirs (e.g., accidental leakage). To address these issues, we have developed continuous monitoring system based on permanent source system and distributed acoustic sensing (DAS). Here we deployed the monitoring source system at the coast of Kamaishi city and used DAS for the seafloor fiber optic cable to record the monitoring source signal in high spatial resolution. Because the fiber optic cable is deployed on the seafloor, the DAS data is sensitive for horizontal motion. We succeeded to receive the monitoring signal by the seafloor cable, although the signal was not clear based on conventional analysis. Therefore, we optimized the deconvolution parameters and developed filter for the monitoring signal enhancement. As a result, we identified the monitoring signal >10km horizontal distance using 1 month data. This monitoring system enables us to continuously monitor multi storage sites with low cost.	10) Continuous monitoring system for offshore CO2 storage site based on continuous seismic source system and DAS
June 9		*Takeshi Tsuji・Fernando Lawrens-Hutapea・Tatsunori Ikeda(Kyushu Univ.),Masanao Shinohara(Univ. Tokyo),Koshun Yamaoka (Nagoya Univ.),Tsunehisa Kimura(Schlumberger),Ryunosuke Kikuchi・Terumitsu　Tsuchiya(NC Geophysical Survey)
June 9	14:20 - 14:40　　　 Monitoring the change in subsurface condition caused by CO2 injection is crucial in CO2 underground storage. Monitoring of pressure buildup can give effective information for injection operation, while monitoring of expansion of CO2 plume is fundamental for reducing leakage risks. At the same time, however, the monitoring work should be applied at a low cost so as to meet the long-term implementation requirements. In recent years, DAS (Distributed Acoustic Sensing) technology has been attracting attention as a low-cost seismic recording technology, which uses permanent fiber-optic cable as distributed vibration sensors. In this study, numerical experiments of “Flex Refraction” (the Flexible 3D seismic refraction method; Kikuchi et al., 2020) by using the seafloor deployed fiber-optic cable were conducted to find applicability of the DAS technology for the pressure buildup monitoring. As a result, it was found that the DAS-based Flex Refraction using the shooting vessel is a promising low-cost and wide-area pressure monitoring technique. This result also suggests that the method, if combined with the Continuous Seismic Source installed on the seafloor (Tsuji et al., 2021), will allow for more frequent pressure monitoring.	11) Pressure monitoring in the sea using DAS; Numerical study by 3D seismic refraction survey.
June 9		*Ryunosuke Kikuchi・Terumitsu Tsuchiya・Rei Sato・Yuya Yamazaki(NC Geophysical Survey),Takeshi Tsuji(Kyushu Univ.),Hajime Yamamoto・Yusuke Hiratsuka・Atsuhiro Miyagi(Taisei Corporation)
June 9	14:40 - 15:00　　　 For underground structure exploration using elastic waves, wave sources and detectors are allocated at appropriate places around target underground. This paper proposes simultaneous parallel measurement using finite length codes without interferences in a certain time interval. The codes are constructed from convolving orthogonal complex valued sequence pairs and a real valued sequence. Transmission codes are constructed from cosine and sine waves weighted by real and complex parts. In a receiver, complex sequences are demodulated and parallelly correlated to outputs for desired signals. Signal processing of senders and receivers are digitally executed, and control and data acquisition are cooperated with networks.	12) A Method of Simultaneously Measuring Multiple Elastic Transmission Characteristics Using Orthogonal Finite Length Codes
June 9		*Yoshihiro Tanada・Masato Iguchi・Tomoki Tsutsui(Kyoto University DPRI)


Session	[Earthquake 1][Disaster Prevention 1]
Chair	Keiichi Suzuki(Kawasaki Geological Engineering)

June 9	15:20 - 15:40　　　 A S-wave velocity structure model of southern Okayama prefecture area was constructed by using conventional published data. In this purpose, dispersion curves of micro-tremor array method and Bouguer anomaly data were used as deep underground information and velocity logging, N-value and AVS30 data were used as shallow underground information. Joint inversion method was applied to integrate all these data. The result derived from the joint inversion is consistent with conventional results by using gravity analysis, which were reported by Koga et al. (2009) and Nishimura et al. (2011). In addition, integrated model reflects shallow data such as N-value and AVS30. These results suggest that the joint inversion method is effective to build a shallow and deep integrated ground model.	13) Construction of 3-D S-wave velocity structure model of southern Okayama prefecture by joint inversion method
June 9		*Yoshihiro Sugimoto(DIA Consultants Co., Ltd.)
June 9	15:40 - 16:00　　　 Understanding the subsurface structure is important issue in a wide range of fields such as earthquake hazard prediction, geological interpretation, and resource exploration. Development of accurate estimation methods of the velocity structures are required. In this study, we estimated surface-wave phase velocities around the Nara basin using the cross-correlation of ambient seismic noises. To obtain robust surface wave dispersion curves, we used both the diagonal and off-diagonal components of the cross-correlation with a zero-crossing method based on the SPAC method. Finally, phase velocity maps were obtained by a phase velocity tomography. As a result, clear contrast in the phase velocity was illustrated: low velocity area in the west and high velocity area in the east, which associated with the difference of the geologic structure.	14) Estimation of the Rayleigh-wave dispersion curve using the off-diagonal components of the cross-correlation function
June 9		*Hiro Nimiya・Masayuki Yoshimi(AIST)
June 9	16:00 - 16:20　　　 Fujikawa-kako fault zone has high earthquake occurrence rate in the next 30 years. It is suggested that the earthquake in Fujigawa-kako fault zone will have a tremendous impact at the national level. In this study, we tried to improve the subsurface structure model of this area by carrying out temporary strong motion observations and microtremor measurements. At first, we modified a previous subsurface structure model by using observed phase velocities of the Rayleigh wave. The spatial distribution of the phase velocities was corrected at the observed points and around the observed points. We used different wavelengths for each phase velocity with spatial smoothing. There for we could construct the modified model where boundary surface changes smoothly as the deep layer. It was found in the strong motion simulation for the previous and new models that the arrival times of the S wave and the reproducibility of the surface wave is higher than previous one.	15) Subsurface structure modeling around the Fujikawa-kako fault zone for strong motion simulation
June 9		*Seina Hashimoto・Hiroaki Yamanaka・Kosuke Chimoto(Tokyo Inst. Tech.),Hiroe Miyake(Univ. of Tokyo),Shigeki Senna(NEID),Masayuki Yoshimi(AIST),Masahiro Korenaga・Seiji Tsuno(Railway Tech. Res. Inst.)
June 9	16:20 - 16:40　　　 The distribution of active faults can be evaluated from that of heat flow. We calculated heat flow values around the Nankai Trough off south Japan from bottom-simulating reflector (BSR) appeared on reflection seismic profile. In order to obtain accurate seismic profile needed for precise heat flow calculation, we applied automated velocity analysis. In this analysis, P-wave velocity can be determined to all CMP gathers, and we could obtain accurate depth model. Our seismic reflection revealed BSR clearly. The estimated heat flow value reached approximately 40-80 (mW/m2). This result is consistent with previous studies. Moreover, We identified an area with anomalous high heat flows value. The area is located in outer ridge region where the strike-slip faults exist. The presence of such faults supports the interpretation of heat flow anomalies estimated by our approach.	16) Heat Flow Estimation Derived from Automatic Velocity Analysis : Insight into Spatial Distribution of Active Faults
June 9		*Takenouchi Shuto・Mukumoto Kota・Tsuji Takeshi(Kyushu University),Kodaira Shuichi・Fujie Gou・Nakamura Yasuyuki・Shiraishi Kazuya(JAMSTEC)
June 9	16:40 - 17:00　　　 Hakodate-heiya-seien fault zone is one of the major active fault zone in Hokkaido, and this active fault continues from western margin of the Hakodate plain to the west coast of the Hakodate bay. In Hakodate bay, high-resolution multichannel sonic survey and ultra-high-resolution single-channel sonic survey were conducted in order to clarify the fault structure beneath the ocean floor. However, it is difficult for sonic survey to clarify the geological structure beneath the ocean floor which is covered with gravel or bedrock. Therefore, the purpose of this sturdy is to conduct the marine magnetic survey at the blank area of sonic survey data that could not obtain by the effect of marine sediment in Hakodate bay and examine the efficacy of the marine magnetic survey in the active fault research.	17) Marine magnetic survey in the offshore extension of Hakodate-heiya-seien fault zone, Hokkaido
June 9		*Junya Maruyama・Naoyuki Higaki(HRO)
June 9	17:00 - 17:20　　　 MT responses on the seafloor are very sensitive to coastal effects and bathymetric effects. Therefore, an inversion analysis with a simple initial model may yield an incorrect resistivity structure. Based on the observed MT responses around the Hyuga-nada area, we performed 3D inversion analyses with two types of initial models: simple two-layer models and a three-layer model. The latter model produced the inverted model of which responses correspond to the observed values well. However, in the former cases, the visualized resistivity structures vary greatly depending on the background resistivity of the initial models. When we set it to high resistivity, the responses from the inverted models did not match the observed values. Hence, it is important to set an appropriate initial model in MT analyses using seafloor data.	18) 3-D resistivity analysis around the Hyuga-nada area based on marine-MT data
June 9		*Hayato Nakamura・Hiroshi Ichihara(Graduate school of environmental studies, Nagoya University),Tada-nori Goto(Graduate School of Life Science, University of Hyogo),Tetsuo Matsuno(Kobe Ocean Bottom Exploration Center, Kobe University),Noriko Tada(Research Institute for Marine Geodynamics, Japan Agency for Marine-Earth Science and Technology),Shinya Sato(Kyoto University, Graduate School of Engineering)



June 10, Room No.1

Session	[Volcano]
Chair	Tatsunori Ikeda(Kyushu Univ.)

June 10	9:30 - 9:50　　　 A pilot seismic experiment for volcanological monitoring is presented, which includes airgun shootings and a borehole seismometer array. The experiment aims at tuning fundamental shooting parameters and obtaining know-how to run continuous surveys at volcanoes. A preferable shooting parameter set and one of the target later arrival from depth of Aira Caldera were successfully obtained through the experiment.	19) A pilot experiment for monitoring of Aira Caldera with using seismic method; An feasibility study of airgun shooting across the caldera
June 10		*Tomoki Tsutsui・Takeshi Tameguri・Daisuke Miki・Masato Iguchi(DPRI, Kyoto University),Masanao Shinohara(ERI, University of Tokyo)
June 10	9:50 - 10:10　　　 It was reported that volcanic activity has been getting active near Hachijojima recently (Kimura et al.,2002). In order to evaluate a risk of eruption, we will analyze underground structure by technique using natural earthquakes. Hence, we need many and accurate first arrivals of seismic waves generated by natural earthquakes to improve the accuracy of the analysis. However, the number of earthquakes near Hachijojima is small and it takes much time to identify first arrivals of seismic waves manually. Therefore, we try to increase the number of earthquakes, to automate the identification and to improve the accuracy of the identification by applying deep learning. Deep learning requires a large-scale dataset. It takes vast time to prepare the dataset. In order to solve the problem, we use the method called fine-tuning. The fine-tuning is a method to retrain a part of a trained model’s parameters with a small amount of data and enables to extract the features of data efficiently. We applied Generalized Seismic Phase Detection with Deep learning (GPD) (Ross et al., 2018) to waveform data observed in Hachijojima. But there were many false detections. Therefore, we applied fine-tuning to solve the problem.	20) Automatic Detection of First Arrival Time of Seismic Waves with the Fine-tuning: Applying to Observed Data in Hachijojima
June 10		*Hikaru Kunimasa・Hiroyuki Azuma・Yoshiya Oda(TMU)+R25Q25:T25
June 10	10:10 - 10:30　　　 Hachijojima Island is a volcanic island belonging to the Izu Islands. In 2002, a volcanic earthquake swarm and a crustal movement were observed. Hachijojima Island is one of the 50 active volcanoes that is continuously monitored its activity by JMA. In order to interpret volcanic activity correctly, it is effective to understand the underground structure of the volcano by means of geophysical method. In this study, in order to estimate the underground structure of Hachijojima Island in higher resolution, we installed 46 stations of seismograph on Hachijojima Island and Hachijo-kojima Island and conducted temporary seismic observation from September 2019 to March 2020 (182 days). As a result of 3D natural earthquake tomography, we estimated detailed 3D velocity structure up to the depth of 8km just beneath Mt. Nishiyama. There are two high Vp anomalies at the depth of 2-5km just beneath Mt. Nishiyama. These anomalies are continuous from deeper area and may represent the detailed structure of dykes penetrated in 2002.	21) Three-Dimensional Crustal Structure beneath Hachijojima Island using dense seismic observation data
June 10		*Shotaro Kanke・Hiroyuki Azuma・Yoshiya Oda(Tokyo Metropolitan Univ.)
June 10	10:30 - 10:50　　　 This paper presents the attenuation characteristics of Hachijojima Island estimated from twofold spectral ratio method. Twofold spectral ratio method removes the source and site amplification characteristics by using multiple earthquake records observed at multiple stations, and extracts path characteristics, attenuation characteristics. Earthquake data observed by a dense seismic network of 46 stations in Hachijojima Island during the period from September, 2019 to March 2020 were used. As the result, we found that the low seismic attenuation area is located in the same area as the dike intrusion model (Kimata, et al., 2004) and the high seismic attenuation area is situated in the area of tuff cones and tuff rings formed by phreatomagmatic explosions (Sugihara and Shimada, 1998). The high seismic attenuation regions in Nishiyama and Higashiyama volcanos roughly correspond to the low-velocity region of the velocity structure of Hagiwara and Watanabe (2019).	22) Attenuation characteristics of Hachijojima Island from twofold spectral ratio method using dense seismic observation
June 10		*Utako Watanabe・Hiroyuki Azuma・Yoshiya Oda(Tokyo Metropolitan Univ.)
June 10	10:50 - 11:10　　　 In this study, as a volcanic disaster prevention study, the surface displacement of Hachijojima Island, a volcanic island belonging to Tokyo Metropolitan, has been estimated by InSAR time series analysis, ALOS-2 PALSAR-2 data observed during the period from August 2014 to August 2020 are used for InSAR analysis. In this study, we employed SBAS(Small Baseline Subset) method, which in one of the methods for time series analysis. 83 pairs for Descending, and 26 pairs for Ascending are used for SBAS analysis. In addition, we analyzed GNSS data observed at JMA “South mountainside of Nishiyama” station to verify the results. The results obtained by time series analysis are described. On Hachijojima Island, no regions with surface displacement of 4 cm or more was observed in 6 years. But it was confirmed that analysis results of both Descending and Ascending showed uplifting surface displacement in the flat area between Nishiyama and Higashiyama and near the summit of Nishiyama. Specifically, a displacement of about 4 cm was confirmed on the flat land, and a displacement of about 2 cm was confirmed near the summit of Nishiyama. Those area correspond to the area of the lavas of the Senjojiki stage and Okoshigahana Scoria fall.	23) Surface displacement in Hachijojima Island since 2014 by InSAR analysis and GNSS
June 10		*Yusei Hamamura・Hiroyuki Azuma・Yoshiya Oda(Tokyo Metropolitan Univ.)
June 10	11:10 - 11:30　　　 Goshogake geothermal area is on the foot of the Akita-Yakeyama volcano and it has fumaroles, mud pots, hot springs etc. Recent years geothermal activity has destroyed the walkway named Nature study route of Goshogake spa and the walkway is partly off limits. In order to investigate the subsurface hydrothermal structure, we conducted geophysical prospecting including SP (Self-Potential), GPR(Ground Penetrating Radar), induction method by the instrument called GEM-2, and ERT (Resistivity method). We found the positive anomalies of SP are in accordance with the geological alteration zones. In addition, GPR scatterings are outstanding at resistive zones beneath a survey line along the shore of the Ooyunuma hot pond.	24) Self-potential and resistivity at Goshogake geothermal area in Towada-Hachimantai National Park, northeast of Japan
June 10		Yuya Tada (Hitachi Construction),Shura Taomote・Masato Sugano・*Shin’ya Sakanaka・Keigo Tsuboe・Ritsu Sakiyama (Akita Univ.),Tomoya Moriwaki (Nittoc Construction), Yusuke Inoue (OYO Corporation)


Session	[Earthquake 2][Disaster Prevention 2]
Chair	Hiro Nimiya(AIST)

June 10	13:00 - 13:20　　　 In addition to microtremor surveys using a miniature array by Yamamoto et al. (2019) in the Aoyama and Motomiya areas of Morioka City, Iwate Prefecture, we have carried out microtremor observations using a three-point array with an interval of 10 to 20 m. S-wave velocity structures from a ground surface to about 30 m depth were estimated and the site amplification factors from the AVS were evaluated. Regarding the S wave velocity up to a depth of 20 m in Motomiya area, the S wave velocity of the Vs 500 m/s layer assumed by Yamamoto et al. (2019) actually were modified to a value of Vs of about 400 to 450 m/s. However, there was almost no difference in the ground amplification factor. In the Aoyama area, regarding the S wave velocity up to a depth of 20 m, the assumed S wave velocities of the Vs400 m/s layer was modified to a value of Vs about 350 m/s on the west side and was modified to a value of Vs about 400 m/s on the east side. Regarding the ground amplification factor, many observation points with a difference of 0.1 or more from Yamamoto et al. (2019) were confirmed. At least, in the Aoyama area, where the ground amplification factor was larger and the seismic intensity distribution was larger, it was not possible to obtain a highly accurate ground amplification factor only by observing a very small array (radius 0.6 m), so it was necessary to perform array observation with an interval of about 10 m.	25) Evaluation of site amplification factors by microtremor array surveys for Aoyama area and Motomiya area in Morioka City, Iwate Prefecture
June 10		*Hidekazu Yamamoto(Iwate University),Rin Sugo(Morioka City),Tsuyoshi Saito(Iwate University)
June 10	13:20 - 13:40　　　 We conducted 2D microtremor array measurements to delineate a deep subsurface S-wave velocity model in Toge area, Monzen-machi, Ishikawa Prefecture. Total length of the survey line was 1100 meter. Forty single-component nodal seismograph (McSEIS-AT) with geophone (natural frequency is 2 Hz, SUNFULL; PS-2B) were used for the measurement of each sub-array. The H/V spectral ratio and the phase velocity were used in a multi-station inversion with Genetic algorithm. The results of the multi-station inversion were consistent with logging data. The inverted depth of the top of the basement varies from about 15m to 40 m.	26) 2D microtremor array measurements for estimation of shallow subsurface structural model in Toge area, Monzen-machi, Ishikawa Prefecture.
June 10		*Haruhiko Suzuki(OYO Corp.),Shinichiro MORI(Ehime University),Chisato KONISHI(OYO Corp.),Takumi KOBAYASHI(PWRI)
June 10	13:40 - 14:00　　　 To quantitatively evaluate the effects of changes in the physical parameters of surface ground due to the freezing in the cold region, we carried out various geophysical observations and surveys such as, continuous seismic observation, surface wave exploration, array microtremor observation, electrical survey, and GPR survey in Kitami city. At first, in this paper, we reported the outline of the observations and the surveys. Secondary, we showed the results of initial analysis of seismic observation data and linear-array vibration data by weight drop.	27) Study on the site amplification characteristics in the cold region by the geophysical explorations at Kitami city : Outline and results of initial analysis of seismic observation data
June 10		*Seiji Tsuno(Railway Technical Research Institute),Hiroaki Yamanaka(Tokyo Institute of Technology),Nobuo Takai(Hokkaido University),Shunzo Kawajiri(Kitami Institute of Technology),Naofumi Nakagawa(Hokkaido University),Shingo Nomoto・Jun Horita(Geo-Technical Research Co.,Ltd),Michiko Shigefuji(Kyushu University)
June 10	14:00 - 14:20　　　 We carried out the surface wave exploration by multi-channel active source in Kitami city of the cold region. We selected two sites inside Kitami Institute of Technology and the vicinity of KiK-net Rubeshibe station (ABSH13). We can see that the recorded waves are changed by several seasons at both sites. We analyzed F-K spectra from these waveforms to get the phase velocity of the Rayleigh waves. From these results, we grasped the variation of the phase velocity from October to March. The phase velocities of the cold season are higher than those of the other seasons. We will discuss site amplification characteristics by comparing them with other geophysical exploration data.	28) Study on the site amplification characteristics in the cold region by the geophysical explorations at Kitami city : Variation on the results of the surface wave analysis by a multi-channel active source
June 10		*Nobuo Takai・Akari Tsukamoto・Naofumi Nakagawa・Momoko Iwazaki(Hokkaido University),Shingo Nomoto・Jun Horita(Geo-Technical Research Co.,Ltd),Michiko Shigefuji(Kyushu University),Shunzo Kawajiri(Kitami Institute of Technology),Seiji Tsuno(Railway Technical Research Institute),Hiroaki Yamanaka(Tokyo Institute of Technology), et al.
June 10	14:20 - 14:40　　　 For evaluating the effect of freezing of the ground on the site amplification characteristics, we have been carrying out the geophysical explorations at Kitami city, Hokkaido, since October, 2020. The results on the process of freezing of the ground are as follows. 1) The electrical survey: The VES-curve changed its shape from K-type to H-type and/or Q-type as the freezing of the ground progressed. Also, the ground surface showed significantly higher resistivity in the period of freezing of the ground. 2) The GPR survey: Change of location of the reflection in depth and its disappearance were recognized due to freezing of the ground. The results suggest that the electrical survey and the GPR survey are useful for estimating the process of freezing of the ground.	29) Study on the site amplification characteristics in the cold region by the geophysical explorations at Kitami city : Variation on the results of electrical survey and GPR survey
June 10		*Tetsunori Kishikawa・Shingo Nomoto・Jun Horita(Geotechnical Research),Hiroaki Yamanaka(Tokyo Institute of Technology),Shunzo Kawajiri(Kitami Institute of Technology),Seiji Tsuno(Railway Technical Research Institute),Nobuo Takai(Hokkaido University),Michiko Shigefuji(Kyushu University)
June 10	14:40 - 15:00　　　 A spatially dense microtremor measurements has been carried out at OYO Tsukuba office using node type 3-component seismometers (McSEIS-AT (3ch)). Horizontal to vertical spectral ratio (HVSR) was computed at 375 measurement points with 10 m spacing in approximately 140 m by 350 m area. The peak frequency of the HVSR in a range between 2 and 10 Hz shows lower value at the area where S-wave velocity is relatively low. MDS is applied for the dimensionality reduction and the visualization of the similarity or dissimilarity of the HVSR at all measurement points. Then, the data after the MDS is classified by k-means classification method, and the low S-wave velocity region is detected by the classification result. These findings indicate the possibility of using HVSR in the automation process for the classification of subsurface structure at near surface.	30) Classification of shallow subsurface by H/V spectral ratio of a spatially dense microtremor measurements
June 10		*Chisato Konishi・Haruhiko Suzuki・Osamu Ishitsuka(OYO Corporation)


Session	[Civil Engineering 2]
Chair	Nobuo Takai(Hokkaido Univ.)

June 10	15:20 - 15:40　　　 As the law for promoting offshore wind power was implemented in April 2019, it is expanding the submarine ground survey market for the construction of offshore wind power generation. In general ground surveys, the geological and physical properties are evaluated by boring, but offshore boring has a longer survey period and is more expensive than land boring, there is a strong demand for simpler survey method. The authors have developed “Offshore Microtremor Array Exploration” for the purpose of grasping S-wave velocity structure in a short time and reported its applicability at this conference in the spring of 2 years ago. In this paper, we will report on the survey conditions, seafloor conditions, and characteristics of offshore microtremor obtained from offshore microtremor array exploration conducted at various sites.	31) The application in offshore ground survey：The case study of Offshore microtremor array method - Characteristics of Offshore Microtremor obtained through data analysis -
June 10		*Mikihiro Imai・Yusuke Inoue・Masanori Ebato・Akesi Hiraide・Yoshikazu Matsubara(OYO)
June 10	15:40 - 16:00　　　 Computing crosscorrelations of ambient noise is an effective approach to extract seismic responses between two stations using seismic interferometry. However, the heterogeneous distribution of the ambient noise sources would interfere in constructing seismic responses in the calculation of the crosscorrelations. In this study, we identified the ambient noise sources recorded in a dense seismic array and utilized the information for better subsurface imaging. The seismic array was composed of 50 stations installed in the Itoshima Peninsula. By analyzing direction of incoming ambient noise, we found the traffic noise generated surface waves in low frequencies (<10 Hz) and P-waves in high frequencies (>15 Hz). We also identified high frequency surface waves at ~60 Hz generated from the construction related activities. Surface wave velocities between each station pair were extracted by considering the distributions of the ambient noise sources. Such surface wave velocities would be used to construct high-resolution velocity maps by applying tomographic analysis. Thus, identifying the sources of the ambient noise acquired with dense seismic arrays is effective to improve the ability of the ambient noise data to image subsurface structures.	32) Extracting surface and body waves from heterogeneous ambient noise field in the Itoshima Peninsula
June 10		*Tatsunori Ikeda・Takeshi Tsuji(Kyushu Univ.),Chisato Konishi・Hideki Saito(Oyo Corporation)
June 10	16:00 - 16:20　　　 The authors had conducted a high-resolution S-wave reflection survey by means of Land Streamer in 1997 at a site where a concealed active fault had been inferred. Supplemental boring, VSP, CPT, and seismic cone measurements was appended to obtain near surface geological data along the seismic line. As a result, flexure structure was delineated around the inferred fault zone. However, it was difficult to interpret the detailed structure due to the influence of abrupt changes in the uppermost surface structure adjacent the inferred fault. We then reprocessed the S-wave VSP and seismic cone data to utilize a modified S-wave velocity model for the seismic data reprocessing. Finally we could improve seismic depth section, and interpret faulting structure up to 60 m in depth. High-performance and low-cost seismic processing software running on windows PC enabled us to reprocess legacy near surface geophysical data and to reinterpret the detailed near surface geological structure, or fault related deformation structure.	33) Reprocessing and re-interpretation of a legacy near surface geophysical dataset based on VSP and seismic CPT data.
June 10		*INAZAKI Tomio・KANO Naomi(Research Institute of geology and Geoinformation, AIST)
June 10	16:20 - 16:40　　　 Seismic wave tomography is used as an inverse approach to image underground information from observed wave travel time. In this study, we consider a case that observation data in space are not enough because of the difficulty to set sources and receivers in a subsurface ground, while data observation may be done to continuously monitor underground conditions for certain time period with appropriate intervals. Assuming that the time-lapse data sets are available for small and local variations in underground condition, the data sparsity can be constructed by taking the difference of data in time and space. For such sparse data, we develop a new LASSO inversion in conjunction with an ADMM scheme, and apply the method to simulated data with signal noise. As the result after some examination, it reveals that the new LASSO produces inverse solutions with better accuracy than the conventional LASSO.	34) Seismic wave tomography based on sparse modeling for time-lapse data
June 10		*Xiaoyuan Qiu・Sohichi Hirose(Tokyo Institute of Technology),Yoshiaki Yamanaka・Tomoko Emoto・Hiroshi Imai・Shunichiro Ito(Suncoh Consultant)
June 10	16:40 - 17:00　　　 Ground Penetrating Radar (GPR) is a type of geophysical surveying technique that investigates the underground conditions based on the shape of the reflected waves of electromagnetic waves. The interpretation of the data is mainly done visually by skilled technicians. Recently, with the development of data acquisition systems, it has become possible to acquire GPR data in three dimensions at high speed, and the number of data to be interpreted has increased. Therefore, there is a need to automate and overcome the time-consuming problem. In the application of deep learning to GPR data, research has been conducted using two-dimensional depth-section images. However, it is considered that a more reliable extraction of event waveforms can be achieved by making a comprehensive judgment from 3D data. Therefore, in this study, we compare and discuss the results of learning 2D data in depth cross section and time slice by handling 3D data.	35) An attempt to detect events by deep learning on 3D GPR data
June 10		*Ito Yoshiki・Azuma Hiroyuki・Oda Yoshiya(Tokyo Metropolitan University),Matsuoka Toshifumi(Fukada Geological Institute),Shinohara Jun(CANAAN GEO RESEARCH)



June 11, Room No.1

Session	[Resource Exploration 2][Geothermal][etc.]
Chair	Takumi Ueda(Waseda Univ.)

June 11	9:30 - 9:50　　　 Electrical impedance tomography (EIT) applies a weak current from the surface of the body, and obtains a cross-sectional view of the inside of the human body from the resistivity distribution. Compared to X-ray CT and MRI, EIT has advantages such as non-invasiveness to the human body and miniaturization of the device. However, there is no detailed discussion on the frequency characteristics of the electrodes, which are essential for EIT, and the differences in stability depending on the electrode material. In addition, experiments using simulated organisms are required before clinical trials, but the frequency characteristics of simulated organisms also have hardly been studied. In this study, we examined the basic properties of electrodes and simulated organisms. The small-size non-polarized electrode was made by ourselves, indicating a suitable stability for EIT. The simulated organisms were also made by ourselves, and used for the test experiments of EIT in the laboratory. As a result, we succeeded to image the internal structure of the simulated organisms by our EIT survey.	36) Application of Electrical Resistivity Tomography to Medical Diagnosis
June 11		*Yuki Oshima・Tada-nori Goto(University of Hyogo)
June 11	9:50 - 10:10　　　 The first magnetotelluric (MT) response function estimator is based on the least-square theory; the robust procedure can improve its performance. The robust M-estimator gives a small weight to reject the outlier based on the residual between the predicted electric field by the LS estimator and the observed electric field. M-estimator can reduce the influence of unusual data in the electric field (outliers) but are not sensitive to exceptional input (magnetic field) data, which are termed leverage points. The bounded influence (BI) estimator combines the standard robust M-estimator with leverage weighting based on the hat matrix diagonal element's statistics. Alan Chave also creates an open-source code (BIRRP), and it is widely used in the MT community. Chave (2004) showed that the BI-estimator would perform better than M-estimator, but not all the cases. The leverage point corresponds to the large variation of the magnetic field. It may be an energetic signal or active noise. This paper will introduce an M-estimator to compare with the BIRRP code and demonstrate a case study that the leverage point corresponds to the energetic signal. At this condition, the M-estimator performs better than the BI-estimator.	37) Comparing the bound influence and maximum likelihood magnetotelluric response function estimator
June 11		*HAO CHEN・Hideki Mizunaga・Toshiaki Tanaka・Gang Wang(Kyushu university)
June 11	10:10 - 10:30　　　 Fluids trapped under supercritical conditions are attractive unconventional geothermal resources that can yield much higher well-productivity. The magnetotelluric (MT) method can image supercritical fluids as deep conductors. However, MT imaging of a deep conductor strongly depends on the conductor geometries and overlying clay layers. The ability of the MT method to image a deep conductor needs clarification to understand the usefulness of MT for exploring supercritical fluids and to correctly relate the imaged conductor to supercritical fluids. Therefore, this study investigated the MT ability to image a deep conductor of supercritical fluids using numerical tests. The numerical tests used a conceptual resistivity model of a promising supercritical geothermal system in the Kakkonda area, northeast Japan. The test results prove that the MT method is useful for revealing the deep conductor of promising supercritical fluids. The results also showed that the MT ability to image deep conductors strongly depends on the size and depth of the target conductor and the thickness of overlying clay layers. The findings from the test results help to relate deep conductors imaged by the MT method to supercritical geothermal systems.	38) Ability of the magnetotelluric method to image a deep conductor: Exploration of a supercritical geothermal system
June 11		*Keiichi Ishizu・Yasuo Ogawa・Toru Mogi(Tokyo Institute of Technology),Yusuke Yamaya・Toshihiro Uchida(AIST)
June 11	10:30 - 10:50　　　 In order to reveal the location of fracture that contributes to the productivity for geothermal power generation, we focus on diffracted and scattered waves in the distributed acoustic sensing (DAS) monitoring record. Diffraction is generated by the faults and fractures, and in principle, they are encoded minute changes that cannot be captured by reflected waves. These waves generally have only a fraction of the amplitude compared to reflected waves, and have been removed from the record in conventional seismic processing. However, the information is useful for estimating the distribution of the fracture systems.In this study, we develop the diffraction enhancement workflow that combines plane wave destruction (PWD) and common reflection surface (CRS) stack method to separate diffractions from direct arrival in the monitoring data. Compared with the distributions of extracted diffractions and the wellbore information, it is confirmed that the proposed method could be worked as the fracture indicator. Moreover, we conduct diffraction imaging to evaluate the diffraction sources. The distribution indicates the appropriate fracture locations that correspond to the wellbore information and surface seismic profile (SSP) attribute. This proposed workflow can be potentially applied to local acoustic emission (AE) events with DAS and SSP survey for imaging fracture.	39) Diffraction extraction and imaging for DAS monitoring data
June 11		*Keisuke Teranishi・Naoshi Aoki・Moeto Fujisawa(JGI),Shinya Onodera(JOGMEC)
June 11	10:50 - 11:10　　　 We carried out a geothermal seismic study using a distributed acoustic sensor (DAS) at the Ohnuma geothermal power plant owned by Mitsubishi Material Co. in September 2020. The Ohnuma geothermal power station, the third commercial geothermal power plant in Japan, was completed in 1974. We installed an optical fiber system for the DAS measurements on the site. We deployed the optical fiber down to the depth of 1,973 m in the O-13R borehole. To enhance the S/N, we stacked the DAS data for 480 times and correlated them with the vibrator source signature. By stacking for a long duration, we obtained excellent DAS records down to the bottom of the boreholes. Using 2D migration of observed and synthetic DAS seismic records, we recognized intense seismic reflections from 2.8-3.0 km depth, suggesting the possibility of geothermal reservoirs. The density decrease in this zone could be 20%, possibly implying that the fracture zone is filled with fluid. The two field studies in the Medipolis and Ohnuma geothermal fields in Japan showed that the DAS-seismic method in the borehole can efficiently image seismically reflective zones, and the findings suggest high possibility of geothermal reservoirs.	40) DAS-Seismic study in the Ohnuma geothermal borehole in Honshu, Japan
June 11		*Junzo Kasahara・Yoko Hasada・Haruyasu Kuzume(ENAA),Hitoshi Mikada(Kyoto Univ.),Yoshihiro Fujise(WELMA)
June 11	11:10 - 11:30　　　 Currently, water on the Moon surface is considered a potential resource. Furthermore, we need to know geological information to build base camps on the Moon. Therefore, estimating the S-wave velocity of the underground of the Moon has been strongly required. In the Moon exploration, it is difficult to bring heavy and large-sized equipment to the Moon because its weight and size are limited on the spacecraft. Therefore, we investigated the applicability of a small extraterrestrial exploration system based on the microtremor survey without a seismic source. This study evaluates the investigation depth for each arrangement of small seismometer arrays on the Moon if we analyze microtremor data by Centerless Circular Array (CCA) method. Finally, we summarized the points and issues to use the CCA method for exploration on the Moon, considering appropriate array arrangement and noise level to analyze the desired depth. For example, our evaluation demonstrates that we can investigate from surface to 3m depth by using the pentagon-type array with 0.3m radius or the triangle-type array with 0.5m radius. Thus, our evaluation approach is useful to design seismic arrays in lunar exploration projects in future.	41) Utilizing a miniature seismic array system based on CCA analysis for moon exploration
June 11		*Marina Orita・Tatsunori Ikeda・Takeshi Tsuji(Kyushu University)

VENUE
	ONLINE

SCHEDULE
	Wednesday, June 9, 2021
		Openning Speech	Room No.1	09:20-09:25
		__Speaker:	Toshiki WATANABE (NAGOYA Univ.; President of SEGJ)
		Technical Sessions	Room No.1	09:30-17:20

	Thursday, June 10, 2021
		Technical Sessions	Room No.1	09:30-17:05
		Student Exchange Event	Room No.2	17:20-18:50
	Friday, June 11, 2021
		Technical Sessions	Room No.1	09:30-11:30
		Executive Sessions	Room No.1	13:00-15:00
		__Chair:	Hidekazu YAMAMOTO (IWATE Univ.; Vice-President of SEGJ)
		__Presentation(1):	Progress of Research on Environmental Pollution and Issues for Environmental Recovery Caused by the Fukushima Daiichi Nuclear Power Plant Accident and Issues for Environmental Recovery Yuichi MORIGUCHI (NIES)
		__Presentation(2):	CCS Challenges and Prospects from a Perspective of Monitoring Masao SORAI (AIST)
		General Meeting	Room No.1	15:20-16:50

Registered and Paid		Pre-Registration (Until May 13, 2021)	Onsite Registration
Conference Meeting	Member / Non-member	JPY 05,500-**	Pre-Registration Only
	Student (Member)	JPY 00,000-**
	Student (Non-member)	JPY 02,750-**

**	Pre-registrant will receive PAPER Proceedings by mail before the conference.
**	PAPER Proceedings can be purchased at the SEGJ OFFICE for a fee.

LIST OF ADVERTISEMENTS
	Geophysical Surveying Co., Ltd.(GSC) Hakusan Corporation ITOCHU Techno-Solutions Corpration(CTC) JGI, Inc. Kawasaki Geological Engineering Co. Ltd.(KGE)

1) TEST for you internet connection to the online conference
2) Guide for Technical Session
3) DO NOT screenshot or record any part of the conference (including the audios).