

May.30 　　Room No.1（Meeting Room 03，2F）

Session	[Civil Engineering][Earthquake 1][Disaster Prevention 1]
Chair	Kentaro Kasamatsu (Kajima Corporation Technical Research Institute)

May.30	9:30 - 9:50　　　 In first arrival traveltime tomography, borehole-based observation and far offset data are often used to obtain information on the deep subsurface structure.However, in first arrival traveltime tomography, far offset data where the source point exists outside the computational domain are difficult to handle because the physical properties between the far offset excitation point and the computational domain are unknown.In this paper, we estimated the far offset travel times that are incident on the side of the computational domain. We then treated this data as far offset excitation data generated from the side of the computational domain, and applied it to adjoint state　first-arrival traveltime tomography.The analysis shows that using travel time data for both near and far offsets can estimate deeper subsurface structures compared to using only travel time data for near offsets.	1) First arrival traveltime tomography using near and far offset traveltimes
May.30		*Yamauchi Yasutomo(1) 1:DIA Consultans Co., Ltd.
May.30	9:50 - 10:10　　　 For the purpose of site investigations of floating offshore wind farms, which will become mainstream in the future, research and development of deep-sea microtremor array exploration was carried out. The applicable water depth for floating offshore wind power farm is 50m or deeper and the current measurement method for offshore microtremor arrays, that is, measuring by dropping ocean bottom seismometers from the sea level onto the three vertices and center of the triangular array, is not possible at deep water depths. We have developed a method for microtremor array exploration by arranging ocean bottom seismometers in a straight line using an undersea positioning device and conducted measurement experiments. A clear dispersion curves were obtained from the acquired data. In addition, we performed vibration excitation by dropping a weight on the extension line of the linear array, and also conducted surface wave analysis, and obtained clear dispersion curve of shallower part.	2) Development of deep-sea microtremor array exploration for floating offshore wind farm
May.30		*Yoshikazu Matsubara・Yasuhiro Asano・Mikihiro Imai・Yusuke Inoue・Akesi Hiraide(1) 1:OYO corporation
May.30	10:10 - 10:30　　　 In recent years, 2D and 3D microtremor surveys have been frequently used in civil engineering. The direction of microtremors is sometimes biased and changes with time; thus, analyzing this temporal change is essential for estimating S-wave velocity when we can use limited receiver pairs with narrow apertures. In this study, we obtained continuous observation data by a triangle receiver array for three days and examined the temporal change. As a result of the data analysis, we found that two receiver pairs on the triangle sides always give clear frequency-phase velocity images during the experiment. We also found that one of the pairs does not provide a clear frequency-phase velocity image in the daytime, but it does in the nighttime. This observation suggests that the calculated phase velocity is affected by the directions of the microtremors. Therefore, we must consider the measurement time when conducting the microtremor monitoring by limited receivers with narrow apertures.	3) Basic experiment of microtremor measurement for building a microtremor monitoring　system
May.30		*Ryusuke Ohnuki・Masaya Yamauchi・Chisato Konishi・Masaki Tsuji・Om Pradhan(1) 1:OYO Corporation
May.30	10:30 - 10:50　　　 We conducted a convenient field survey to estimate near-surface 2-D S-wave velocity structure at Ikejiri-gawa Depression, close to Nojiriko, Nagano Prefecture. The survey consisted of single-point microtremor measurements and multichannel surface wave surveys (MASW). Downhole Vs measurements using a direct push type seismic cone and SH seismic reflection survey data were supplementally utilized. HVSR curves were derived from the microtremor measurements at a total of 61 points in the depression, 1.5 km long and 500 m wide. In combination, down-sized MASW surveys were carried out at 15 short lines in the depression to estimate 1D Vs structure down to 20 m. Mode of occurrence of multi phase, especially SV reflection events discriminated in the shot records were utilized to constrain the Vs models. We verified the 1D Vs model by checking whether the synthesized HVSR curve matched with the observed one as well as whether the synthesized waveform was concordant with the shot record with special reference to SV reflection events. Correlation with a downhole Vs profile and an SH seismic reflection section was quite effective for the estimation of Vs models. We here propose this combined survey method as a simple field survey tool for estimating near-surface 1D Vs structure.	4) Near-surface Vs modeling in combination of microtremor measurements with surface wave surveys: A case study at Ikejiri-gawa Depression, Nagano Prefecture.
May.30		*INAZAKI Tomio(1),HAYASHI Koichi(2),KONDO Yoichi・SEKI Megumi(3),INOUCHI Yoshio(4) 1:Independent,2:Geometrics,3:Nojiriko Naumann Elephant Musium,4:Formerly Waseda Univ.
May.30	10:50 - 11:10　　　 We developed new methods based on reverse time migration (RTM) for imaging subsurface structures using passive seismic data. Since the proposed technique use only receiver-side information to extrapolate wavefields forward and/or backward in time from receiver points, the detailed information of the original seismic sources is not required for the subsurface imaging with reflection waves or transmitted-conversion waves. In this study, we introduce a case study of the passive seismic reflection imaging using local earthquakes recorded with the MeSO-net dense seismic stations. We achieved the passive seismic RTM imaging of the deep crustal structures related to the dual plate subduction beneath the Kanto region. The result shows the deeper reflection profile than using artificial sources and more detailed structures than from earthquake tomography. The proposed method is promising to be applicable not only to deep structures, as in this study, but also to various subsurface investigations with passive seismic observations at different scales.	5) Subsurface imaging by passive seismic reverse time migration
May.30		*Kazuya Shiraishi(1),Toshiki Watanabe(2) 1:JAMSTEC,2:Nagoya Univ.
May.30	11:10 - 11:30　　　 A FMCW SAR-equipped drone observation system was developed and assessed as a new observation method to complement satellite SAR data, and we used SAR analysis to create SLC images that can identify a point target and topography. We succeeded in clearly estimating the topography of mountainous slopes. The difference in azimuth resolution at different drone speeds was clarified, and it was found that the higher the drone speed, the better the resolution of the topography. To quantitatively evaluate the generated SLC images, three types of coherence were calculated, and we found that the phase coherence was relatively small. The phase coherence needs to be enhanced for InSAR analysis in the future.	6) Point target and terrain estimation from FMCW SAR-equipped drone observations
May.30		*Yutaro Shigemitsu・Kazuya Ishitsuka・Weiren Lin(1),Tomoyuki Sugiyama・Munemaru Kishimoto・Takeharu Takahashi(2) 1:Kyoto univ,2:Nittetsu Mining Consultants Co., Ltd


Session	[Case Study]
Chair	Kazuya Shiraishi (JAMSTEC)

May. 30	13:00 - 13:20　　　 Application of AI and Deep learning to real geophysical survey data has been actively application and discussed. Sakaguchi et al. (2022)1, the possibility of application to automated BSR detection for reflection seismic data was discussed. Subsequently, Sakaguchi et al. (2022)2 discussed its application to BSR detection for data presenting complex geological structures in the 3D seismic processing data. However, application requires not only knowledge of AI but also knowledge of the programing skills and there were several challenges to its use. Therefore, we developed a tool (call by Deep BSR Scanner, DBS) that automatically extracts BSR seismic events. The tool (DBS) can be used by many users and help social implementation. This paper discusses the challenges and results of the tool development.	7) Development of automatic BSR extraction tool using deep learning model(Deep BSR Scanner)
May. 30		*Hironori Sakaguchi・Kazuyoshi Takaichi・Satoshi Kodama・Yusuke Shimono・Hisako Shimizu(1),Shohei Ishinabe・Yoshitaka Nakayama・Ebitani Akira(2) 1:CTC,2:JOGMEC
May. 30	13:20 - 13:40　　　 Seismic reflection surveys were conducted as a part of the “Comprehensive Research Project for the Byobuyama, Enasan Fault Zone and the Sanageyama Fault Zone (Enasan - Sanageyama-North Fault Zone)”. The purpose of the survey was to elucidate the shallow structure of the fault zone to understand the geometry of the multiple faults that compose the fault zone and their continuity in the subsurface. The survey was conducted in September 2021 on main three survey lines, each about 5-7 km long. Some reflection sections lack a clear indication of the fault structure due to the geological structure composed of a granite basement covered by a thin topsoil layer or a weathered layer, and a near-vertical fault structure without a thick fractured zone. In the survey Line 3, several east-dipping reverse faults were identified that correspond to the surface fault traces of the Enasan Fault. It forms the boundary between the basin and the mountainous area and indicates the formation of a structure due to fault activity.	8) Seismic reflection survey of the Enasan Fault, the Sanageyama North Fault, and the Sanage-Saikagawa Fault, and the Sanage-Saikagawa Fault
May. 30		*Toshiki Watanabe・Keiichi Tadokoro(1),Tatsuya Ishiyama(2),Nobuhisa Matta(3),Yasuhiro Suzuki・Koshun Yamaoka・Hiroshi Ichihara(1),Hideo Saito・Mamoru Nakata・Susumu Abe(4) 1:Nagoya Univ.,2:Univ. of Tokyo,3:Okayama Univ.,4:JGI., Inc.
May. 30	13:40 - 14:00　　　 Bursa is the first capital of the Ottoman Empire with the historical importance for Turkey. The city has been suffered from earthquakes. The 1855 Bursa earthquake is one of the large earthquakes in Bursa province. Historical buildings, such as mosques and minarets, in the old city part of Bursa were damaged during the 1855 earthquake (Morita et al., 2023). We have conducted shallow S-wave velocity explorations at 9 damaged sites with microtremor array measurements and vertical electric sounding. We also have conducted the explorations at five strong motion stations in the area. Phase velocities of Rayleigh wave were estimated in a frequency range of 6 to 30 Hz from the SPAC analysis of the array records if vertical microtremors. The phase velocities at sites in the basin are about 200 m/s at a high frequency, while high-frequency phase velocity is 300 m/s at a site in the mountains. The apparent resistivity curves observed at the sites in the basin and the mountains are also different indicating the differences in the shallow geological conditions.	9) Microtremor exploration at damaged mosques in old city of Bursa during the 1855 Bursa earthquake
May. 30		*Hiroaki Yamanaka(1),Masashi Morita(2),Kentaro Kasamatsu(3),Mai Kozu(1),Hiroe Miyake(4),Daiki Sato(1) 1:Tokyo Inst. Tech.,2:Yokohama National Univ.,3:Kajima Tech. Res. Inst.,4:Earthq. Res. Inst., Univ. Tokyo
May. 30	14:00 - 14:20　　　 Two-dimensional microtremor array measurement was conducted along Tone River embankment where some buried valleys are thought to be existed. Although it is thought that consolidation settlement could be caused by weak marine clay layer under the embankments, there are not enough boring data at the site. Through applying two-dimensional microtremor array measurement, we achieved to obtain a wide area of S-wave velocity distribution deeper than that obtained by MASW which has been widely used for river embankment survey. The result is comparable with previous and newly obtained geological data. We conclude that two-dimensional microtremor array measurement is effective for obtaining geological information along river embankments.	10) Estimation of buried valley distribution under river embankment through 2D microtremor array measurement
May. 30		*Atsushi Suzaki・Hidetake Tanaka・Masaya Yamauchi・Sennichi Sato・Takamasa Niibe(1) 1:OYO
May. 30	14:20 - 14:40　　　 In the geological disposal project, data obtained from in-situ and laboratory tests using rock cores are used as characteristic data for constructing a geological environment characteristic model. We need to evaluate the representativeness of the data and the variability of the data. In this paper, we evaluated the representative values and variations of hydraulic conductivity (permeability coefficient) and mechanical strength (uniaxial compressive strength) using the physical logging data conducted in the coastal area of Suruga Bay and the laboratory geophysical test data from Horonobe site. First, we applied a rock physics model by Lima (1995) to derive hydraulic conductivity profiles from multiple physical logging data at Suruga Bay site. In addition, we confirmed the effectiveness of this method in comparison with on-site permeability test data. Next, we applied a rock physics model based on Takahashi et al. (2016) for estimating the unconfined compressive strength from seismic velocity measured by rock cores at Horonobe site and compared with the unconfined compression test data by using rock cores. As a result, we were able to estimate the maximum uniaxial compressive strength, which reflect the property of the rock matrix that is not affected by microcracks in rock cores.	11) Estimation of permeability and mechanical strength characteristics of layers around boreholes using geophysical logging data
May. 30		*KOICHI SUZUKI(1),Kenzo Kiho・Reo Ikawa(2) 1:Kawasaki geological engineering Co., Ltd.,2:AIST
May. 30	14:40 - 15:00　　　 We carried out ERT (Electrical Resistivity Tomography) at Futatsui mine site (quarry site), Noshiro City, Akita Prefecture. Formaly mining of natural asphalt and oil sand was operated here but now ceased. The symptom of seepage of natural oil can be seen even now on the wall of outcrop of the tuff in Pliocene to Pleistocene. The IRIS instrument of Syscal Pro with 48 ch was used for ERT and installed three survey lines at the site and digital data were collected by dipole-dipole and Wenner electrode arrays. The data were inverted with the software Res2Dinv and obtained subsurface resistivity models. The models reveal a certain characteristics feature including a thin resistive sheet on the conductive body. This feature can be interpreted as a kind of water reservoir and small amount of natural oil floated on its water in the porous soil.	12) Subsurface exploration of ERT and GPR at a quarry of oil sand, in Futatsui, Nshiro city, Akita prefecture
May. 30		Kazumasa Nishida(1),*Shin'ya Sakanaka(2),Keigo Tsuboe(3),Shion Tsunaga(4),Kanju Bahati Haji(2) 1:ANA Okinawa Airport,2:Akita Univ,3:HRS,4:SOUGOUCHISHITSU
May. 30	15:00 - 15:20　　　 ALIS is a hand-held dual sensor developed by Tohoku University. Since 2018, it has been tested in Cambodia for humanitarian demining, and our team started a scientific program to introduce this new sensor into humanitarian demining activities effectively to Bosnia and Herzegovina, under the support of NATO SPS Project G5607 ”Accelerating mine clearance by introducing a user-friendly and cost-effective dual-sensor detector in humanitarian demining operations”. This project started in December 2020 and will finish in December 2023. We have trained deminers in Bosnia and Herzegovina and preparing a new SOP for ALIS. Japan International Cooperation Agency (JICA) started a pilot project to support the humanitarian demining activities in Ukraine in January 2023. This project aims introduction of ALIS to landmine clearance in Ukraine. We trained 8 Ukraine deminers in Cambodia with corporation with Cambodian Mine Action Centre (CMAC), where ALIS has been used since 2018. We are using the experience in Bosnia and Herzegovina into the activities in Ukraine. In this paper, we will introduce our activities using dual sensor for Ukraine and Bosnia and Herzegovina.	13) Support of Humanitarian Demining in Ukraine by Dual Sensor
May. 30		*Motoyuki Sato(1)(2) 1:Tohoku University,2:ALISys Co.,Ltd.


May.30 　　Room No.1（Meeting Room 03，2F）

Session	[Poster]
Chair	Keiichi Ishizu (Univ of Hyogo)

May. 30	15:40 - 17:20　　　 We generally estimate the geothermal reservoir, heat source and path of geothermal fluid by conducting the magnetotelluric survey or controlled source magnetotelluric survey. However, these surveys are aimed at deep structure, so it is difficult to estimate the shallow subsurface structure in detail. Therefore, the purpose of this sturdy is to reveal the shallow subsurface structure from a few meters to a few hundreds of meters using magnetic exploration in geothermal area, Hokkaido. We measured survey line at 1m intervals to get the detailed subsurface structure. As a result, we were able to quantitatively evaluate the shallow fault structure. Thus, measuring methods of magnetic exploration obtained by this study are expected to be useful for the geothermal survey.	P-1) A Shallow Subsurface Structure Survey in Geothermal Area Using Magnetic Exploration
May. 30		*Junya Maruyama(1) 1:HRO
May. 30	15:40 - 17:20　　　 I propose a simple method to evaluate the random component of the uncertainty of magnetotelluric (MT) forward modeling for practical three-dimensional (3D) electrical conductivity structure models in a Cartesian coordinate system. The method is based on the idea that the horizontal coordinate system can be selected arbitrarily for a general 3D structure. The synthesized MT responses are ideally identical irrespective of the selection but are different because of the difference in discretization angles, boundary values, and numerical errors. By synthesizing MT responses to the model in several different coordinate systems, the mean, standard deviation, and coefficient of variation can be calculated. These statistics provide quantitative information on how stably the forward calculations synthesize MT responses under the given conditions of the structure model, observation array, periods, numerical algorithm for the forward modeling, and mesh design. The proposed method was applied to three practical situations for marine and land MT arrays to show how the method works. All application showed that the coefficient of variations of synthesized MT responses are significantly different in site, frequency, and the elements of MT impedance tensor. The uncertainty of forward calculation should not be ignored in the future MT studies.	P-2) Evaluation on uncertainty in forward calculation of magnetotelluric response to three-dimensional electrical conductivity structure
May. 30		*Kiyoshi Baba(1) 1:ERI, UTokyo
May. 30	15:40 - 17:20　　　 Microtremor array observations were conducted along a north-south traverse of the Kochi plain, and the S-wave velocity structure of the subsurface were estimated using the array data. The depths to the engineering basement were estimated to be 20 to 30 m, and the layer from the surface to the top of the engineering basement were estimated to consist of a layer with S-wave velocity of 200 m/s for more than 10 m, and a very thin layer with velocity of 150 m/s.	P-3) Estimation of subsurface structure of S-wave velocity in urban Kochi
May. 30		*Nobuyuki Yamada・Sora Kadowaki(1) 1:Kochi Univ.
May. 30	15:40 - 17:20　　　 The Tatsutayama Fault is an active fault located in the northern part of Kumamoto City, but detailed geological and geophysical surveys have not been conducted in the southern part of the fault. We conducted microtremor observations around the Tatsutayama Fault. We confirmed that the peak period of microtremor H/V changes due to fault displacement of the Tatsutayama Fault. No clear change in the peak period was observed at the southern extension of the fault. Although further investigation is needed to estimate the geometry of the southern extension of the Tatsutayama Fault, it is a powerful tool for estimating active fault location.	P-4) Earthquake ground motion characteristics around the Tatsutayama Fault in Kumamoto.
May. 30		*Haruhiko Suzuki・Hisanori Matsuyama(1) 1:OYO Corp.


May.31 　　Room No.1（Meeting Room 03，2F）

Session	[Geothermal][Resource Exploration]
Chair	Junya Maruyama (HRO)

May. 31	9:30 - 9:50　　　 We carried out the geophysical survey in Kijiyama geothermal field using optical fiber in the KJ-5 geothermal well. We installed optical fiber to 2,000 m depth. The temperature was 288°C near the bottom. We obtained DAS and surface seismic records for twelve seismic excitations north of the wellhead. Vp depth gradient increases at 1,000 m depth showing the geological layer boundary. The Vp profile along the A-A’ line was obtained and is consistent with geological information by drilling. We obtained SH arrivals by DAS to 1,200 m depth using horizontal excitation. The VP/VSH at the shallower depth than 1,200 m was 1.75, which is usual in most rocks.	14) Geothermal exploration at Kijiyama geothermal field in Akita, Japan
May. 31		*Junzo Kasahara・Yoko Hasada(1),Hitoshi Mikada(2),Hiroshi Ohnuma(1),Yoshihiro Fujise(3) 1:ENAA,2:Kyoto Univ.,3:WELMA
May. 31	9:50 - 10:10　　　 A large number of seismic first motions are needed to analyze the subsurface structure. It is significant to use AI to detect earthquake data from a huge amount of continuous observation data. On islands, it is difficult to detect earthquakes by AI due to noise. Therefore, we applied EQTransformer (hereinafter referred to as "EQT"), which is considered to be usable on islands, to Kozushima island to examine the applicability of automatic seismic wave detection. To evaluate the applicability of EQT, we conducted manual seismic surveys during the four days when EQT detected many earthquakes, and calculated a confusion matrix with the manual seismic results as positive. As a result, we found the precision value is 91% and the recall value is 78%. Next, since we were able to evaluate precision value and recall value in Kozushima island, we then compared the performance of the AI in Mousavi et al. (2020) with that of EQT in this study. Both the precision and the recall were comparable. From these results, we can say that EQT can detect earthquakes even for noisy island data.	15) Automatic seismic detection of dense seismic observation data at Kozushima Island using EQTransformer
May. 31		〇Nonaka Shota・Azuma Hiroyuki・Oda Yoshiya(1) 1:Tokyo Metropolitan University
May. 31	10:10 - 10:30　　　 Distribution of hydrothermal altered layer (caprock) and fracture zones (faults) in geothermal area, related to upwelling of hydrothermal fluid from a geothermal reservoir, are often clearly imaged as conductors. Therefore, the resistivity structure can prove the relationship among the caprock, faults, and upwelling fluid. However, such detailed relationship has not been clarified due to low spatial density of MT observations and distortion of MT response (static Shift) due to near surface heterogeneous. In this study, we carried out high-density audio-frequency magnetotelluric (AMT) explorations at 83 sites with site separations of about 50 to 150 meters in a geothermal area, then estimated the detailed three-dimensional resistivity structure shallower than 1km. Here, we developed a new method to correct for the static shift, and applied it to AMT data. We found that a distinct variation in resistivity on both sides of a known active fault. This resistivity contrast is suggested to be due to the sharp variation of hydrothermal circulation around a fault. The spatial heterogeneities of low resistive zone, possibly corresponding to the caprock, would be due to the non-uniform upward paths of hydrothermal fluid along a fault.	16) Estimation of shallow resistivity structure in a geothermal area using a novel Static Shift correction method and a scheme to stabilize three-dimensional magnetotelluric inversion.
May. 31		〇Nagi Yamashita・Tada-nori Goto(1),Kodo Umakoshi・Hiroshi Sasaki(2) 1:University of Hyogo,2:Nagasaki Univ.
May. 31	10:30 - 10:50　　　 Induction coils used in measuring magnetic fields of conventional MT survey systems are known to be heavy and theoretically less sensitive at lower frequencies. Based on years of research to solve these problems, three types of compact and lightweight MT survey systems using MI sensors instead of induction coils are being developed. Commercial MI sensors intentionally have low sensitivity below 0.1 Hz to reduce the influence of the Earth's strong magnetic field. In this study, commercial MI sensors were improved and evaluated to measure magnetic field variations in the frequency band lower than 0.1 Hz, assuming an exploration depth of several kilometers below the ground surface. Based on the evaluation of the improved MI sensor, it was confirmed that it can effectively measure magnetic field variations with sufficient sensitivity in the very low frequency range.	17) Improvement of MI sensor for measuring very-low-frequency magnetic fields
May. 31		*Toshiaki Tanaka・Hideki Mizunaga・koji Hashimoto・Hiroyuki Furukawa(1),Shinya Onodera(2) 1:Kyushu University,2:JOGMEC
May. 31	10:50 - 11:10　　　 In recent years, research and development of geophysical exploration technology using drones (small UAVs, uncrewed aerial vehicles) have progressed. In the electromagnetic (EM) survey method using drones, the transmitter is placed on the ground surface or towed by a drone, and its response characteristics vary widely. On the other hand, few studies have investigated and evaluated their response characteristics by numerical calculations. In this study, we have calculated the responses of various drone survey methods by our forward modeling code and carefully examined the responses and depth of investigation depending on the transmitter/receiver configuration. Such calculations and investigations will enable us to study the response characteristics and optimal transmitter/receiver configuration for drone EM surveys, leading to more effective uncrewed exploration.	18) Determination of Optimal Transmitter and Receiver Configuration for Time-Domain Airborne Electromagnetic Survey Using UAVs
May. 31		*Yasuyuki Kuwai・Takumi Ueda(1) 1:Waseda Univ.
May. 31	11:10 - 11:30　　　 While gas hydrates (GH) are expected to be used as an energy resource, the greenhouse effect of methane generated from GH is also a concern. It is known that lumpy GH (shallow-type GH) exists near the seafloor surface in the eastern margin of the Japan Sea, and its impact on the global environment has been pointed out. However, the spatial extent of shallow-type GH in the horizontal and vertical directions is not known. We focused on marine direct current resistivity (MDCR) survey that can investigate the resistivity structure under the seafloor. By analyzing MDCR data obtained in the eastern margin of the Japan Sea, we estimated the resistivity structure. It was found that in areas where GH and gas plumes have been identified on the seafloor surface, high resistivity anomalies (>10 Ωm) with a thickness of about 30 m are localized just below the seafloor. The GH saturation rate was estimated using Archie's equation, and it was found that this high resistivity anomaly area shows high GH saturation rates ranging from 60% to 95%.	19) Imaging of shallow gas hydrate in the eastern margin of Japan Sea by marine direct current resistivity survey
May. 31		Ayako Oda ・*Keiichi Ishizu ・Tada-nori Goto(1),Takafumi Kasaya(2) 1:Univ of Hyogo,2:JAMSTEC


Session	Executive Sessions
Chair	Jun Matsushima (The Univ. of Tokyo, GSFS)

May.31	15:00 - 16:00	Satoru Nakazawa (JAXA)
May.31	15:00 - 16:00
May.31	16:00 - 17:00	Norihiko Saeki (METI)
May.31	16:00 - 17:00


May.31 　　Room No.2（Meeting Room 05，2F）

Session	[Earthquake 2][Disaster Prevention 2][Volcano]
Chair	Nobuyuki Yamada (Kochi Univ.)

May. 31	9:30 - 9:50　　　 Additional microtremor array observations using a small-size array were conducted in the northwestern part of Morioka City where the exploration depth was insufficient in previous our studies. From estimated S-wave velocity structure model, 30m averaged S-wave velocities (AVSs) were obtained, and the ground amplification factors from the engineering bedrock were calculated using estimated AVSs. As a result, the AVSs were ranging from 133 m/s to 384 m/s. These were slower than J-SHIS's ones. The amplification factors obtained from AVS were ranging from1.18 to 2.55. The amplification factors observed at 42 of the 48 observation points was larger. became. The result implies that the amplification factor of J-SHIS was underestimated and dense microtremor array observations were important.	25) Re-evaluation of site amplification factors in northwest Morioka area by microtremor array surveys
May. 31		*Hidekazu Yamamoto・Tsubasa Sato・Tsuyoshi Saito(1) 1:Iwate University
May. 31	9:50 - 10:10　　　 We conducted repetitive high-resolution surface wave survey in Hinode District, Itako City, where had been attacked by the 2011 East Japan Earthquake and was severely damaged by liquefaction. The purpose of the survey was to discriminate spatial and temporal changes in S-wave velocity accompanied with the artificial lowering of water table. A total of 7 short lines were surveyed from September 2016 to December 2017 set in and around the district. We adopted the “walk-along” surface wave survey method which was proposed by the first author targeting near surface shallower than 10 m. CMP-SPAC analysis was employed to calculate high-resolution dispersion curve of surface wave phase velocity. Although obvious variations of dispersion curves were not recognized, we could distinguish slight changes of dispersion curves at specific frequency zones which were correlated with the sensitive layers for the change in water table and rainfall. This indicated that high-resolution “walk-along” surface wave survey is capable of detecting local or temporal changes in S-wave velocities of the near surface caused by changes in water table or in water content.	26) Repetitive multichannel surface wave surveys at Hinode District, Itako City.
May. 31		*INAZAKI Tomio(1),YOKOTA Toshiyuki(2) 1:Independent,2:AIST
May. 31	10:10 - 10:30　　　 We developed a compact ambient noise measurement system that consists of a node-type seismograph with a mobile cell phone sim card for transferring the data to a cloud database every minute. The stored data in the cloud are analyzed by the SPAC method, and the phase velocity is automatically calculated every hour. We have carried out a field test using four to ten receivers for about one year. A notable difference in the ambient vibration level between daytime and nighttime was observed, and it is due to the surrounding construction activities near the test field. The calculated phase velocity using four receiver records showed slight fluctuation, but a 24-hour moving average was stable during the field test period of about one year. Several large earthquakes were recorded during the observation period, and the waveform was comparable to the F-net broad-band seismograph waveform. These results demonstrate that the developed real-time monitoring system can monitor the ambient vibration level and the phase velocity changes associated with natural phenomena or construction activities.	27) Development of a compact real-time seismograph system
May. 31		*Osamu Ishitsuka・Chisato Konishi・Koichi Hayashi・Haruhiko Suzuki・Yoshiaki Tanazawa(1) 1:OYO Corporation
May. 31	10:30 - 10:50　　　 The S-wave velocity structure model must be appropriately constructed for the ground response analysis that removes the influence of surface ground. We generally apply a one-dimensional (1D) analysis for that. However, the 1D analysis is not always applicable to all stations due to the irregularity of S-wave velocities. In this study, we developed a method to construct the model with a two-dimensional (2D) analysis and applied the method to KiK-net Kamaishi. It was shown that the reproducibility of observation records, which was insufficient in the conventional 1D analysis, could be improved through the evaluation of the 2D model.	28) Evaluation of two-dimensional S-wave velocity structure model with borehole array records
May. 31		*Kentaro Kasamatsu(1),Takashi Nozawa(2),Tetsushi Watanabe(3) 1:Kajima Corporation Technical Research Institute,2:Kajima Corporation Nuclear Power Department,3:Kobori Research Complex Inc.
May. 31	10:50 - 11:10　　　 This paper estimates P-wave attenuation structure of Kozushima Island from the twofold spectral ratio method. The twofold spectral ratio method removes the source and site amplification characteristics by using multiple earthquake records observed at multiple stations, and extracts only path characteristics, i.e., attenuation characteristics. Earthquake data observed by a dense seismic network of 31 stations in Kozushima Island during the period from October 2020 to March 2021 were used. Twofold spectral ratios were calculated for a total of 23 station pairs data, and the Q structure was estimated by BPT as an initial model for the inverse analysis and SIRT for iterative correction reducing the error. As a result, beneath the center of Tenjosan volcano tends to have higher Q-values. In addition, the Q-values in the central part of the island (southwest of Tenjosan) were relatively high, corresponding to the high velocity area estimated from the natural earthquake tomography.	29) P-wave Attenuation Structure of Kozushima Island from twofold spectral ratio method using dense seismic observation data
May. 31		*Haruna Inoue・Utako Watanabe・Hiroyuki Azuma・Yoshiya Oda(1) 1:Tokyo Metropolitan Univ.


Jun.1 　　Room No.1（Meeting Room 03，2F）

Session	[Buried Objects][Slopes][Others]
Chair	Motoyuki Sato (Tohoku University)

Jun. 1	10:00 - 10:20　　　 A creep movement with about 3 cm/year has been observed in a landslide site in Tokushima Prefecture. The electrical resistivity tomography was repeatedly carried out there (in 2015 and 2022) along the same survey lines to detect subsurface structural changes associated with the creep. As a result, the change of the resistivity structure was not large for about these 7 years; however, an increasing tendency of the resistivity was seen in a part of the presumed slip surface. In addition, the change in resistivity tended to be large on the slope where the slip surface seems not to be developed. Based on the above, we discuss the possibility of future landslides in this area.	20) Temporal change of subsurface resistivity distribution in a landslide area based on time-lapse electrical resistivity tomography
Jun. 1		Sena Fujimoto・*Tada-nori Goto・Keiichi Ishizu・Miku Sakurai・Soma Yamamoto(1),Tomomi Terajima(2) 1:Univ. Hyogo,2:Kyoto Univ.
Jun. 1	10:20 - 10:40　　　 We investigated the formulation of a hyperbolic reflection image that was detected when the underground penetrating radar was crossing underground pipes. When the underground pipe is parallel to the ground, the reflection point at the top of the hyperbolic reflection image is located directly below the radar. When it isn’t parallel, that reflection point isn’t directly located below the radar and it is difficult to infer the position of underground pipes. It is necessary to formulate the reflection image in order to infer the buried pipe position from the data of hyperbolic reflection image. In this paper, the parameters to identify the position of underground pipes were defined and assumed reflection images were formulated based on the measurement principle. As a result, it is confirmed that the exploration position detecting the top of the hyperbolic reflection image is not equal to the point dicrectly above the underground pipe when the pipe isn’t parallel to the ground. The detail of parameters and formulation result is described.	21) Formulation of hyperbolic image on underground pipe detection by ground penetrating radar
Jun. 1		*Soraya Shizumi・Atsushi Isobe(1) 1:Hitachi, Ltd.
Jun. 1	10:40 - 11:00　　　 The transient electromagnetic method using long grounded-wire (GW) sources as a transmitter can be used at a depth of investigation up to several kilometers and can be applied in a wide range of targets such as mineral resources, geothermal reservoirs, and natural gas resources deep underground. In cal-culating the response of GW signal sources, an approximation method using a horizontal electric dipole (HED) signal with equivalent moments can also be adopted under specific conditions. Still, not much research has been done on the characteristics and distribution of the response between the two signal sources. Therefore, in this study, transient responses (secondary magnetic fields) generated by HED and GW sources are calculated by our forward modeling code, and their distribution and characteristics are carefully examined and evaluated. Then, we aim to understand the conditions under which the response of the GW source can be approximated using the response of the HED source and the differences in the response characteristics of the two sources	22) Calculation and Characterisation of Electromagnetic Fields Generated
Jun. 1		*Haoran LIU (1),Takumi Ueda(2) 1:Graduate school, Waseda Univ.,2:Waseda Univ.
Jun. 1	11:00 - 11:20　　　 Deep Learning (DL) is a multilayer neural network (NN), and when used as a regression problem, it corresponds, for example, to inverse　analysis in electromagnetic exploration. We have constructed new NN models, particularly Deep Neural Network (DNN), Convolutional Neural Network (CNN), and Recurrent Neural Network (RNN), to invert survey data by airborne time domain electromagnetic (ATEM) data. The training data for DL network construction are consisted of three sets of data with different numbers of data. For each data, several validation models were constructed and trained. Each layer was modified to find effective hyperparameters for the different NN models. We have constructed new NN models from these validations, including DNN, CNN, and RNN, and carefully compared and examined their learning accuracy. Then we applied our NN models to actual field data observed using AEM (Helitem system) by JOGMEC. Results show that our three types of NN models can estimate subsurface geo-electrical structures similar to standard deterministic inversion results. Future research includes a detailed examination of training data, hyperparameters, and consideration of data noise.	23) Inversion of time-domain airborne electromagnetic data using deep learning
Jun. 1		Shu Kitsukawa(1),Kengo Nakanishi・Tamaki Nishino(2),*Takumi Ueda(1) 1:Waseda Univ.,2:Graduate school, Waseda Univ.
Jun. 1	11:20 - 11:40　　　 Due to the rapid development of drone technology, drones have come to be used in a wide range of fields, such as topographical surveying, pesticide spraying, and bridge inspection. In this study, we utilize UGVEM system, combining multi-coil type electromagnetic equipment (CMD-EXPLORER) and uncrewed ground vehicle (UGV) for shallow subsurface exploration such as river levee surveys. Based on the measurement data obtained by our UGVEM system, inversion has been performed using two inversion codes (1) commercial-based software and (2) newly developed in this study with Python. The results of the resistivity structure estimation by both inversion codes are harmonic, showing that the UGVEM exploration can be applied effectively even for the long survey line data.	24) Numerical methods for unmanned ground vehicle electromagnetic survey system and application to river levee exploration
Jun. 1		Yukei Shirasawa(1),Junya Shinohara(2),*Takumi Ueda(1),Ryousuke Umezawa・Motoharu Jinguji・Toshiyuki Yokota(3) 1:Waseda Univ.,2:Graduate School, Waseda Univ.,3:GSJ, AIST


Jun.1 　　Room No.2（Meeting Room 05，2F）

Session	[CO2][Methane hydrate][Oil and Natural Gas]
Chair	Yamauchi Yasutomo (DIA Consultans Co., Ltd.)

Jun. 1	10:00 - 10:20　　　 Recent climate changes have led to a growing international movement to regulate carbon dioxide emissions. However, in the present, when fossil fuels are the primary energy source, it is difficult to reduce carbon dioxide emissions remarkably. Therefore, CCS, a technology to capture and store carbon dioxide into deep underground, has been attracting attention. On CCS, monitoring subsurface carbon dioxide behavior is required as well as drilling and injection, and seismic tomography is considered one of the essential monitoring methods. In this study, we have developed forward and inversion programs of seismic tomography. For coding, we have adopted Julia, a new scripting language, as a programing language that aims for both ease of writing and calculation speed and can also visualize numerical results. Furthermore, we have analyzed the pseudo-model of CCS using developed codes and evaluated the applicability of seismic tomography to CCS monitoring.	30) Development of numerical methods for cross-well travel time tomography by Julia and its application to CCS monitoring
Jun. 1		*Kento Kamimura(1),Takumi Ueda(2),Masashi Kodama・Toshiyuki Yokota(3) 1:Graduate School, Waseda Univ.,2:Waseda Univ.,3:GSJ, AIST
Jun. 1	10:20 - 10:40　　　 To verify the applicability and performance of the developed joint inversion, our previous study proposed a laboratory scale measurement by a combination of muons and seismic waves, and then outlined the conceptual design of combined laboratory measurements of muons and seismic waves. In this paper, we demonstrate preliminary results of laboratory measurements of cosmic ray muons by investigating the water depth by using muon counts. Our repeated laboratory measurements shows that average muon counts macroscopically correlate with water depth. By further conducting atmospheric and temperature correction, we can improve the results.	31) Combined measurement of cosmic ray muons and elastic waves on a laboratory scale: a follow-up report
Jun. 1		*Jun Matsushima(1),Masashi Kodama(2),Mohammed Y. Ali・Fateh Bouchaala(3),Hiroyuki K.M. Tanaka(4),Tadahiro Kin(5),Toshiyuki Yokota(2),Makoto Suzuki(6) 1:The Univ. of Tokyo, GSFS,2:AIST,3:Khalifa University,4:The Univ. of Tokyo, ERI,5:Kyushu Univ., IGSES,6:The Univ. of Tokyo, SoE
Jun. 1	10:40 - 11:00　　　 Machine learning has made remarkable progress in recent years and has demonstrated high performance in various fields. It has also been successfully applied in geophysical exploration. In this paper, we examine the applicability of a method called Physics-Informed Neural Networks (PINN), which has recently attracted much attention. PINN incorporates constraints based on knowledge of physical laws into the loss function of a neural network. By incorporating these constraints, the neural network undergoes a learning process and is eventually optimized to satisfy the given physical laws. A common application example is the problem of solving partial differential equations. This paper addresses the problem of simulating one-dimensional diffusion and wave equations. However, the PINN method is still in the research stage and has many problems to be solved. It is not certain whether it is useful for large-scale problems. In addition, the design of the loss function and its integration into the network is often difficult, and the need for hyperparameter tuning also makes applying PINN difficult. It is expected that PINN will be applied in a wide range of fields in the future.	32) Applicability of Physics-Informed Neural Networks (PINN) to geophysical exploration
Jun. 1		*Matsuoka Toshifumi・Iso Shinichiro(1) 1:Fukada Geological Institute
Jun. 1	11:00 - 11:20　　　 A knoll in the Mogami Trough on the eastern margin of the Sea of Japan, called the Sakata Knoll, is known as a field of shallow methane hydrate (MH). There are still many unknowns regarding the occurrences of the shallow MH. For better understanding of the occurrences of shallow MH, we conducted multiple types of geophysical survey in the same area. Although it is difficult to directly compare the results of these data due to differences in target, possible depth of exploration, resolution, and depth index, an integrated interpretation is possible with enough attention of the characteristics of each geophysical observation: it can be possible to discuss the rough distribution of shallow MH and related materials based on multiple physical properties obtained from multiple types of geophysical surveys. As a result, part of the subseafloor structures can be estimated as follows; (1) horizontally continuous sedimentary layers and also horizontally continuous shallow MHs are developed above the BSR, (2) MHs are distributed over the limited area narrower than the BSR displayed, (3) bubbles widely exist below the BSR, and (4) fluid containing bubbles exists along faults and may lower the resistivity.	33) Challenge to predict the subseafloor shallow methane hydrate distribution, using LWD, acoustic, seismic, and CSEM explorations
Jun. 1		*Miho Asada・Shogo Komori・Toshiyuki Yokota(1) 1:AIST
Jun. 1	11:20 - 11:40　　　 Amplitude Versus Offset (AVO) analysis is the method to detect hydrocarbons within reservoirs by using amplitude variation of reflection events. When the reservoir getting thinner, interferences between top and bottom of the reservoir should be taken into account. The conventional procedure to examine such the effect is convolving a source wavelet with reflection coefficients calculated using the Zoeppritz equation or its approximations. This procedure, however, does not consider horizontal interferences which occur in actual field data acquisition. We investigated this AVO behavior by using a shot gather generated by finite difference method (FDM) which can provide a more realistic simulation compared to the convolution model. We compared both results using a cross-plot of Intercept and Gradient values. When the reservoir getting thinner, the FDM gives complex trajectories, while the convolution model gives simple straight lines. We conclude that we should use the FDM method to evaluate more realistic AVO responses for thin layers as it considers the effect of horizontal interferences.	34) Synthetic study of thin layer effects on AVO
Jun. 1		*Shuhei Nishijima(1),Takao Nibe(2),Akihisa Takahashi(3) 1:JGI,2:JAPEX,3:Wavelet
Jun. 1	11:40 - 12:00　　　 MDRS processing adequately accommodates the so-called conflicting dip problem inherent in CRS methods, while keeping their striking performance to enhance S/N in noisy seismic data without degrading lateral resolution. Prestack MDRS, recently extended to the 3D application, enables to produce prestack seismic gathers of the MDRS quality, which allows to perform more accurate velocity analysis and subsequent prestack imaging such as PSTM/PSDM. An example from a land 3D seismic survey demonstrates that a significantly improved PSTM result is delivered through improved prestack gathers, uniform offset-azimuth distributions, and better-defined velocity information, all resulted from 3D Prestack MDRS processing.	35) Quality improvement in land seismic data through prestack MDRS
Jun. 1		*Shigeru Ino・Kenichi Akama(1),Fuminori Honda(2) 1:JGI, Inc.,2:JAPEX

VENUE
	BLDG.63, 2F,NISHIWASEDA CAMPUS, 3-4-1 Okubo, Shinjuku-ku, TOKYO, JAPAN and ONLINE

SCHEDULE
	Tuesday, May 30, 2023
		Openning Speech	Room No.1 (Meeting Room 03, 2F)	09:20-09:25
		__Speaker:	Keiichi SUZUKI (Kawasaki Geological Engineering Co. Ltd.; President of SEGJ)
		Technical Sessions	Room No.1 (Meeting Room 03, 2F)	09:30-15:20
		Poster Core Time	Room No.1 (Meeting Room 03, 2F)	15:40-17:20
		Exhibition	Room No.3 (Meeting Room 04, 2F)	11:00-17:20

	Wednesday, May 31, 2023
		Technical Sessions	Room No.1 (Meeting Room 03, 2F)	09:30-11:30
		Technical Sessions	Room No.2 (Meeting Room 05, 2F)	09:30-11:10
		General Meeting	Room No.1 (Meeting Room 03, 2F)	13:00-14:40
		Executive Sessions	Room No.1 (Meeting Room 03, 2F)	15:00-17:00
		__Chair:	Jun MATSUSHIMA (The University of Tokyo; Vice-President of SEGJ)
		__Presentation(1):	Asteroid Explorer "HAYABUSA 2", Challenge, Return, and Challenge Satoru NAKAZAWA (JAXA)
		__Presentation(2):	The CCS policy of Japan Norihiko SAEKI (METI)
		Exhibition	Room No.3 (Meeting Room 04, 2F)	09:30-17:00
	Thursday, June 1, 2023
		Technical Sessions	Room No.1 (Meeting Room 03, 2F)	10:00-11:40
		Technical Sessions	Room No.2 (Meeting Room 05, 2F)	10:00-12:00
		Exhibition	Room No.3 (Meeting Room 04, 2F)	10:00-11:00

Registered and Paid		Pre-Registration (Until April 27,2023)	Onsite Registration
Conference Meeting	Member / Non-member	JPY 7,700-*	JPY 8,800-**
	Student (Member)	JPY 0,000-*	JPY 0,000-**
	Student (Non-Member)	JPY 3,850-*	JPY 4,400-**

**	Pre-registrant will receive PAPER Proceedings at the registration desk of the conference.
**	PAPER Proceedings can be purchased at a price of JPY 1,650-.

LIST OF EXHIBITORS
	ACT Electronics Corp. Geofive Co.,Ltd. Hakusan Corporation Superconducting Sensing Technology Corporation(SUSTEC)

LIST OF ADVERTISEMENTS
	Geophysical Surveying Co., Ltd.(GSC) 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).