Since 1997, we, the Japan Agency for Marine-Earth Science and Technology (JAMSTEC), have been conducting marine seismic survey around Japan to reveal the detailed seismic structure of the plate subduction zone to understand the mechanisms of the earthquake occurrence. The target depth is about 10 km and more, and it is significantly deeper than that of the seismic exploration of resources, such as oil and gas. Therefore, our marine seismic surveys have been characterized by the use of wide-angle seismic reflection and refraction surveys using ocean bottom seismometers (OBSs) in addition to the multi-channel seismic reflection survey using a hydrophone streamer cable.

　Traditionally, refraction data has been processed using travel-time to construct seismic velocity structure models. However, because of the lower spatial resolution of travel-time analysis compared to that of the seismic reflection profiles, conventional structural studies have been limited to qualitative comparisons and discussions of the resultant models and seismic reflection image side by side.

　Recently, full-waveform inversion (FWI) of refraction data has been put to practical use, and the spatial resolution of seismic velocity structure models obtained from refraction data has dramatically improved, making it possible to interpret the results from the refraction data more closely related to seismic reflection profiles.

　The spatial resolution of the FWI results improves with denser OBS spacing, but denser OBS spacing requires more resources, including ship time, number of OBS, and budget. To investigate an optimal OBS spacing that balances the spatial resolution sufficient to discuss the key geological structures of the plate subduction zone and the resources required, we conducted a thinning out test (a decimation test) in which we applied FWI to an actual refraction data obtained in the Nankai Trough with decreasing numbers of OBSs (i.e., with increasing the OBS spacing). From this test, we conclude that the cost-effective OBS spacing for our current study objective is 2 km although denser spacing such as 1 km or less would certainly provide higher spatial resolution.

　We also present the results of combined processing of refraction data and reflection data in Nankai Trough that shows the refraction data is effective for estimating appropriate seismic velocities at depth that cannot be constrained by the reflection data alone. The results again demonstrate the effectiveness of combining FWI of refraction data with seismic reflection imaging in the study of the seismogenic plate subduction zone.

　In the oil and natural gas industry, the reduction of greenhouse gas emissions in the exploration and development (E&P) is an important perspective to achieve both the social demand for carbon neutrality and stable resource supply. Further business development of aquifer CCS and CCUS with enhanced recovery will be inevitable in the next 10 years. In addition, it is also crucial to improve the efficiency of E&P costs through the adoption of digital technology, and to monetize E&P projects at an early stage. Under these circumstances, the construction of a detailed subsurface digital twin that incorporates structure, lithology, physical properties, hydraulics, and mechanical properties is essential for improving the accuracy of understanding the time history and predicting the behavior of subsurface fluid flow, and there is a need to accelerate the development of various technological elements related to the acquisition, processing, and analysis of seismic exploration data. In this paper, we will discuss the current status and future prospects of elemental technology, focusing on the optimization of spatial sampling and the expansion of broadband seismic exploration, mainly for seismic data acquisition, while looking at the complementary development of data acquisition and processing that has resulted from the advancement of innovative technologies such as Full Waveform Inversion (FWI). In addition, we will discuss recent advances in sensing technologies such as fiber optics and IoT technologies that enable automated data acquisition from the perspective of efficient seismic data acquisition and integration, as well as approaches to seismic exploration for reducing environmental impact and promoting environmental protection.

　According to estimates by the Cabinet Office, the domestic market for the space industry is expected to reach approximately 4 trillion yen in 2024. The World Economic Forum also predicts that the international market for space business will reach approximately 260 trillion yen in 2035, about three times the current market size. The space industry can be divided into two main categories. The first is the space equipment industry, which manufactures satellites and launch vehicles, and the second is the space applications industry, which uses satellites to transmit and receive data. The main areas of the space application industry include telecommunications broadcasting, positioning and remote sensing, with consumer-oriented services accounting for by far the largest proportion. As a result, attention is focusing on the creation of new related markets using technologies such as big data and artificial intelligence.

　This paper focuses on remote sensing technology, which has been utilized along with the recent development of space technology, with the aim of understanding its measurement principles and introducing examples of domestic and international technology utilization and research and development, mainly in the area of terrestrial remote sensing.

　We developed a method that spatially interpolates one-dimensional (1D) S-wave velocity (Vs) profiles to a 3D Vs model based on deep neural network to predict regional shallow three-dimensional (3D) subsurface models. The method uses dispersion curves obtained by active and passive surface wave measurements, and horizontal-to-vertical spectral ratio (H/V) obtained by single station three-component microtremor measurements. Since the number of sites with dispersion curves is smaller than those with H/V measurements, the deep neural network consists of two stages. The first stage (A) predicts Vs profiles from H/V using training data of Vs profiles obtained from dispersion curves, and the second stage (B) predicts Vs profiles from surface topography and geomorphological classification etc. using training data of Vs profiles obtained from the first stage. Estimation procedure for a 3D regional shallow Vs model can be summarized as four steps and two stage deep learnings as follows. At the first step, we estimated 1D Vs profiles by the inversion of dispersion curves at sites where both dispersion curve and H/V were observed. At the second step, the first stage deep learning (A) predicts 1D Vs profiles from H/V spectra based on the first stage network trained by H/V spectrum-1D velocity profiles together with other regional information including coordinate, surface elevation, geomorphology, and bedrock depths in community velocity model. The first stage (A) training predicted 1D Vs profiles from H/V spectra measured at sites without surface wave methods. At the third step, we used 1D Vs profiles predicted in the second step as initial profiles, and applied non-linear inversion using H/V to finalize 1D Vs profiles. At the last step, the second stage deep learning (B) predicts 1D Vs profiles in the investigation area from geological and other regional information, based on the second stage network trained by the geological information-1D Vs profile pairs. We applied the proposed method to the Eastern part of Tokyo Metropolitan area to Southeastern part of Saitama prefecture, using dispersion curves, H/V and Vs profiles open to public as digital data, and predicted Vs profiles to 90 m deep with 200 m grid intervals. The predicted Vs model was reasonably consistent with surface topography, surface geology and geomorphological classification.

　Fiber optic sensing technology is attracting attention due to its high sensor density, long-term installation capabilities, and low-cost effectiveness. Its application to CO₂ underground storage projects is being explored. Distributed Acoustic Sensing (DAS) is effective for detecting changes in subsurface velocity caused by injected CO₂, which is crucial for monitoring CO₂ migration and leakage in reservoirs and seal formations.

　In this study, we conducted technical verification of DAS measurement using an optical fiber installed on the ground surface (Surface DAS; S-DAS). An experiment was designed to detect velocity changes caused by gas injection in a test field through S-DAS measurement. Specifically, 7,000 liters of nitrogen gas were injected to a depth of approximately 5 meters below the water table over about 2 hours. Continuous observation of seismic records generated by a small electrical hammer was carried out for about 8 hours.

　The experiment was repeated twice with different offsets between the gas injection well and the shot point to examine the impact of offset variation on the recorded data. The observed waveforms changed over time, likely due to the underground spread of the injected gas. This result was consistent with numerical simulations conducted prior to the experiment. The changes in the records can be interpreted as P-S conversion reflection waves, suggesting the importance of considering P-S reflection waves in S-DAS monitoring plans.

　This study demonstrates that S-DAS measurement is a promising method for CO₂ monitoring. To further improve the effectiveness of S-DAS monitoring, it is essential to develop adequate noise reduction techniques and imaging processes tailored to S-DAS measurements, along with ensuring the repeatability of the shot waveform to mitigate general 4D noise.

　The accurate picking of seismic phase arrivals is crucial for hypocenter determination and subsurface imaging. However, the growing demand for seismic monitoring in CCS projects and the increasing adoption of DAS for high-density seismic observations have revealed the limitations of manual picking methods. This study aims to achieve accurate first arrival picking under limited training data conditions by evaluating the effectiveness of data augmentation, self-training, and fine-tuning techniques applied to deep learning models (PhaseNet and EQTransformer). Experiments using Hi-net data demonstrate that increasing the amount of training data generally improves accuracy. However, even when the training data was limited to one month, combining the proposed methods enabled the models to achieve F1 scores comparable to those obtained with one year of training. Among the methods, fine-tuning using pretrained models on the large-scale STEAD dataset yielded the most significant improvement, with additional gains from self-training and data augmentation. Furthermore, we validated the models using DAS data, which were acquired under observational conditions distinct from those of the seismic records used for pretraining. The results demonstrated that EQTransformer exhibited high adaptability through fine-tuning, whereas PhaseNet showed no improvement. This difference is attributable to the architectural characteristics of each model, highlighting the importance of selecting appropriate model architectures and training strategies based on the characteristics of the target data and intended application. The findings of this study provide guidelines for effectively leveraging deep learning under limited data conditions and are expected to contribute to the advancement of seismic monitoring and disaster mitigation efforts.

　This technical paper was contributed to the 75th Anniversary Symposium of the Geophysical Exploration Society of Japan and presents the latest in geophysical exploration in the resources session (oil, natural gas, metals, geothermal). JOGMEC has developed the DAS-VSP technology as part of its geothermal power generation technology research and development project. As a technique for exploring fracture systems in an area of about 500 m around a well, we have developed a new exploration method using diffraction imaging, which utilizes the phenomenon in which the opening part becomes a prominent secondary source when seismic waves pass through the fracture and have achieved field implementation. In addition, DAS-VSP can be applied not only in the active type using artificial seismic sources, but also in the passive type using AE in the area as the seismic source, and the applicability of each method may differ depending on the field, so we have classified them according to the surface conditions of geothermal areas, microseismic activity, and the number of wells that can be used for observation, and have proposed a survey method suitable for each type.

　The purposes of this study are to elucidate the resistivity characteristics of rocks from the volcanic area by laboratory resistivity experiments using drilled core samples, and to estimate the depth distribution of porosity at the same borehole by integrating the core experiment results with log data. The target borehole is located in the Aso volcanic area in Kumamoto. In the laboratory resistivity experiments, the effects of porosity, saturation, and pore water resistivity on rock resistivity were quantitatively evaluated. The effect of each physical property on rock resistivity by this study were generally agreement with the previous study, but the m value, called the cementation factor, ranged from 2.31 to 3.22 for sedimentary rocks, which is larger than the commonly used value of 2, and smaller values (less than 2.0) with smaller porosity for volcanic rocks. In addition, the ratio of apparent formation factor to the true formation factor was compared with the pore water resistivity, indicating that at least for pore water resistivity larger than approximately 30 Ωm, the effect of surface conductivity should be taken into account even for rocks from volcanic area.

　Based on the high drilling mud resistivity (52.1 Ωm at 15.8 °C), which was inferred as the pore water resistivity in borehole wall rocks, and the experimental results in this study, we estimated the depth distribution of the porosity of the borehole by incorporating the values of surface conductivity and m of the core samples into the log data. The results of this study are consistent with the results of the estimated porosity by sonic log in the depth range of 302-562 m, indicating that porosity can be estimated from electrical log even in areas where rocks from volcanic regions are distributed and pore water resistivity is high. The estimated porosity using drilling mud resistivity of this study was larger than the core porosity and the estimated porosity by sonic log in the depth range of 562-653 m. The reason may be that the drilling mud resistivity is higher than the pore water resistivity, because penetration of drilling mud into the surrounding formation was insufficient in this depth range.

　For the purpose of site investigations of floating offshore wind farms, which will potentially 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 50 m 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 the 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 (linear array) using an undersea positioning device and conducted measurement experiments. A clear dispersion curves were obtained from the acquired data. In addition, a surface wave exploration measurement experiment was also conducted in which vibration was generated by dropping a weight on the extension of the end of the linear array to actively generate surface waves. It was revealed that good data could be obtained using this method, and that in addition to the S-wave velocity structure obtained by the microtremor array survey, the velocity structure in shallower layers could also be obtained with high accuracy.

　To promptly issue seismic intensity information and earthquake warnings immediately after an earthquake occurs, it is necessary to construct and maintain an earthquake observation network and a telecommunication network, and to install seismographs that can retrieve high-quality data in an appropriate environment. In the railway field, to improve the reliability of early earthquake warning (EEW) systems, EMC standards are applied to railway seismographs that measure and record seismic motions, mainly installed in substations to reduce the influence of electromagnetic waves. However, as for seismographs that measure and record seismic motions, an investigation that the seismic motions recorded satisfy certain standards, such as the seismic intensity inspection by the Japan Meteorological Agency, has not been performed. Thus, the reliability of seismographs installed in railways has not been verified. Therefore, in this study, to comprehend the reliability of earthquake warnings in the railway field, we investigated seismic data quality affecting earthquake warnings applied to railways through temporary earthquake ground motion observation with multiple types of seismographs at the Kasama station (KSM) in the Metropolitan area maintained by the Railway Technical Research Institute.

　First, we investigated the self-noise of seismographs by the calibration test for microtremors. We showed that the seismographs need to measure the power spectral density of microtremors of about －120 dB in the frequency range of more than 0.12 Hz, to properly record microtremors at KSM on the rock site. In addition, the average power of microtremors recorded by all seismographs was less than 10^-4 cm²/s⁴, indicating no particular problems with data accuracy when measuring vibrations with an amplitude of 0.01 cm/s² or more. Second, we investigated seismic data quality affecting two types of earthquake warnings applied to railways (the threshold method and the C－Δ method). As for the threshold method, the variability can be seen in the case of values of 10.1 to 10.7 cm/s² exceeding 10 cm/s² for all seismographs (average: 10.4 cm/s², standard deviation: 0.2 cm/s²). It is revealed that the timing of exceeding the threshold varies by a few tenths of a second among the seismographs in this observation data. And, the residual average ± the standard deviation with the unified earthquake catalog of Japan Meteorological Agency (JMA) of the epicentral distance estimation by the C－Δ method were within the range of 0.05 to 0.6 km (Common logarithm), and the impact of this observation data affecting the C－Δ method to estimate the epicentral distance was not very large. On the other hand, the residual average ± the standard deviation with the unified earthquake catalog of JMA of the epicentral direction estimation by the C－Δ method were within the range of －90° to 90°, and the impact affecting the C－Δ method to estimate the epicentral direction was relatively large compared to the epicentral distance estimation.

　Seismic observation was conducted on Hachijojima Island, the second largest volcanic island among the Izu Islands, Japan. This temporary observation took place twice, with each session occurring within a 7-months period in 2019 and 2021. We have densely installed 46 temporary stations across the island, in addition to 9 permanent stations, to collect seismic data. In this paper, our primary focus is to describe the seismic observation and establish an earthquake catalog, with initial evaluation of seismicity in relation to the magmatic system and tectonic setting. The arrival times of P- and S- wave were manually picked using WIN System, a processing system that handles multi-channel seismic waveform data. Subsequently, the hypocenters of the local earthquakes were located directly during manual pick using the ‘hypomh’ program within WIN System. A total of 179 local earthquakes were localized where 119 took place in 2019 and the remaining 60 occurred in 2021. It was observed that most earthquake events' hypocenters were situated primarily in the northwestern region of the island, located at the northern edge of the rift margin. These earthquake events appear to be associated with the long-distance lateral magma transport in the middle to lower crust at depths of 10-20 km, which is driven by the regional tectonic conditions within the deeper crust. The catalogue obtained from this study can be utilized for investigating the subsurface structure around Hachijojima Island.

　We propose a method to automatically estimate regression lines indicating the average depths of the causative layers and their wavenumber ranges from the power spectrum of potential fields, such as gravity anomalies or magnetic anomalies, by setting the adjusted coefficient of determination (R²_adj) as the indicator. In the proposed method, the data in all wavenumber zones are set as the initial data, and the regression line indicating the depth of the deepest layer is then estimated. The regression lines are estimated by sequentially reducing the data from the higher- wavenumber side, and R²_adj is calculated each time the regression line is estimated. When R²_adj of one line assumes the highest value, this line is considered the optimum regression line. The same process is then applied to the residual data (high- wavenumber side), resulting in a regression line indicating the second deepest layer. By repeating these processes, the number of causative layers is automatically determined, and the regression lines are obtained, indicating the layer depths and wavenumber ranges. In numerical tests, we confirmed that our method could estimate the model parameters correctly using L₁ and L₂ norm minimisation. We applied our method to the spectrum analysis of the Bouguer anomaly in central Kyushu, Japan, and obtained the following results as the average depth of causative layers: 8.9 km, 2.2 km, 0.6 km by L₁ norm minimizations and 9.2 km, 2.2 km, 0.6 km by L₂ norm minimizations. There were no significant differences among the results estimated by L₁ and L₂ norm minimisation and no differences in the wavenumber range for each regression line estimated by each minimisation.

　Permeability is an important rock physical property for understanding geological phenomena and for resource assessment. However, it is not feasible to measure permeability compared to other rock physical properties (e.g., porosity, elastic wave velocity, and electrical resistivity), especially for impermeable rocks. This study developed an experimental system that can measure permeability rapidly and simply using only inexpensive commercial products (less than 500,000 yen in total). In this system, a sample holder is placed in an acrylic pressure vessel and a water pressure pump is used to simultaneously pressurize the confining and pore pressure. This makes it possible to measure permeability under relatively high water pressure gradient (0.3 MPa) and near atmospheric pressure conditions. The constant water pressure method is adopted as the measurement method. Test measurements using sandstone, granite, and 3D printed sample showed that the results were consistent with those obtained using a syringe pump and with literature values. Our results under near-atmospheric conditions makes it possible to compare with other physical properties (e.g., porosity, elastic wave velocity, and electrical resistivity) under ambient pressure. The measurement range of permeability verified by this system is ~10^-10－10^-15 m², or smaller when increasing the measurement time or extending the measurement method.