Distributed Acoustic Sensing Data Research Articles

Detection of a CO2 plume by time-lapse analysis of Rayleigh-wave amplitudes extracted from downhole DAS recordings of ocean microseisms

Amplitudes of Rayleigh waves are known to decrease exponentially away from the surface. If elastic properties of the solid are stratified on a scale much smaller than the wavelength, the Rayleigh-wave amplitude at a given depth should also be affected by the formation properties at that depth: the softer the medium, the larger the amplitude. An increasingly widespread deployment of fiber-optic cables in wells makes it possible to record this depth variation of Rayleigh-wave strain amplitudes and their variations over time (due to changes of formation properties) using distributed acoustic sensing (DAS). To investigate this opportunity, we have explored temporal variations of downhole DAS amplitudes of ocean-generated Rayleigh waves during injection of CO2 into a water-saturated thin porous layer. Analysis of these data clearly shows changes of these amplitudes related to the CO2 injection. To understand these amplitude variations, we perform a theoretical analysis and numerical simulations, which explain our field observations. In particular, analysis of downhole DAS data and theoretical modeling show that Rayleigh-wave amplitudes measured with vertical fiber-optic cables can be used to detect thin layers in the subsurface. Furthermore, time-lapse analysis of these amplitudes indicates temporal changes of stiffness of these layers, such as changes in saturation or pressure of the fluid in a porous layer in the vicinity of the borehole. In particular, Rayleigh-wave amplitudes are sensitive to the presence in the vicinity of the wellbore of a CO2 plume created as a result of a small injection into a thin porous reservoir layer. Our analysis also shows that the effect of a thin layer on Rayleigh-wave amplitudes strongly depends on frequency so that at different frequencies the amplitude is affected by different combinations of elastic properties. This opens an opportunity to use amplitude variations with depth at different frequencies to separately estimate changes in bulk and shear moduli.

New wiggles from old cables: Data processing and imaging considerations

Distributed acoustic sensing (DAS) using fiber-optic cables as sensors has become a standard measurement technique in many industrialized settings. Geophysical applications have been mainly commercialized in the borehole environment. When deployed on the seabed (SB), horizontal analysis of DAS data poses new challenges due to the particularity of the receiver response and characteristics of DAS data. Recent acquisition of a 2D towed-streamer (TS) survey following the path of a communication cable (installed in 1998), which is interrogated continuously throughout the TS acquisition, enables us to jointly analyze SB-DAS and conventional TS data recorded using modern hydrophone streamer from acquisition, through different processing stages up to a migrated image. The SB-DAS recording from the old communication cable enables us to generate excellent P-wave and converted wave (PS) images of the subsurface. However, care is needed when recovering the high spatial frequencies damped by the nature of DAS measurements (which also may depend on the hardware used) and demonstrating the recovery of the signal within the data processing (DP) sequence. Our focus is on the challenges associated with the DP of entrenched marine cables. Given a simple TS processing sequence as a reference, we examine an equivalent recommended practice for signal analysis, DP, and imaging of SB-DAS data.

Integration of distributed acoustic sensing for real-time seismic monitoring of a geothermal field

To accelerate the energy transition, the exploitation of deep geothermal reservoirs is becoming a priority to supply district heating networks in areas with high potential for geothermal applications. However, the sustainable development of the resource exploitation implies minimizing the associated risks, in particular related to induced seismicity, while optimizing operational processes. Besides, the growth of this energy sector, often supported by financial aid programs, provides resources to the industry that were not available in the past to implement advanced monitoring strategies. In this context, we present a monitoring system establishing Distributed Acoustic Sensing (DAS) as an effective component of the seismic network used for the monitoring of the geothermal field of Schäftlarnstraße (Munich, Germany). We also investigate its potential for real-time seismic monitoring in an urban environment and for risk mitigation. The monitoring system is based on a data management system linking the on-site acquisition infrastructure, including the fiber optic cable deployed in an injection well and the associated DAS interrogator, to a cloud Internet-of-Things (IoT) platform. The latter is designed to deliver both a secure storage environment for the DAS recordings and optimized computing resources for their processing. The proposed solution has been tested over a six-month period under operating conditions of the geothermal field. The survey proves the feasibility of efficiently acquiring and processing the large flow of continuous DAS data. The processing outcomes, emphasized by two detected local seismic events, demonstrate the suitability of DAS, cemented behind the casing of a flowing well, for (micro-) seismic monitoring of the geothermal site. The processing applied to the data takes advantage of the high spatial density of the acquisitions for their de-noising and for the detection of events. We find that the DAS monitoring system is capable of successfully detecting an event that could not be detected by the standard surface or shallow-borehole 3C-seismometers, despite noisy conditions associated with the urban environment and the field operation. The six-month test period demonstrates the potential of DAS to be integrated as a routine seismic monitoring component of an operating geothermal field. In addition, it highlights its advantageous role as a complement to surface seismometer-based networks, particularly in urban environments.

An Object Storage for Distributed Acoustic Sensing

Abstract Large-scale processing and dissemination of distributed acoustic sensing (DAS) data are among the greatest computational challenges and opportunities of seismological research today. Current data formats and computing infrastructure are not well-adapted or user-friendly for large-scale processing. We propose an innovative, cloud-native solution for DAS seismology using the MinIO open-source object storage framework. We develop data schema for cloud-optimized data formats—Zarr and TileDB, which we deploy on a local object storage service compatible with the Amazon Web Services (AWS) storage system. We benchmark reading and writing performance for various data schema using canonical use cases in seismology. We test our framework on a local server and AWS. We find much-improved performance in compute time and memory throughout when using TileDB and Zarr compared to the conventional HDF5 data format. We demonstrate the platform with a computing heavy use case in seismology: ambient noise seismology of DAS data. We process one month of data, pairing all 2089 channels within 24 hr using AWS Batch autoscaling.

Multistage residual network for intense distributed acoustic sensing background noise attenuation

Owing to its advantages in acquisition properties, distributed acoustic sensing (DAS) is gradually being applied in seismic exploration. Unfortunately, the acquired DAS records are usually contaminated by various unwanted interferences, which are considered one of the main obstacles for subsequent processing, such as inversion and imaging of the seismic data. In general, conventional signal processing methods cannot satisfy the requirements of DAS noise attenuation owing to the limitations in presumption and denoising accuracy. Therefore, deep learning, especially convolutional neural networks (CNNs), has been used to eliminate intense background noise and improve the quality of DAS records. Nevertheless, most CNN-based methods often rely on single-scale features and cannot guarantee denoising performance when dealing with complex DAS data. In this study, we develop a multistage residual network (MSR-Net) aiming to enhance denoising ability for complex DAS background noise. More precisely, the backbone of MSR-Net adopts a multiscale architecture to capture informative features within the DAS data. A modified ringed residual module is also used to enhance the representation ability of potential features. In addition, a feature aggregation spatial attention module is designed to refine and reinforce the primary features, thereby positively impacting the denoising performance. Meanwhile, an authentic training data set is generated based on field noise data and synthetic records obtained by the forward-modeling method. Compared with conventional methods and typical denoising networks, MSR-Net demonstrates superiority in weak signal recovery and intense DAS noise attenuation.

Field assessment of elastic full-waveform inversion of combined accelerometer and distributed acoustic sensing data in a vertical seismic profile configuration

Seismic data are a significant facilitator for monitoring in carbon capture and sequestration projects, providing high-resolution images of fluid migration, using, for example, full-waveform inversion (FWI). Distributed acoustic sensing (DAS), a relatively novel technology for wavefield sampling, is well suited for this type of monitoring. Using noninvasive optical fibers, DAS allows for dense spatial sampling along the entire length of the wellbore, without disrupting operations. Permanently installed in the wellbore, typically behind casing, DAS offers highly repeatable and dense sampling of the transmitted wave modes crucial to seismic monitoring of injected carbon dioxide ([Formula: see text]). However, the DAS data consist of measurements of strain along the tangent of the fiber and therefore do not transfer directly to conventional FWI algorithms. Incorporation of DAS data in their native strain (or strain-rate) form in standard FWI algorithms, requires changing the definition of the receiver sampling operator to use geometric information about the fiber to supply tangential strain measurements to the FWI residual. The theoretical developments are applied to invert field vertical seismic profile data acquired with DAS fiber and accelerometers at a [Formula: see text] sequestration site in Newell Country, Alberta. Our method incorporates DAS data and accelerometer data in one objective function and allows us to tune the relative importance we wish to place on each data set. This method also transfers to noncollocated sensors, for example, surface-deployed geophones and borehole fiber. The inverted models contain features expected from the geology of the field site, and data modeled in the inverted models compare favorably with the field data for these sensor types. The models are derived from data acquired prior to [Formula: see text] injection, representing baseline models for future time-lapse studies planned at the field research station.

Development of fracture diagnostic methods for fluid distribution based on quantitative interpretation of distributed acoustic sensing and distributed temperature sensing

Multistage hydraulic fracturing design on horizontal wells has significantly evolved with larger fluid volume, more fracturing stages, and tighter perforation cluster spacing to efficiently stimulate unconventional reservoirs. From the published field observations, the recent fracturing design results in complex fracture networks or swarm of fractures. Fracture treatment evaluation is extremely challenging in such a case because of the large amount of variables in well completion and stimulation design. Combined measurements from different technologies can help in fracture diagnosis. Fluid distribution, either during fracture injection or during production, directly relates to the stimulation efficiency at the cluster level and at the stage level. Because it is unlikely in the real world to distribute the injected fluid uniformly among all the clusters, we need diagnostic techniques to generate the flow profile along a lateral. Fiber-optic measurements, such as distributed acoustic sensing (DAS) and distributed temperature sensing (DTS), are currently used to diagnose downhole flow conditions. This technology allows us to qualitatively confirm the fluid flow profile and other issues occurring downhole during fracturing such as leakage through plugs. For optimizing a fracturing design, we also need to understand how the design parameters are correlated with the stimulation efficiency. In this study, we combine two sets of models of DAS and DTS data interpretation for injected fluid volume distribution. The DAS is interpreted based on an empirical correlation between fluid flow rates and frequency band energy from the acoustic signals. The DTS is interpreted by performing temperature history match-based thermal energy conservation. Because of the completely different physics behind the interpretations, the confirmation of two interpretations provides confidence in fluid distribution.

Directional sources and distributed acoustic sensing deployments for near-surface characterization

Active near-surface surveys using distributed acoustic sensing (DAS) have been practically limited to 2D inline acquisitions followed by Rayleigh-wave dispersion analysis. In this study, we analyze different acquisition setups and the subsurface properties that can be estimated from them. First, we find a surface survey dedicated to Love-wave acquisition. We use a horizontal source and common-receiver sorting of the DAS data for a pure Love-wave velocity dispersion analysis yielding an S-wave velocity profile. Common-receiver sorting has the added benefit of practically eliminating the effect of the gauge length. Next, downhole DAS is used in a vertical seismic profile-type survey to recover a 1D P-wave velocity through checkshot analysis. Finally, we find that the horizontal portion of a fiber deployed in a deviated borehole records high-frequency data from surface sources. We use first arrivals for diving-wave tomography, yielding a 2D P-wave velocity model above and below the fiber. We validate the inversion results of all acquisitions through a comparison with geophone-derived velocity models. We conclude that source orientation plays a crucial role and should be considered during survey planning in conjunction with DAS directivity. As anticipated, fiber coupling can strongly influence the signal-to-noise ratio and needs to be adequately planned. In vertical boreholes, infilling offers a viable solution for deployment in existing wells. For deviated boreholes, installation inside the casing yields suboptimal coupling, but more adequate deployment protocols may yield high-frequency data useful for velocity model building and imaging. From a practical point of view, the acquisitions that we describe are difficult to justify due to their operational costs and limitations. Nonetheless, they highlight the potential of DAS in going beyond traditional analysis.

Complete perception self-attention network for weak seismic signal recovery in distributed acoustic sensing vertical seismic profile data

Distributed acoustic sensing (DAS) is a new technology for acquiring seismic data with high spatial resolution at low cost. Furthermore, in real downhole seismic exploration, DAS can receive some weak signals reflected from deep and thin layers. Unfortunately, some real downhole seismic data received by DAS often are characterized by low quality. Specifically, in real DAS records, desired signals with weak energy often are contaminated by some new noise not presented in seismic data received by conventional electronic geophones. Due to the characteristics of seismic data, such as frequency band aliasing, low signal-to-noise ratio, and complex noise wavefield, existing linear or nonlinear denoising methods based on information processing theories cannot effectively eliminate this complex and multitype noise. Recently, the deep-learning method has been regarded as a powerful tool for background noise attenuation in seismic data. Most of the existing deep-learning methods are concerned with local features and ignore the global features that can be used to enhance their performance further. To simultaneously extract global and local features, we design a novel complete perception self-attention network (CP-SANet) based on the transformer framework and apply it to the denoising of downhole DAS records. The network embeds a transformer module into multilevel encoder-decoder framework. Depth-wise convolution is applied to enhance the local perception capability. Given the transformer’s requirement for a large amount of data, we specifically design abundant seismic data samples using formation models with different parameters. The noise in the data sets is obtained from actual field DAS data. The effectiveness and feasibility of CP-SANet are verified on synthetic and field DAS records. All of the experimental results prove its satisfactory performance compared with some classical and network methods.

Performance of distributed acoustic sensing relative to co-located hydrophone measurements

Distributed acoustic sensing (DAS) is a technology that transforms telecommunication fiber optic cables into dense sensor arrays by continuously transmitting pulses of light down the cable and measuring backscattering from natural inhomogeneities within the fiber cable. The technology can densely sample the acoustic field over long ranges (up to 100 km), providing a means for large scale passive acoustic monitoring. The DAS Calibration Experiment 2022 (DASCAL22) provides 9 days of co-located hydrophone and DAS data in Puget Sound, WA. The DAS data were recorded with a 2kHz sampling rate over ∼3.5 km of cable in a channel reaching close to 100 m depth and with DAS channels spaced 6.38 m. A combination of an active source broadcast from 1, 5, and 10 m depths, and passive signals (primarily boat noise) are used to investigate the capabilities of DAS to sense a range of frequencies and the impacts of the environment on those capabilities.

Borehole fibre-optic seismology inside the Northeast Greenland Ice Stream

SUMMARYIce streams are major contributors to ice sheet mass loss and sea level rise. Effects of their dynamic behaviour are imprinted into seismic properties, such as wave speeds and anisotropy. Here, we present results from a distributed acoustic sensing (DAS) experiment in a deep ice-core borehole in the onset region of the Northeast Greenland Ice Stream, with focus on phenomenological and methodological aspects. A series of active seismic surface sources produced clear recordings of the P and S wavefield, including internal reflections, along a 1500 m long fibre-optic cable that was placed into the borehole. The combination of nonlinear traveltime tomography with a firn model constrained by multimode surface wave data, allows us to invert for P and S wave speeds with depth-dependent uncertainties on the order of only 10 m s−1, and vertical resolution of 20–70 m. The wave speed model in conjunction with the regularly spaced DAS data enable a straightforward separation of internal upward reflections followed by a reverse-time migration that provides a detailed reflectivity image of the ice. While the differences between P and S wave speeds hint at anisotropy related to crystal orientation fabric, the reflectivity image seems to carry a pronounced climatic imprint caused by rapid variations in grain size. Further improvements in resolution do not seem to be limited by the DAS channel spacing. Instead, the maximum frequency of body waves below ∼200 Hz, low signal-to-noise ratio caused by poor coupling, and systematic errors produced by the ray approximation, appear to be the leading-order issues. Among these, only the latter has a simple existing solution in the form of full-waveform inversion. Improving signal bandwidth and quality, however, will likely require a significantly larger effort in terms of both sensing equipment and logistics.

DAS with telecommunication fibre-optic cable in urban areas can record storm-induced seismic noise

SUMMARY Extreme weather events threaten life and property in populated areas. Timely and precise weather event monitoring and risk assessment are critical, but can be hampered by limited meteorological stations in cities. Recent studies have shown that seismic stations are sensitive to storm-induced noise. This study aims to investigate the sensitivity of distributed acoustic sensing (DAS), a technology capable of turning existing optical fibres into dense seismic sensors, for recording storm-induced seismic noise. We analyse 4-month continuous DAS recordings (June–September 2021) from a 4.2-km-long underground fibre-optic array in State College, PA. We calibrate the DAS data by comparing it to various meteorological data (rainfall and wind speed) from nearby weather stations. We first show DAS-recorded low-frequency wind-induced noise (0.5–8 Hz) probably caused by light poles swaying in the wind, as observed resonant frequencies agree with theoretical natural frequencies of nearby light poles. We find a strong linear correlation between DAS energy and wind speed. We further characterize rain-induced noise. Detailed observations from two rain events: a moderate rain and a heavy rain from Hurricane Ida, suggest that rain-induced noise is not generated by direct raindrops hitting the ground. Instead, the low-frequency noise (2–8 Hz) is attributed to the acoustic noise generated by water flow in stormwater drainage systems. Strong high-frequency noise up to 125 Hz is likely related to the rapid rainwater filling from the surface to the drainage system during heavy rain. We show linear relations between rain-induced DAS energy and rainfall rate, where the slopes of relations are related to the volume of rainwater, suggesting the influence of surface water and rainwater flow in the drainage system on DAS signals. Our results show the possibility of using DAS-equipped pre-existing telecom fibre-optic cables for sensing windstorms and rainstorms in urban areas and their interactions with urban infrastructures.

SLKNet: An attention-based deep-learning framework for downhole distributed acoustic sensing data denoising

Distributed acoustic sensing (DAS) is a novel and fast-developing seismic acquisition technology, which enjoys many advantages compared with traditional geophones. However, DAS data often suffer from severe and diverse types of noise with varying amplitudes, resulting in a low signal-to-noise ratio (S/N) and making the extraction of hidden signals a challenging task. Therefore, exploring a high-efficiency and high-generalization denoising method is crucial for improving the S/N of DAS data and subsequent processing. We develop a dense connection network with the kernel-wise attention mechanism to denoise complex and diverse noise (e.g., high-amplitude erratic, high-frequency, random, and horizontal noise) on real DAS data sets. We use an integrated denoising framework that is suitable for attenuating DAS noise to generate labels for network training. Our network consists of five types of blocks, i.e., convolutional, dense, transition down, transition up, and selective kernel blocks (SKB). In particular, the SKB is used to fuse multiscale features by weighting, thereby improving denoising accuracy. The computational efficiency and denoising performance are further augmented by using a patching method to segment the DAS data and generate many small-scale patches. Our network is trained on a small DAS data set and tested on the synthetic and field data from vastly different geographic areas. The comparisons of our network with three state-of-the-art deep-learning-based benchmark models demonstrate more robust performance and superior signal extraction ability.

Subglacial volcano monitoring with fibre-optic sensing: Grímsvötn, Iceland

We present a distributed acoustic sensing (DAS) experiment at Grímsvötn, Iceland. This is intended to investigate volcano-microseismicity at Grímsvötn specifically, and to assess the suitability of DAS as a subglacial volcano monitoring tool in general. In spring 2021, we trenched a 12 km long fiber-optic cable into the ice sheet around and within the caldera, followed by nearly one month of continuous recording. An image processing algorithm that exploits spatial coherence in DAS data detects on average ~100 events per day, almost 2 orders of magnitude more than in the regional earthquake catalog. A nonlinear Bayesian inversion reveals the presence of pronounced seismicity clusters, containing events with magnitudes between −3.4 and 1.7. Their close proximity to surface volcanic features suggests a geothermal origin. In addition to painting a fine-scale picture of seismic activity at Grímsvötn, this work confirms the potential of DAS in subglacial volcano monitoring.

Seismic monitoring of rockfalls using distributed acoustic sensing

Distributed acoustic sensing (DAS) using fiber-optic cables shows promise for rockfall monitoring by potentially offering high spatial resolution and wide coverage at lower costs. However, further research is needed to fully evaluate DAS performance for rockfall applications. This study assesses DAS capabilities for rockfall monitoring by analyzing DAS and seismometer recordings of 38 artificially triggered rockfalls with block masses ranging from 139.6 to 638 kg. Results demonstrate strong correlations between DAS waveforms and spectra with colocated seismometers, validating DAS data for detecting rockfall impacts. Fiber configuration impacted DAS sensitivity: fully in-contact cables improved signal-to-noise ratios and detection of high frequencies from 55 to 86 Hz. Analyses of seismic waves identified three distinct rock block propagation modes associated with intact, fragmenting, and collapse-triggering movements. While DAS signal feature values agreed with seismometers, no clear mass correlation emerged within the examined range. However, spectral features like frequency centroid (48.8 ± 6.5 Hz) and predominant frequency (41.0 ± 12.1 Hz) exhibited spatial stability across channels, indicating potential for event identification. Preliminary attempts to locate rockfall sources from DAS arrays proved feasible but require optimization. Overall, enhanced fiber coupling, identification of propagation modes, spatially-dependent analyses, and dense-array source location may enable DAS to complement sparse seismic networks for higher resolution, cost-effective rockfall monitoring. These insights could inform the development of improved rockfall monitoring strategies leveraging distributed fiber-optic sensing.

High‐Resolution Near‐Surface Imaging at the Basin Scale Using Dark Fiber and Distributed Acoustic Sensing: Toward Site Effect Estimation in Urban Environments

AbstractNear‐surface seismic structure, particularly the shear wave velocity (Vs), can strongly affect local site response, and should be accurately estimated for ground motion prediction during seismic hazard assessment. The Imperial Valley (California), occupying the southern end of the Salton Trough, is a seismically active basin with thick surficial lacustrine sedimentary deposits. In this study, we utilize ambient noise records and local earthquake events for high‐resolution near‐surface characterization and site effect estimation with an unlit fiber‐optic telecommunication infrastructure (dark fiber) in Imperial Valley by using the distributed acoustic sensing (DAS) technique. We apply ambient noise interferometry to retrieve coherent surface waves from DAS records, and evaluate performances of three different surface wave methods on DAS ambient noise dispersion imaging. We develop a quality control workflow to improve the dispersion measurement of noisy portions of the DAS data set by using a data selection strategy. Using the joint inversion of both the fundamental mode and higher overtones of Rayleigh waves, a high resolution two‐dimensional (2D) Vs structure down to 70 m depth is obtained. We successfully achieve an improved Vs30 (the time‐averaged shear‐wave velocity in the top 30 m) model with higher spatial‐resolution and reliability compared to the existing community model for the area. We also explore the potential for utilizing DAS earthquake events for site amplification estimation. The preliminary results reveal a clear anti‐correlation between the approximated site response and the Vs30 profile. Our results indicate the potential utility of DAS deployed on dark fiber for near‐surface characterization in appropriate contexts.

Characterizing hydraulic fracture growth using distributed acoustic sensing-recorded microseismic reflections

Hydraulic fracturing is crucial for enhancing hydrocarbon production from unconventional reservoirs. The characterization of fracture geometry and propagation has significant value in understanding reservoir response and designing more efficient completions. Distributed acoustic sensing (DAS) is a rapidly developing technology that can be used for this purpose because it provides wide-aperture observations of microseismic wavefields that contain direct P and S arrivals as well as converted and reflected waves. In addition to traditional approaches for microseismic event location and source mechanism analysis, the high spatial resolution of DAS microseismic recordings allows the imaging of induced fractures with reflected waves. Reflections are generated by waves radiated from microseismic events that impinge on hydraulic fractures created during prior treatment stages. We use a straightforward method based on f- k filtering and ray tracing to map reflected S waves from the time domain to reflectivity in the space domain. A case study of fracture imaging indicates that inferred fracture development, based on reflection imaging, is consistent with fracture-driven interactions observed using low-frequency DAS (LF-DAS) data. In addition, this study reveals reflection images of apparent distal fractures that do not reach the fiber and thus are not directly observed by LF-DAS. Fracture images obtained from several microseismic events during the same stage provide the opportunity to observe snapshots of dynamic fracture evolution processes.

Monitoring seismicity triggered by geothermal site shutdown with a surface DAS array at Brady Hot Springs

SUMMARY Seismicity induced by fluid injection including wastewater injection, hydrofracking and enhanced geothermal system (EGS) site production draws public attention. Dense arrays have been deployed to improve monitoring capability. In 2016 March, the PoroTomo experiment deployed an 8.6-km-long fibre-optic cable for distributed acoustic sensing (DAS) in the geothermal field at Brady Hot Springs, Nevada, covering an area of 1.5 by 0.5 km. The goal was to assess an integrated technology for characterizing and monitoring temporal changes in the rock mechanical properties of an EGS reservoir in three dimensions. We applied a neural network designed for earthquake detection called ADE-Net2 to the DAS data set to detect seismic events in continuous records. We were able to detect a total of 90 earthquakes, which included 21 events that had not been reported by a previous template-matching study. Additionally, we were able to successfully detect almost all of the active source signals, with only seven events being missed. We used the STA/LTA (short-/long-term average) method to pick arrivals and a clustering method to remove outliers. We initially tried a standard event location algorithm, but the low signal-to-noise ratio resulted in significant picking uncertainty that is up to ∼0.5 s, leading to large location uncertainty. Therefore, we developed a new location method based on the similarity between the theoretical traveltime curve and picked moveout. A grid search scheme was adopted to find the optimal point at which the traveltime curve is most similar to the picked one. Most newly detected earthquakes locate southwest of the DAS array, where five earthquakes were reported by a local seismic network. The plant began shutting down at 19:15 UTC on the March 13, and most earthquakes occurred on the March 14, indicating a relationship between the seismicity and the pressure changes caused by the shutdown of the plant. The pressure changes at epicentres obtained from a simplified model range from 71 to 157 kPa, exceeding a typical earthquake trigger threshold of 10 kPa.

Moment tensor inversion of perforation shots using distributed-acoustic sensing

Distributed-acoustic-sensing (DAS) fibers have enabled various geophysical applications in unconventional reservoirs. Combined with perforation shots, a DAS fiber can record valuable guided waves that propagate in the reservoir formation and carry information about its properties. However, the representation of perforation shots as seismic sources, needed to conduct quantitative analysis, remains unknown. We model such sources using a superposition of three mechanisms for which we derive the moment-tensor (MT) representation. Using field DAS data recorded in the same well where the perforations are located, we establish a workflow to invert the resolvable components of the total MT for 100 different perforation shots. By scrutinizing the inversion results, we conjecture that the MT can indicate how effectively a perforation shot creates microcracks in the surrounding rock. Furthermore, we observe a regular spatial pattern in the inverted MTs, which correlates with the relative location of the perforation shot within each stimulation stage.

A Self-Supervised Deep Learning Approach for Blind Denoising and Waveform Coherence Enhancement in Distributed Acoustic Sensing Data.

Fiber-optic distributed acoustic sensing (DAS) is an emerging technology for vibration measurements with numerous applications in seismic signal analysis, including microseismicity detection, ambient noise tomography, earthquake source characterization, and active source seismology. Using laser-pulse techniques, DAS turns (commercial) fiber-optic cables into seismic arrays with a spatial sampling density of the order of meters and a time sampling rate up to one thousand Hertz. The versatility of DAS enables dense instrumentation of traditionally inaccessible domains, such as urban, glaciated, and submarine environments. This in turn opens up novel applications such as traffic density monitoring and maritime vessel tracking. However, these new environments also introduce new challenges in handling various types of recorded noise, impeding the application of traditional data analysis workflows. In order to tackle the challenges posed by noise, new denoising techniques need to be explored that are tailored to DAS. In this work, we propose a Deep Learning approach that leverages the spatial density of DAS measurements to remove spatially incoherent noise with unknown characteristics. This approach is entirely self-supervised, so no noise-free ground truth is required, and it makes no assumptions regarding the noise characteristics other than that it is spatio-temporally incoherent. We apply our approach to both synthetic and real-world DAS data to demonstrate its excellent performance, even when the signals of interest are well below the noise level. Our proposed methods can be readily incorporated into conventional data processing workflows to facilitate subsequent seismological analyses.