Research Advances on Distributed Acoustic Sensing Technology for Seismology

Abstract

Distributed Acoustic Sensing (DAS) has emerged as a groundbreaking technology in seismology, transforming fiber-optic cables into dense, cost-effective seismic monitoring arrays. DAS makes use of Rayleigh backscattering to detect and measure dynamic strain and vibrations over extended distances. It can operate using both pre-existing telecommunication networks and specially designed fibers. This review explores the principles of DAS, including Coherent Optical Time Domain Reflectometry (COTDR) and Phase-Sensitive OTDR (ϕ-OTDR), and discusses the role of optoelectronic interrogators in data acquisition. It examines recent advancements in fiber design, such as helically wound and engineered fibers, which improve DAS sensitivity, spatial resolution, and the signal-to-noise ratio (SNR). Additionally, innovations in deployment techniques include cemented borehole cables, flexible liners, and weighted surface coupling to further enhance mechanical coupling and data accuracy. This review also demonstrated the applications of DAS across earthquake detection, microseismic monitoring, reservoir characterization and monitoring, carbon storage sites, geothermal reservoirs, marine environments, and urban infrastructure surveillance. The study highlighted several challenges of DAS, including directional sensitivity limitations, vast data volumes, and calibration inconsistencies. It also addressed solutions to these problems, such as advances in signal processing, noise suppression techniques, and machine learning integration, which have improved real-time analysis and data interpretability, enabling DAS to compete with traditional seismic networks. Additionally, modeling approaches such as full waveform inversion and forward simulations provide valuable insights into subsurface dynamics and fracture monitoring. This review highlights DAS’s potential to revolutionize seismic monitoring through its scalability, cost-efficiency, and adaptability to diverse applications while identifying future research directions to address its limitations and expand its capabilities.

1. Introduction

Seismology is crucial in geophysical research as it provides most of our understanding of Earth’s structure and is widely employed in studying the subsurface [1]. To fully understand plate tectonics and earthquakes, it is crucial to study long-term lithospheric deformation cycles. Observing these cycles and earthquake behaviors requires examining earthquakes across all active faults globally. This need for widespread data makes seismology a global field [2]. Studies have also demonstrated that weaker, man-made seismic waves could be effectively used to explore the Earth’s shallow subsurface, aiding in the identification of mineral, water, and hydrocarbon resources [3]. This led to the formation of the seismic exploration industry. Over time, its applications have expanded beyond resource discovery. Seismic technology is now employed in identifying suitable waste disposal sites, assessing ground stability for industrial developments, and even in archaeological research [3]. The use of seismology to reduce earthquake damage is crucial. Accurate earthquake predictions would be very effective in reducing damage, but they are difficult to make due to the complex interactions involved in the nucleation and rupture processes of earthquakes [4]. Real-time seismology, which involves collecting and analyzing seismic data quickly after an earthquake, can be used for post-earthquake emergency response and, in some cases, early warning [4]. The timeframe for real-time seismology typically ranges from minutes to hours. This means that by the time information is released, the earthquake has already concluded. The information is primarily used for post-earthquake response efforts, such as planning field work, deploying instruments, and disseminating public information. However, if information can be obtained within seconds or minutes, it can be utilized for early warning purposes [4,5,6,7].

Distributed Acoustic Sensing (DAS) has become a groundbreaking technology for seismic monitoring, enabling the transformation of fiber-optic cables into extensive seismic arrays that can cover distances of up to 100 km [8,9,10]. By emitting laser pulses and analyzing the phase shift in Rayleigh backscattering caused by inherent impurities within the fiber cable, a DAS interrogator unit detects and measures the dynamic strain or strain rate along either specialized fiber cables or existing telecommunication fiber cables [8,9]. The changes in the intensity of backscattered light over time are called distributed acoustic signatures. These signatures can identify and describe acoustic events, such as seismic waves, along the fiber [11]. The extensive dynamic range and broad frequency coverage of DAS have made it useful for monitoring seismic activities such as earthquakes, volcanic events, and glacier-related seismicity [12,13,14,15,16,17,18]. Research in this area has primarily concentrated on detecting, locating, and estimating the magnitude of seismic events using spectral data, travel time analysis, and amplitude measurements [19]. DAS technology is increasingly being adopted in seismology, with its application expanding quickly. Initially, DAS-based seismic monitoring focused on analyzing artificially induced seismic events, such as reservoir monitoring and acquiring vertical seismic profiling data from boreholes and wells [12,20]. In recent years, DAS has been employed to monitor natural seismic occurrences. Many scientific investigations have concentrated on showcasing the efficacy of DAS technology in detecting seismic events under various circumstances [21,22,23]. This involves detecting earthquakes with existing fiber-optic cables in remote regions, urban telecom networks, and even undersea cables. Unlike conventional seismic networks, which rely on seismographs spaced several kilometers apart, DAS technology offers a vastly improved spatial resolution for seismic wave monitoring, increasing it by approximately three orders of magnitude, with detection down to just a few tens of meters [24,25,26,27,28]. To monitor and study seismic activities, three-axis seismometers are commonly employed. Enhancing the precision of seismic measurements can be achieved by increasing the number of these devices. However, the high costs associated with deploying each seismometer, including the need for power supply and data connectivity make expanding a seismic sensing network both time-intensive and costly [28]. Moreover, placing seismometers in densely populated urban areas or in marine environments demands more effort, time, and increased liability. In such scenarios, fiber-optic DAS technology can serve as a more feasible alternative or complement to seismometers [29].

This review highlights the significance of seismology in geophysical research and hazard mitigation. It also discusses the role DAS plays in seismology, explaining the fundamental principles of DAS technology, particularly the use of fiber-optic cables as sensors. The review also explores recent research advancements related to DAS in seismology, focusing on acquisition, modeling, preprocessing, and processing. Additionally, it presents case studies, addresses current challenges, and discusses potential future directions for the field.

2. Fundamentals of DAS

DAS stands as an advanced technology in distributed sensing, employing light as the conveyer of information and standard telecommunications-grade optical fibers as the medium for seismic data capture [30]. DAS encompasses techniques employing optical interferometry with laser light propagation within an optical fiber to gauge the dynamic strain or strain rate at numerous points along the fiber [9]. Typical DAS systems comprise an optoelectronic instrument called the interrogation unit (IU) and a sensing cable. These units continuously emit laser pulses into the optical fiber cables. As light traverses the fiber cores, it scatters in varied directions due to spatial fluctuations in the fiber cores’ refractive index, generating diverse scattered light types, as depicted in Figure 1 [30,31]. Any disturbances, such as strain, temperature, or vibrations, experienced by the optical fibers lead to alterations in the scattered light’s properties, such as the wavelength, light intensity, and frequency. By analyzing these light characteristics, changes in multiple physical parameters, including temperature, axial strain, and strain rate, can be discerned [30]. DAS interrogators detect Rayleigh backscattered light within the fibers and analyze the phase details of coherent Rayleigh scattered light to derive measurements of dynamic strain, encompassing vibrations, acoustic waves, etc. [32]. Additionally, Brillouin and Raman scattering light is generated at every point along the optical fibers. These scattering phenomena enable the implementation of distributed strain sensing (DSS) and distributed temperature sensing (DTS), expanding the technology’s capabilities [30,31].

The underlying principle of distributed sensing relies on Optical Time Domain Reflectometry (OTDR) [33], as illustrated in Figure 2. As a laser pulse travels along an optical fiber, a small fraction of the light is inherently scattered through Rayleigh, Raman [34], and Brillouin [35] interactions, subsequently returning to the optoelectronic sensor unit. The precise measurement location is determined by the time taken for the laser pulse to travel down the sensing fiber and for the backscattered light to return to the optoelectronic sensor unit. To achieve this, the sensing fiber is stimulated with a coherent laser pulse, and the interference caused by Rayleigh backscattering along the fiber is detected and digitized. When an acoustic wave elongates the fiber, it alters the optical phase shift between backscatter components originating from the leading and trailing parts of the optical pulse. Consequently, due to interference, the intensity of the returning light varies from one pulse to another. Additionally, it is possible to calculate the optical phase to recover the acoustic phase, leading to two categories of DAS based on the detection methods of optical intensity and optical phase [36].

2.1. Coherent Optical Time Domain Reflectometry

Coherent Optical Time Domain Reflectometry (COTDR) is an advanced fiber-optic sensing technique that detects perturbations along an optical fiber through measurements of the variations in the backscatter intensity from pulse to pulse [36], as shown in Figure 3. COTDR processes the beat frequency signal generated between the local signal and the Rayleigh backscattering signal. This processing enables real-time sensing of both the phase information and the position of external vibration signals [33]. COTDR has been employed to detect alterations in temperature profiles [37] as well as acoustic vibrations [38] along fiber cables spanning several kilometers [39].

In the context of COTDR, the main distinction from traditional OTDR lies in the use of a narrow linewidth laser with a highly stable frequency and significantly longer coherence length compared to the fiber under test [33]. The operating principle of the COTDR system becomes clear when examining the radiation produced by localized scattering centers [36]. The COTDR system captures coherent scattered light resulting from two reflections with random amplitude and phase. When the fiber experiences strain, the backscatter intensity varies in accordance with the strain rate (Figure 3) but with unpredictable amplitude and phase changes along the fiber [39]. This nonlinear response poses challenges in effectively accumulating signals for multiple seismic pulses due to the difficulty in directly matching changes in amplitude and phase to the original fiber strain.

2.2. Phase-Sensitive Optical Time Domain Reflectometry

Phase-Sensitive Optical Time Domain Reflectometry (ϕ-OTDR) stands at the forefront of advanced fiber-optic sensing techniques, offering unparalleled capabilities in detecting and monitoring perturbations along optical fibers with high precision and spatial resolution [40]. At its core, ϕ-OTDR relies on the observation of coherent Rayleigh backscattering in the time domain. In this method, the sensing fiber is conceptualized as a collection of random scattering elements, each characterized by a reflectivity ($r_i\ll 1$). When a coherent pulse traverses the fiber, these elements backscatter light, and due to the coherence of the light, the fields from adjacent elements interfere with the receiver. This interference results in a distributed coherent speckle pattern whose local phase and intensity are exceptionally sensitive to local disturbances [41].

The electric field backscattered from the coherent pulse propagating along the fiber is a sum of contributions from these scattering elements [42,43]. The variation in local intensity ($\Delta I$) of the backscattered field depends on the phase variation at the scattering element, a change that can occur due to external factors like vibration or heating. These factors alter either the refractive index or the optical path length at the specific location. Therefore, the measurement of $\Delta I$ is based on the phase differences between adjacent elements, making the technique highly sensitive to even subtle local disturbances [41]. A schematic of distributed measurement using this technique is depicted in Figure 4.

One of the key challenges addressed by ϕ-OTDR is the ability to respond to external interference effectively. Unlike traditional Optical Time Domain Reflectometry (OTDR) techniques, ϕ-OTDR provides real-time monitoring by utilizing a narrow linewidth laser source [40]. This laser emits highly coherent light, ensuring precise phase measurements. The coherence length of the laser is significantly longer than the pulse width, enhancing the system’s sensitivity [33]. The technique involves injecting pulses from this coherent source into the sensing fiber through a circulator, which also taps the coherent Rayleigh backscattering signal to a photodiode. This signal is then digitized and processed in real-time, enabling the rapid and accurate acquisition of data [41].

Furthermore, ϕ-OTDR configurations can vary, including additional components like modulators, amplifiers, optical filters, and polarization controllers (Figure 5). An acousto-optic modulator (AOM) is employed to transform continuous light into a probing pulse. An erbium-doped fiber amplifier (EDFA) is utilized to offset the losses incurred in the preceding optical path and to compensate for the power lost in the optical path devices [33]. These elements are integrated to enhance system performance and accuracy. The method allows for distributed measurements of vibrations at a high repetition rate, making it suitable for real-time, high-frequency vibration monitoring over extensive fiber lengths [41].

In a DAS system, strain or acoustic signals are converted into optical signals through the principle of Rayleigh backscattering in standard fiber-optic cables. A coherent laser pulse is injected into the fiber, where microscopic imperfections in the fiber structure cause a small fraction of the light to be backscattered. External disturbances, such as strain or acoustic waves, induce minute deformations that are like elongations or compressions along the fiber, altering the phase, frequency, or intensity of the backscattered light [44]. The DAS interrogator continuously monitors these changes, using interferometric or coherent detection techniques to analyze the optical signals and detect vibrations or strain events [8]. By processing the backscattered light, the system converts optical variations into strain or acoustic data, enabling the identification and localization of events such as seismic activity, sound, or mechanical vibrations [12]. This approach effectively transforms the fiber into a dense array of virtual sensors, offering continuous, distributed measurements over long distances without the need for physical sensors along the cable.

One of the main differences between ϕ-OTDR and COTDR is that in COTDR, the laser’s frequency and polarization are intentionally varied to suppress fading noise from coherent interference, ensuring reliable measurements. In contrast, ϕ-OTDR requires a stable laser frequency and polarization for accurate phase discrimination, as any instability introduces noise and reduces sensitivity [45,46].

2.3. Optoelectronic Interrogator Unit

The Optoelectronic Interrogator Unit (IU) in DAS systems plays a pivotal role in converting optical signals into actionable data. This unit employs optical interferometry to precisely measure the phase or phase rate, and its design varies based on specific DAS approaches [9,36]. In the initial approach described by Dakin [47], two laser pulses with different frequencies, $f_1$ and $f_2$, are sequentially launched down the fiber, as shown in Table 1 (a). The backscattered signal is measured at the beat frequency ($\Delta f=\left|f_1-f_2\right|$). The temporal separation of these pulses results in a backscattered signal combining light from different locations ($x_1$ and $x_2$) along the fiber, separated by the gauge length ($x_g$) [48]. The backscattered signal phase ($\Delta \Phi$) is linearly related to the gauge strain (${\epsilon }_{xx}$) and expressed as

$${\epsilon }_{xx}\left(t,x\right)={\displaystyle \frac{\lambda }{4\pi n_c{x}_g\psi }}\Delta \mathsf{\Phi },$$

(1)

Here, $t$ and $x$ denote the positions for measuring axial strain along the fiber axis (+x direction). The parameter λ represents the frequency utilized for measurement (specifically, the beat frequency in this context), while $n_c$ and $\psi$ correspond to the refractive index and Pockels coefficient of the single-mode fiber glass (where $\psi =0.79$) [44]. Optical dispersion effects are considered for multi-frequency setups or ignored for single-frequency ones.

Table 1. Function of the Optoelectronic Interrogator Unit modified after Shatalin [36].

No	Method	Adapted Schematic Diagram	References
a	Two pulses with shifted frequencies and embedded delay		[49]
b	Interferometer with 3 × 3 coupler and embedded delay		[50]
c	Heterodyne		[51]
d	Different frequency comparator		[50,52,53]

Table 1. Function of the Optoelectronic Interrogator Unit modified after Shatalin [36].

No	Method	Adapted Schematic Diagram	References
a	Two pulses with shifted frequencies and embedded delay		[49]
b	Interferometer with 3 × 3 coupler and embedded delay		[50]
c	Heterodyne		[51]
d	Different frequency comparator		[50,52,53]

Another approach incorporates an embedded delay line, defining spatial resolution, as shown in Table 1 (b). Posey [38] presented this approach that involves the injection of a single pulse, followed by analysis of the backscattering using a Mach–Zehnder interferometer and a 3 × 3 coupler. This technique measures the change in optical phase over time, referred to as the optical phase rate. In this approach, the gauge length is regulated by gating the backscattered signal, a control mechanism restricted by the pulse width [50,54,55]. A third approach used optical heterodyne (Table 1 (c)), employing a slightly frequency-shifted local oscillator laser, measures elongation along the fiber by computing the difference in accumulated optical phase between two sections of fiber at differential frequency. Although this technique offers flexible spatial resolution, it necessitates a laser source with extremely high coherence for optimal signal-to-noise ratio performance over extended fiber lengths [44].

Additionally, multiple pulses of different frequencies can be sent either in series or from pulse to pulse, computing the phase of the backscatter signal, as shown in Table 1 (d). This phase calculation method is like the first scheme mentioned. In summary, within DAS IUs, the optical phase is measured either between consecutive gauges during a single pulse, denoted as the fast axis in Figure 6, or by combining signals from the same gauge across two repeated laser pulses, known as the slow axis [48]. Along the fast axis, the measured parameter is the optical phase (strain), whereas along the slow axis, it represents the change in phase per pulse separation time (strain rate) [9].

2.4. Benefits and Limitations of DAS

Traditional seismic sensors are devices designed to detect and measure ground motion caused by seismic waves resulting from geological events such as earthquakes, volcanic eruptions, or human-made activities like explosions. These sensors are crucial tools in seismology, geophysics, and various engineering applications [56]. They are used to detect and monitor seismic activity for scientific research, earthquake early warning systems, and oil and gas exploration.

Seismic sensors offer exceptional advantages in seismic monitoring. Their high sensitivity enables the detection of subtle ground vibrations and seismic waves, making them invaluable for earthquake research and early warning systems [57]. Arrays of these sensors provide precise location capabilities, accurately pinpointing earthquake epicenters and identifying sources of ground disturbances [58]. Moreover, traditional seismic sensors can detect seismic activity at different depths, offering valuable insights into the Earth’s subsurface layers [56]. This technology is proven and reliable, having been extensively used and validated, making it a trusted choice for scientific research and a wide range of engineering applications [58].

However, traditional seismic sensors, while highly valuable, come with several limitations. Their coverage is inherently limited [57], necessitating the deployment of multiple sensors for extensive monitoring, leading to increased costs [58]. Installation can be problematic, especially in challenging terrains or densely populated urban areas, potentially compromising the sensors’ effectiveness [58,59]. Regular maintenance and calibration are essential, adding to the operational demands. Moreover, the initial investment in high-quality seismic sensors and their installation can be prohibitively expensive, particularly for large-scale monitoring projects, presenting a significant financial challenge [28,58].

DAS is an emerging technology with enormous prospects in various scientific and technological areas. It can detect any perturbation that affects the optical path along the length of an optical fiber, such as vibrations, strain, and temperature variations. DAS requires a single interrogation unit placed at one end of the fiber, typically composed of optical and relatively low-frequency (<1 GHz) electro-optical components. The length of the fiber is typically limited to tens of kilometers due to attenuation and nonlinearity of fibers. Compared to traditional seismic sensors, DAS has several advantages. DAS can provide much denser sampling than traditional sensors, with a strain sensing unit every few meters [28]. This high spatial density allows for more detailed imaging of subsurface structures and monitoring of the built environment [60]. The high spatial sampling and information of DAS offer multidimensional features such as distance, time, frequency, and wavenumber, which can provide more meaningful and precise information than traditional sensors [28].

DAS can perform remote sensing, allowing the interrogation unit to be placed in a secure location away from harsh or inaccessible areas, which makes their deployment less intrusive [61,62]. DAS is an inexpensive solution for monitoring over long distances due to its long lifespan and low cost per monitoring point [63]. DAS can use pre-existing fiber-optic cables (dark fibers) for data acquisition, which is particularly useful in urban areas and within infrastructures [13]. This feature allows DAS to establish an unprecedented monitoring network by leveraging dark fibers [61]. Dark fibers are unused optical fibers that are available for use in optical communication systems. By using these fibers, DAS can significantly reduce the cost of deploying new cables and increase the coverage area of the monitoring network [64].

However, DAS technology, while promising, encounters several challenges. Firstly, DAS has a lower signal-to-noise ratio (SNR) in comparison to traditional geophones, limiting its ability to detect weak signals. DAS data can be contaminated by a variety of noise sources, including optical noise, seismic noise, coupling noise, and thermal noise [63]. This can make it difficult to extract the signal of interest. Efforts have been made to enhance DAS SNR by minimizing background noise or utilizing specialized engineered cables [61]. Secondly, most standard fiber-optic cables can only measure the dynamic strain or strain rate in a single direction along the fiber-optic cable. This means that it cannot directly measure the full three-component motion of the ground. Although certain specially designed cables offer broadside sensitivity [65] and multicomponent sensing [66], their performance requires further evaluation in practical field applications [61].

Additionally, the high sampling rates in DAS systems result in substantial data volumes. For instance, a DAS array with 12,000 channels sampled at 500 Hz generated a massive 128 TB of raw data over a three-month acquisition period [13]. Analyzing such a vast amount of continuous data generated by DAS systems can be computationally intensive [67]. Therefore, managing such extensive data necessitates careful considerations in terms of storage, sharing, and processing, limiting the technology’s practical application. DAS data require sophisticated processing techniques to extract meaningful information [28]. Another challenge lies in the spatial distance measurement. DAS determines the position of disturbances along the cable based on the time of Rayleigh Backscatter (RBS) flight. However, the obtained distance corresponds to the length of the fiber-optic cable rather than the actual spatial length due to the cable’s non-ideal straightness and numerous redundant segments.

Furthermore, the amplitude response of DAS systems is influenced by various factors, including the type of fiber-optic cable [68], installation procedures [12,69,70,71], and optical interferometry setups [72]. Recent studies have explored calibration methods, such as utilizing broadband seismometers [18], yet recalibration remains necessary under diverse monitoring conditions. Presently, there exists no standardized approach for calibrating DAS amplitudes, adding to the complexity of its amplitude response. The challenges faced by DAS technology underscore the necessity for continuous research and development efforts aimed at effectively overcoming its limitations. Despite these challenges, DAS has demonstrated significant promise in seismic activity monitoring, which makes it a compelling choice. With ongoing enhancements in signal processing techniques, DAS is poised to evolve into an increasingly vital tool in the realm of seismology, offering a valuable solution for comprehensive and cost-effective seismic monitoring.

3. Evolution and Development

DAS technology has undergone a significant evolution, transitioning from its initial applications in other industries to becoming a valuable tool in seismology (Figure 7) (Table 2). It was initially proposed by Barnoski [73] as the OTDR technique based on the design concept of LiDAR for monitoring fiber-optic networks. The concept of coherent OTDR (COTDR) was then introduced by Healey and Malyon [43], enhancing the system’s performance by responding to phase modulation information resulting from interference events. Subsequently, Taylor and Lee [74] developed the high-sensitivity ϕ-OTDR technique, marking the qualitative detection stage of DAS. To extract external physical information, phase demodulation technologies were introduced for demodulation of the interferometric signal within the RBS.

Masoudi [55] proposed phase demodulation based on a 3 × 3 coupler, followed by the introduction of a phase-generating carrier (PGC) demodulation method by Fang [75]. Further advancements came in 2016 when Dong [76] utilized I/Q demodulation to process optical fiber stretching signals, leading to the quantitative detection stage of DAS.

This evolution has enabled DAS technology to transition from its initial applications, broadening its scope and finding diverse applications, particularly in seismology, where its high sensitivity and resolution make it a valuable tool for monitoring seismic activities over extended distances. Mestayer [77] introduced the initial application of DAS in Vertical Seismic Profiling (VSP). Subsequent field studies have showcased its effective utilization in various practical scenarios, and for decades, DAS deployments have been utilized in VSP applications, both in onshore and offshore settings [11,20,54,78,79,80,81,82,83,84]. In these setups, optical fibers are placed within boreholes, ensuring the comprehensive coverage of wells. Traditionally, this method has been utilized for exploration endeavors and reservoir monitoring. However, its application in other areas of geosciences was limited until recently [28].

The initial discussions regarding the potential utilization of surface DAS for seismic monitoring can be traced back to 2013. Hornman [85] introduced a novel DAS cable with broadside sensitivity, enabling the acquisition of seismic reflection data at the surface with a horizontal DAS cable. Subsequently, multiple tests and experiments have been conducted to assess the effectiveness of surface DAS in earthquake detection and surface seismic monitoring [12,84,86]. In recent years, DAS technology has advanced with the development of new interrogation techniques [54,87,88,89] and data processing algorithms [90,91,92,93]. This has made it possible to use DAS for a wider range of seismic applications and to record data with higher spatial and temporal resolution.

DAS has now become a valuable tool for a wide range of seismic applications, including microseismic monitoring [94,95,96,97], wellbore integrity monitoring [98,99,100], real-time seismic monitoring [101], and induced seismicity monitoring [102]. DAS is also being used for new and emerging seismic applications, such as the monitoring of carbon capture and storage (CCS) sites [103,104,105,106,107,108,109,110], the monitoring of geothermal reservoirs [102,111,112,113,114], the monitoring of permafrost regions [115], and the monitoring of urban infrastructure [102,116,117,118,119]. The advancement in terms of technology development and various research trends are discussed in detail in the next section.

DAS has evolved significantly since its inception, driven by advancements in optical fiber technology and signal processing. Initially, the principle of optical time-domain reflectometry (OTDR) laid the foundation in the 1980s for detecting changes in backscattered light within fiber-optic cables. A major breakthrough came in the 2000s with the development of coherent OTDR (COTDR), which enabled high sensitivity to strains and vibrations over long distances, marking the transition to distributed sensing applications. Researchers such as Hartog [120] demonstrated DAS’s potential for real-time monitoring by converting standard telecom fiber into an acoustic array, opening doors for oil and gas exploration, seismic monitoring, and perimeter security.

In the energy sector, DAS gained prominence for its ability to monitor pipeline integrity and hydraulic fracturing operations, as highlighted in research by Mateeva [11,121]. Breakthrough applications in seismology emerged when Lindsey [12] successfully used DAS to detect earthquakes and surface waves over long distances, proving the system’s feasibility for dense seismic networks. Recent milestones include urban infrastructure monitoring and traffic flow analysis using pre-existing telecom fiber networks, reducing deployment costs, and enabling smart city innovations [102]. Additionally, DAS has extended to underwater applications for submarine cable sensing, supporting environmental and military surveillance [27,122]. These advancements, along with machine learning integration for enhanced signal processing, continue to broaden DAS’s applications, making it a transformative tool across industries.

Table 2. The evolution, development and applications of DAS over the years.

Stage	Key Development/Contribution	Year (s)	References
Initial Concept	OTDR technique proposed for monitoring fiber-optic networks	1976	[73]
Introduction of COTDR	COTDR introduced for phase modulation detection	1982	[43]
High-Sensitivity ϕ-OTDR	High-sensitivity ϕ-OTDR developed for qualitative detection	1993	[74]
Phase Demodulation	Phase demodulation using 3 × 3 couplers and PGC demodulation	2013–2015	[55,75]
Quantitative Detection	I/Q demodulation for quantitative detection	2016	[76]
Seismology Applications—VSP	First DAS application in VSP for seismic monitoring; practical VSP deployments in onshore/offshore wells	2011–2016	[77,83,84]
Surface DAS for Seismic Monitoring	Surface DAS with broadside sensitivity for seismic reflection data; tests on earthquake detection and surface seismic monitoring	2013–2017	[12,84,85]
Technological Advances in DAS	New interrogation techniques; advancements in data processing algorithms	2014–2019	[88,92,93]
Recent Applications in Seismology	Microseismic monitoring, wellbore integrity, real-time seismic monitoring, and induced seismicity monitoring	2020–2022	[97,101,102]
Expansion to Emerging Applications	Applications in CCS, geothermal reservoirs, permafrost monitoring, and urban infrastructure	2020–2023	[109,114,118]
Integration into Various Industries	Integration with machine learning for enhanced signal processing; urban traffic analysis, underwater cable sensing, and military/environmental surveillance	Recent Years	[27,102]

Table 2. The evolution, development and applications of DAS over the years.

Stage	Key Development/Contribution	Year (s)	References
Initial Concept	OTDR technique proposed for monitoring fiber-optic networks	1976	[73]
Introduction of COTDR	COTDR introduced for phase modulation detection	1982	[43]
High-Sensitivity ϕ-OTDR	High-sensitivity ϕ-OTDR developed for qualitative detection	1993	[74]
Phase Demodulation	Phase demodulation using 3 × 3 couplers and PGC demodulation	2013–2015	[55,75]
Quantitative Detection	I/Q demodulation for quantitative detection	2016	[76]
Seismology Applications—VSP	First DAS application in VSP for seismic monitoring; practical VSP deployments in onshore/offshore wells	2011–2016	[77,83,84]
Surface DAS for Seismic Monitoring	Surface DAS with broadside sensitivity for seismic reflection data; tests on earthquake detection and surface seismic monitoring	2013–2017	[12,84,85]
Technological Advances in DAS	New interrogation techniques; advancements in data processing algorithms	2014–2019	[88,92,93]
Recent Applications in Seismology	Microseismic monitoring, wellbore integrity, real-time seismic monitoring, and induced seismicity monitoring	2020–2022	[97,101,102]
Expansion to Emerging Applications	Applications in CCS, geothermal reservoirs, permafrost monitoring, and urban infrastructure	2020–2023	[109,114,118]
Integration into Various Industries	Integration with machine learning for enhanced signal processing; urban traffic analysis, underwater cable sensing, and military/environmental surveillance	Recent Years	[27,102]

4. Advancements and Research Trends

Addressing the challenges associated with DAS technology has led to consistent research efforts and attempts to resolve these issues to achieve superior data quality in different areas. At the acquisition stage, efforts are made to diminish noise by enhancing the signal source type, optimizing the source-receiver geometry, and refining the methods employed for optical cable deployment [123]. Further improvements involve the design of IU (reduction in noise level) and cables (broadside sensitivity) and lowering the cost of fiber-optic cable deployment [11]. In the data processing stage, personnel employ f-x deconvolution (F-X) to mitigate random noise. Additionally, they utilize fitting inversion methods, sparse optimization, and median filters to suppress coupled noise. Furthermore, a noise model trace estimation method is applied to address horizontal noise and checkerboard noise [123]. The various advancements in DAS data acquisition reported in the literature are summarized in Table 3.

4.1. Acquisition

Ensuring high-quality DAS data across diverse applications necessitates significant attention to the acquisition stage. Within this critical phase, several key advancements have been made, and these can be categorized into distinct themes. Notable areas of improvement encompass enhancements in IU design, innovations in fiber-optic cable design, optimization of cable deployment methods, and meticulous consideration of acquisition parameters. These focused developments collectively contribute to elevating the overall efficacy and reliability of DAS data acquisition processes, paving the way for more robust and precise outcomes in various applications.

4.1.1. Improvements in Interrogator Unit

In terms of improvement on IU design, Parker [54] reported the intelligent distributed acoustic sensors (iDAS) technology. The iDAS utilizes the same principle as DTS and OTDR, involving sending a light pulse into the optical fiber and analyzing scattered light for changes in the axial strain. The main difference lies in the iDAS’s ability to determine the position of each component of the returning light, allowing for a dynamic profile of the strain along the fiber. This distinct feature, coupled with repeated pulses, enables the measurement of strain changes at acoustic speeds throughout the fiber. The technology demonstrates adaptability across different applications, encompassing surface, seabed, and downhole measurements, highlighting its versatility in diverse environments. In downhole applications, iDAS provides several advantages, such as flow profiling and condition monitoring. These tasks can be seamlessly executed using the same optical fiber cable, underscoring the technology’s efficiency and continuity.

Another significant improvement in DAS IU was introduced in 2016 by Pastor-Graells [87]. A novel ϕ-OTDR interrogation approach, known as chirped-pulsed (CP-) ϕ-OTDR, was introduced and tested. The schematic of this proposed technique is shown in Figure 8. This method, based on traditional ϕ-OTDR with direct detection, differs fundamentally by employing probe pulses with a linear chirp—indicating a linear variation in instantaneous frequency across the pulse width—instead of the transform-limited pulses used in the traditional approach [124]. The operating principle is illustrated in Figure 9. Adjusting the pulse chirp profile enables tuning of the measured resolution and sensitivity. This technique eliminates the need for a frequency sweep, significantly reducing the measurement time and system complexity. It retains the potential for metric spatial resolutions over tens of kilometers, like conventional ϕ-OTDR. The method facilitates measurements at kHz rates while ensuring reliability over extended periods [87]. However, in (CP-) ϕ-OTDR, the SNR is directly linked to the coherence of the laser source [125]. This connection arises because the linear chirp in the pulse transforms frequency fluctuations of the laser (originating from its finite linewidth) into temporal shifts in the resulting trace, introducing measurement errors [88].

The issue of laser phase noise in the (CP-) ϕ-OTDR was addressed by Fernández-Ruiz [88]. A new technique to reduce the first-order term of laser noise in (CP-) ϕ-OTDR sensors was successfully introduced and tested. The schematic of the proposed phase noise cancellation technique is shown in Figure 10. This advancement results in a remarkable enhancement, achieving up to a 17 dB improvement in the SNR. Their study underscores the robustness of (CP-) ϕ-OTDR, allowing for high-quality, single-shot, and quantitative strain measurements. Importantly, this capability persists even with low-coherence laser sources, if coherence times surpass pulse width, and when utilizing direct detection methods [88]. A major limitation of this technique is compensating for higher-order phase-noise components, particularly in lasers characterized by exceptionally high levels of noise. The limitations arise due to the complexity and intensity of the noise present in these lasers, making it difficult to effectively mitigate the influence of higher-order phase-noise components. As the noise level increases, the precision required for compensation escalates, posing a significant obstacle in achieving accurate and reliable results. Numerous analyses and contributions have been made over the years to improve (CP-) ϕ-OTDR’s performance. Detailed discussion on (CP-) ϕ-OTDR can be found in Fernández-Ruiz [124].

Another common issue associated with DAS data is optical fading. Fading occurs at random positions along the fiber, where the amplitude of the backscattered signal becomes very small due to the cancellation of scattered electric fields. This leads to anomalously noisy traces in a common source gather. To address the issue of fading, Hartog and Liokumovich [127] presented a novel optical arrangement of the instrumentation (i.e., the heterodyne distributed vibration sensor (hDVS)), which enables quasi-simultaneous measurements at multiple optical interrogation frequencies. The schematic of this optical arrangement is shown in Figure 11. Utilizing distinct fading properties at various frequencies, the data are aggregated to substantially reduce noise. This approach not only alleviates the impact of fading but also improves overall measurement accuracy and linearity. Experimental and modeling results indicate that employing three independent pulses achieves a nearly zero fading probability in the aggregated signal. Enhancements in SNR and linearity are notable, with an increase in the number of frequency pulses, particularly up to around 10. However, beyond 10 frequencies, the improvement in signal quality becomes less substantial [89].

4.1.2. Cable Deployment Technique

A notable constraint of DAS lies in its lower SNR compared to geophones, as highlighted in Section 2.4. Achieving an effective transfer of energy from the seismic source to the receiver (optical fiber) becomes crucial, necessitating a secure coupling of the cable to the surrounding medium [128]. In DAS acquisition, various cable deployment techniques are implemented, which offer different levels of coupling. For borehole DAS, these include cementing the cable permanently behind the well casing, clamping the cable to production tubing inside the casing, and deploying via wireline or slickline, where the cable is installed loosely inside the borehole (Figure 12) [128]. For surface DAS, the fiber-optic cable can be deployed on the surface with poor coupling due to rough topography or buried in trenches for proper mechanical coupling [129].

Table 3. Various advancements in DAS data acquisition techniques.

Authors	Aspect	Focus	Findings/Contributions
[87,88,89,124,125,127]	Improvements in Interrogator Unit	- Φ-OTDR system with unparalleled strain sensitivity - Sensitivity variations in traditional DAS systems from noise and fading - Distributed strain sensing for linear and dynamic measurements - Enhancing SNR in high-resolution Φ-OTDR systems using optical pulse compression - Principles and Benefits of Chirped-Pulse Φ-OTDR in DAS	- Achieved record-breaking picostrain per √Hz sensitivity for distributed strain sensing. - Maintained uniform sensitivity along the fiber length - Revolutionized sensing with real-time, distributed acoustic signal measurements via optical fiber - Distributed Strain Sensor using Intensity-Only Measurements was achieved - Optical pulse compression for improved signal-to-noise ratio (SNR) - Chirped-pulse Φ-OTDR enhances performance by reducing fading points and improving sensitivity and detection accuracy
[71,84,128,129]	Cable deployment techniques	- Efficient energy transfer from seismic sources to the fiber-optic cable - Comparison of DAS data with conventional geophone data - Surface deployment strategies for DAS systems - Effects of surrounding media on DAS performance	- Optical fibers can act as efficient seismic sensors when deployed correctly - With proper deployment and coupling, DAS rivals geophone performance - Fibers on grass showed large response variations, while ground-coupling methods like sandbags or tape provided greater stability - Dense, rigid materials enhance coupling and SNR, while loose soil dampens energy transfer, reducing SNR
[66,85,128,130,131,132,133,134,135,136,137,138,139]	Cable designs	- Improving mechanical and electrical properties of cables - Analyzing electromagnetic fields to optimize cable design for telecommunications and power transmission - Advancing cable efficiency and safety - Advanced techniques for high-voltage, extreme-temperature-resistant cables - Dynamic loading: Cable fatigue and lifecycle management - Introducing lightweight, robust designs that enhance deployment and data quality in tough environments - Bio-inspired cable designs using tendons and vines to create stronger, more efficient systems - The use of recyclable materials and eco-friendly manufacturing to minimize cable production’s ecological footprint - Advanced cable coatings for enhanced durability in deep-sea exploration, oil and gas, and underwater robotics. - Integrating electrical and optical functions to enhance telecommunications and energy networks.	- Engineered fibers outperformed standard fibers in sensitivity and dynamic range - Using standard telecom fibers and data stacking provides a cost-effective alternative to engineered fibers - Choosing suitable materials, like high-strength alloys or composites, ensures cables endure mechanical stresses. - Thorough testing is essential to assess cable performance and durability before deployment - Cable design (geometry, materials, insulation) affects signal efficiency and performance - Smart cable sensors detect and diagnose faults in real time, minimizing critical system downtime - Utilizing advanced composites and nanomaterials improve cable thermal stability and conductivity - Designing specialized coatings and structures enhance fiber-optic cable durability under stress - Bio-inspired designs enhance load distribution, minimizing stress points and reducing cable failure risk. - Replace plastic insulations with eco-friendly alternatives to maintain performance and reduce environmental impact - Corrosion-resistant materials are essential to extend cable lifespan. - Combine electrical conductors and optical fibers in a single hybrid cable for power transmission and data communication

Table 3. Various advancements in DAS data acquisition techniques.

Authors	Aspect	Focus	Findings/Contributions
[87,88,89,124,125,127]	Improvements in Interrogator Unit	- Φ-OTDR system with unparalleled strain sensitivity - Sensitivity variations in traditional DAS systems from noise and fading - Distributed strain sensing for linear and dynamic measurements - Enhancing SNR in high-resolution Φ-OTDR systems using optical pulse compression - Principles and Benefits of Chirped-Pulse Φ-OTDR in DAS	- Achieved record-breaking picostrain per √Hz sensitivity for distributed strain sensing. - Maintained uniform sensitivity along the fiber length - Revolutionized sensing with real-time, distributed acoustic signal measurements via optical fiber - Distributed Strain Sensor using Intensity-Only Measurements was achieved - Optical pulse compression for improved signal-to-noise ratio (SNR) - Chirped-pulse Φ-OTDR enhances performance by reducing fading points and improving sensitivity and detection accuracy
[71,84,128,129]	Cable deployment techniques	- Efficient energy transfer from seismic sources to the fiber-optic cable - Comparison of DAS data with conventional geophone data - Surface deployment strategies for DAS systems - Effects of surrounding media on DAS performance	- Optical fibers can act as efficient seismic sensors when deployed correctly - With proper deployment and coupling, DAS rivals geophone performance - Fibers on grass showed large response variations, while ground-coupling methods like sandbags or tape provided greater stability - Dense, rigid materials enhance coupling and SNR, while loose soil dampens energy transfer, reducing SNR
[66,85,128,130,131,132,133,134,135,136,137,138,139]	Cable designs	- Improving mechanical and electrical properties of cables - Analyzing electromagnetic fields to optimize cable design for telecommunications and power transmission - Advancing cable efficiency and safety - Advanced techniques for high-voltage, extreme-temperature-resistant cables - Dynamic loading: Cable fatigue and lifecycle management - Introducing lightweight, robust designs that enhance deployment and data quality in tough environments - Bio-inspired cable designs using tendons and vines to create stronger, more efficient systems - The use of recyclable materials and eco-friendly manufacturing to minimize cable production’s ecological footprint - Advanced cable coatings for enhanced durability in deep-sea exploration, oil and gas, and underwater robotics. - Integrating electrical and optical functions to enhance telecommunications and energy networks.	- Engineered fibers outperformed standard fibers in sensitivity and dynamic range - Using standard telecom fibers and data stacking provides a cost-effective alternative to engineered fibers - Choosing suitable materials, like high-strength alloys or composites, ensures cables endure mechanical stresses. - Thorough testing is essential to assess cable performance and durability before deployment - Cable design (geometry, materials, insulation) affects signal efficiency and performance - Smart cable sensors detect and diagnose faults in real time, minimizing critical system downtime - Utilizing advanced composites and nanomaterials improve cable thermal stability and conductivity - Designing specialized coatings and structures enhance fiber-optic cable durability under stress - Bio-inspired designs enhance load distribution, minimizing stress points and reducing cable failure risk. - Replace plastic insulations with eco-friendly alternatives to maintain performance and reduce environmental impact - Corrosion-resistant materials are essential to extend cable lifespan. - Combine electrical conductors and optical fibers in a single hybrid cable for power transmission and data communication

Daley [84] conducted a series of field tests to assess the application of DAS in both borehole and surface measurements. The goal was to showcase various cable deployment options, emphasizing the need to increase cable sensitivity, particularly for deep wells or extensive surface arrays. These tests were conducted at three different CO₂ storage monitoring pilots’ sites, including Citronelle, Otway, and Ketzin. At Citronelle, DAS with tubing deployment to 2.9 km showed lower sensitivity, emphasizing the need for improved SNR. Otway’s tubing-deployed fiber in a borehole with a more energetic source demonstrated enhanced VSP data with potential sensitivity improvement. Ketzin’s loop of fiber cable on the casing, especially when cemented, provided the best data quality, emphasizing the benefits of cementing and deploying behind casing strings. These case studies showed that DAS technology can yield useful data with heightened source effort. Furthermore, cementing the fiber in place enhances data quality [84].

Surface seismic tests at Otway showcased repeatability, indicating stacking possibilities for SNR improvement [84]. Therefore, Daley [71] introduced an optimal approach termed adaptive stacking to enhance SNR and flatten the noise spectrum within the signal band. The approach involves monitoring noise power while simultaneously recording multiple independent channels with nearly identical signals. The process includes creating output channels through optimally weighted averages of these independent input channels, a technique known as weighted stacking. Further, the integration over time serves to flatten noise and convert the signal to dimensionless strain. This method offers a systematic and effective way to enhance signal quality and minimize the impact of noise during the recording process.

Despite the cementing behind casing proving to provide the best data quality due to good mechanical coupling, each of the traditional deployment techniques involves inherent trade-offs among data quality, installation cost, complexity, and ease of removal. The cementing behind the casing provides higher-quality data at high cost and complexity, and it is not removable. The clamped-on casing provides intermediate-quality data at moderate cost and is semi-removable. The wireline or slickline method is the cheapest and removable; however, it provides the lowest quality data due to poor coupling. In view of this, Munn [128] proposed a novel cable deployment technique that addresses the challenges of achieving good cable coupling in borehole DAS (Figure 12). They employed a flexible borehole liner to ensure continuous cable coupling against the borehole wall. Field experiments reveal that coupling the cable with the flexible borehole liner effectively suppresses cable waves, leading to improved visibility of clear P-wave arrivals and enhanced seismic data quality. In their study, they reemphasized the essence of cable coupling on DAS data quality using field data collected from the same borehole but under different coupling conditions. This proposed technique is most suitable for shallow boreholes with depths of 425 m or less due to the physical constraints of the liners.

Harmon [129] also investigated different coupling methods for a temporary DAS system on the surface. The methods explored include uncoupled, tension-pinned, weighted carpeting, and weighted with sandbags. Uncoupled fiber-optic cables successfully capture seismic waves generated by a nearby hammer source within a 10 m radius. Enhanced coupling, achieved using custom-made couplers or sandbags, significantly improves the SNR, extending the observation range for P-waves, S-waves, and Rayleigh waves up to 23 m from the seismic source. The alignment of DAS shot records with estimated horizontal component data confirms the cable’s capability to record horizontal motions along the fiber line. However, there were differences between the amplitudes of DAS records and estimated horizontal components from the geophones. These disparities were attributed to uncertainties that may arise from uncertainties in geophone deployment angles or fluctuations in the fiber’s responsiveness to ground motion, potentially linked to coupling challenges.

4.1.3. Cable Design

Optical fibers composed of fused silica glass exhibit limited sensitivity to normally incident seismic waves due to the relatively high rigidity of glass (Figure 13a) [130]. This lack of broadside sensitivity is generally inconsequential for borehole applications in vertical or slightly deviated wells, where impinging waves primarily travel parallel to the borehole along the installed fiber-optic cable. However, in surface seismic scenarios, where body waves propagate nearly vertically, conventional horizontal cables with straight fibers face challenges in detecting such waves effectively. This limitation also extends to the detection of distant microseismic events when using a straight fiber in a vertical well (Figure 13b) [140]. According to Kuvshinov [130], the broadside sensitivity of DAS cables can be improved by designing a configuration where seismic waves independently deform the optical fiber regardless of the angle of incidence on the cable. This improvement can be achieved by arranging the surrounding material in a manner that the normal motion of the cable boundary results in the axial motion of the cable interior. Alternatively, shaping the fiber itself offers another avenue to enhance DAS broadside sensitivity.

Kuvshinov [130], therefore, presented the theoretical concept of a helically wound fiber-optic cable for DAS in near-surface measurements (Figure 14). This study showed that enhancing sensitivity to broadside waves in DAS is achieved by helically winding fibers around cables through a reduction in the fiber wrapping angle. The optimal wrapping angle, approximately 30 degrees for plastic cables, minimizes the influence of Rayleigh waves on the measured signal. To ensure reliable seismic wave detection, establishing good mechanical contact between the cable and the surrounding medium is crucial. However, placing the cable in a cemented borehole can decrease DAS sensitivity to primary waves [130]. To validate Kuvshinov’s theory of the helically wound fiber-optic cable, Hornman [65] conducted a field trial. This study sought to validate in a qualitative sense the theoretically predicted angle-dependent response of a helically wound cable. Trial results provided qualitative confirmation of the cable’s effectiveness, showcasing its potential for seismic detection in surface seismic reflection methods. The helically wound cable, characterized by its slim design and absence of active components, emerges as a cost-effective solution for permanent seismic monitoring on land [65]. Furthermore, Spikes [131] proved that DAS cables laid on the ground alongside geophones result in high similarity, particularly for reflection energy with helically wrapped cables. Single-strand fibers show less similarity, with differing frequency content in raw and processed gathers.

Another issue associated with conventional DAS data in boreholes is the limited use for reservoir characterization. This limitation is due to the lack of multicomponent data to quantify various wave modes. This limitation was first addressed by Ning and Sava [132], who proposed solutions that involve the use of either multiple parallel optical fibers or a helical optical fiber to capture multicomponent DAS data. The shape sensing method employed with parallel optical fibers faces challenges in accurately reconstructing displacements for multiple wave modes due to insufficient information, particularly the lack of a connection between curvature and incident wave polarization. In contrast, helical optical fibers, relying on angle-dependent strain measurements, presented an effective solution for reconstructing the strain tensor, even in the presence of multiple wave modes. The helical optical fiber, assuming a seismic wavelength significantly larger than the helix period, enables angle-dependent measurements, offering improved data reconstruction [132].

Despite the successful reconstruction of the complete strain tensor, a limitation of this method arises from the assumption that the seismic wavelength is significantly larger than the defined analysis window [66]. This assumption implies a strain tensor that is considered invariant within the window. Consequently, the approach proposed by Ning and Sava [132] may not be readily applicable to the acquisition of short seismic wavelengths, such as those encountered in microseismic scenarios. To address this limitation, Ning and Sava [66] suggest a setup incorporating five equally spaced helical optical fibers with a constant pitch angle, along with a straight optical fiber (Figure 15). With this configuration, it becomes possible to reconstruct every component of the 3D strain tensor at any point along the fiber. Numerical examples served to illustrate the efficacy of this proposed method in successfully reconstructing the complete strain tensor from intricate elastic wavefields. The suggested configuration facilitates a systematic exploration of design parameters for helical optical fibers, accounting for factors like diameter and pitch angle. The objective is to identify parameters that minimize the condition number of the Gram matrix while meeting engineering constraints inherent in optical fiber construction. The design approach allows for multiple configurations of comparable robustness and quality, providing flexibility in tailoring optical fiber designs to diverse applications [66].

The sensitivity of DAS can also be improved through cable design and structure, hence improving the SNR of the DAS data acquired. Munn [128] presented a comparative study between two different cable structures: a tight-buffered composite cable and the loos tube composite cable (Figure 16). Both cables contain single-mode fiber, multimode fiber, and copper conductors, making them possible for DTS data collection as well. The key distinction between the cables lies in how the fibers are integrated into the cable structure, as shown in Figure 16. The tight-buffered cable captures a higher level of dynamic strain compared to the loose tube cable when exposed to a similar seismic signal at a matching depth. Although both cables generated interpretable VSP data, the tight-buffered cable exhibited a strain rate amplitude three times greater than that of the loose tube cable. The tight-buffered cable demonstrates a more efficient transfer of seismic energy due to entirely solid cable materials, resulting in higher sensitivity and SNR. However, tight-buffered fibers are subject to optical signal attenuation with depth due to hydrostatic pressure, imposing limitations on their maximum deployment depth in boreholes [128]. Therefore, the cable design and structure play a substantial role in influencing DAS seismic performance, necessitating careful evaluation for various deployment scenarios.

Newly engineered fibers are in development, aiming to boost sensitivity and lower the noise floor in comparison to the standard fibers employed in telecommunication networks [133,134]. Richter [134] presented a high-resolution engineered fiber called Constellation with an improved sensitivity of 100× or 20 dB compared to standard fibers. A specialized interrogator called Carina is utilized to sample the engineered fiber. Carina is specifically designed to capitalize on the improved performance of the fiber [135]. Constellation is designed with enhanced backscatter along its length, ensuring a greater reflection of light back to the interrogator. This is accomplished without imposing significant losses on the forward-propagating laser pulses. The performance of DAS using this engineered fiber is comparable to geophones around 10 Hz, yet it significantly surpasses geophone responses in the range below 1 Hz [134].

However, the widespread presence of standard fibers globally serves as a motivation to leverage existing infrastructure in DAS surveys, leading to cost-savings and streamlined logistics. Therefore, Diaz-Meza [133] presented a comparison between stack instances of standard multi-fiber cable and a co-located single-fiber engineered cable (i.e., Constellation). Analysis reveals that optical noise can be reduced by up to 20% through the stacking of DAS records from five instances of standard fiber. To address artifacts in time series arising from dynamic range saturation, an algorithm is proposed. While stacking aids in noise reduction, it falls short of restoring strain-rate amplitude from saturated signals in standard fiber DAS. On the other hand, the algorithm proves effective in restoring strain-rate amplitude from saturated signals of the engineered fiber, surpassing the dynamic range limitations of the recording [133].

Another distinct method employed to enhance DAS performance involves the use of ultra-weak fiber Bragg gratings (UWFBG) [137,141,142]. Initially reported by Zhang [142], UWFBG inscribed in the sensing fiber has demonstrated an improved SNR in DAS systems. Li [137] showcased a remarkable 21.1 dB increase in SNR, effectively measuring both the vibration and temperature simultaneously. However, the performance improvement achieved through UWFBG comes at the expense of heightened signal attenuation. Consequently, UWFBG fibers are more suitable for applications with relatively short ranges or for extending the range of conventional DAS systems by a few tens of kilometers [137]. Furthermore, the employment of fiber Bragg grating (FBG) arrays in distributed sensing introduces unwarranted complexity and necessitates highly precise manufacturing processes, potentially leading to increased costs.

Therefore, Hicke [143] introduced an innovative approach to mitigate nonlinear phase responses by employing a series of localized point reflectors (Figure 17). A similar approach is reported by Redding [138]. More recently, Masoudi [139] demonstrated a long-range DAS system utilizing ultra-low-loss enhanced-backscattering (ULEB) fiber, which also relies on point reflectors. ULEB fibers leverage point reflectors to enhance back-reflected light, ensuring the controlled inscription of reflectors to avoid excessive losses. For instance, the optical attenuation of the ULEB fiber, as reported by Masoudi [139], was measured at 2.05 dB/km, equating to an excess loss of 0.05 dB/km compared to a standard telecom fiber. The minimal attenuation of ULEB fibers positions them as an ideal solution for applications requiring high sensitivity across an extended sensing range [144].

4.2. Modeling

Modeling and simulation play pivotal roles in both research and industrial contexts, providing a versatile toolkit for predictive analysis, cost-effective testing, and risk reduction. These tools enable researchers and industries to optimize designs and processes, fostering innovation and pushing the boundaries of what is achievable. Simulation-based training enhances skill development, while the ability to understand complex systems is greatly facilitated. The time efficiency of simulations accelerates research and development, and their application in environmental impact assessment contributes to sustainable practices. Moreover, modeling drives technological advancements, particularly in fields like engineering, materials science, and medicine. The seamless integration of modeling and simulation enhances efficiency, reduces costs, and catalyzes breakthroughs, underscoring their indispensable role in advancing knowledge and industry practices. In this section, the various advances in DAS technology through modeling and simulation, as well as the advances in modeling and simulation in DAS-technology-related research, are discussed, as summarized in Table 4.

4.2.1. Interrogator Unit Design

In the advancement of DAS technology, modeling and simulation have played a significant role. In terms of IU design and improvements, Liokumovich [145] created a statistical model to analyze signals in phase-sensitive optical time domain reflectometry (OTDR) using highly coherent sources. The backscattering process was intricately modeled through discrete scatterers with meticulously selected parameters. Explicit equations were formulated to calculate the amplitude and phase of the backscattered signal. The model successfully predicted the spectral and autocorrelation characteristics of amplitude signals, aligning well with experimental results. Additionally, they presented and analyzed the characteristics of phase signals, which are crucial for OTDR sensing applications, demonstrating a strong correspondence with experimental data. This development formally opened avenues for more detailed modeling of DAS systems, offering insights into their responses to disturbances along optical fibers.

However, their examination of the sensing system was conducted under static conditions and did not encompass the influence of dynamic perturbations. In view of this, Masoudi and Newson [146] presented an alternative method for the numerical analysis of phase-sensitive DAS. In their study, a numerical model of a distributed optical fiber acoustic sensor is developed, enhancing flexibility by separately modeling the key building blocks of the sensing system. These components are then integrated into a comprehensive numerical model, enabling a systematic evaluation of each block’s impact on the overall sensor output. The model is employed to investigate the effects of various parameters, including laser source linewidth, probe pulse width, and perturbation frequency and amplitude, on the performance of the sensing system. Results reveal that precision and accuracy are influenced by these parameters, emphasizing their significance in optimizing the sensor’s response.

4.2.2. Cable Deployment

As discussed earlier, cable deployment techniques have been a crucial area for improving the quality of DAS data. Schilke et al. [147] addressed challenges faced by DAS when functioning independently, with a focus on SNR issues that can impede the identification of clear seismic first arrivals in VSP surveys. The authors highlighted the coupling of sensors with the medium as a critical factor affecting the quality of seismic recordings. Specifically, their study concentrated on DAS deployment inside tubing, utilizing numerical simulations to understand the impact of coupling on recording quality. They delved into the dynamics of wave motion in fluid-filled boreholes and analyzed seismic signals detected by the fiber-optic cable within the borehole. The overarching goal was to determine optimal signal recording conditions for DAS systems, drawing insights from both real and numerical data. Their study revealed that achieving a favorable SNR with DAS systems is contingent upon optimal contact between the cable and the borehole wall. To enhance cable contact, they proposed adding extra cable, with the quantity determined by factors such as borehole length, diameter, cable dimensions, and elastic properties influencing stiffness. Furthermore, they emphasized the consideration of crucial parameters, including the coefficient of friction and cable elasticity, in determining contact forces that prevent cable slipping and ensure effective coupling [147].

4.2.3. Cable Design

In terms of cable design, Eaid [148] addressed challenges in Multi-parameter Full Waveform Inversion (FWI) with DAS, emphasizing the limitation of directional information from shaped fibers (e.g., helically wound fiber cables) due to extended gauge lengths required for a meaningful SNR. They developed an analytical description of the relationship between the fiber shape, gauge length, and elastic wave sensitivity, integrating DAS into FWI and understanding the impact of fiber geometry on parameter resolutions. Through 2D simulations with helical fibers, they discerned that the fiber wind rate significantly influences parameter accuracy, stressing the importance of smaller gauge lengths compared to the wind rate for the optimal utilization of fiber shape advantages in FWI. Their study showed that the integration of DAS data in FWI presents valuable opportunities for seismic acquisition. Furthermore, for the optimal utilization of shaped fibers, it is essential to incorporate sub-period gauge lengths.

With a more realistic model of a DAS system (discussed in the next subsection), van Putten [144] examined the impact of ULEB fibers on the efficiency of DAS systems. They conducted a comparative analysis of the responses of DAS systems, utilizing both standard single-mode fibers and ULEB fibers. Investigating the influence of noise levels and the different fiber types on the performance of the DAS system, they showed that ULEB fibers can significantly enhance SNR by nearly one order of magnitude without necessitating costly components or advanced signal processing techniques. Notably, ULEB fibers exhibited the potential to improve the precision of strain measurements without compromising the accuracy of frequency and strain measurements in the DAS system. The results demonstrated a threefold enhancement in measurement precision when employing ULEB fibers, and noise reduction is visibly observed in the time-domain data. The study underscores the effective utilization of ULEB fibers to augment SNR, emphasizing their efficacy even in scenarios where vibrations directly impact enhanced reflectivity points [144].

4.2.4. DAS Data Interpretation Aided by Modeling

DAS has emerged as a crucial diagnostic technique for detecting and characterizing hydraulic fractures, contributing to a more profound understanding of fracture propagation and interference [149,150,151]. This insight is instrumental in optimizing well-completion designs and overall field development [150,152]. The ability of DAS to characterize the geometry and orientation of hydraulic fractures offers a means to identify potential issues during and after well completion. This information is not only valuable for estimating production but also facilitates the design of cost-effective wells that enhance overall production [150]. Recent advancements in DAS technology have enabled the direct monitoring of subsurface deformation during hydraulic fracturing (HF), marking a significant leap forward in the field [69,150,153,154].

In HF monitoring, interpreting field DAS results can be challenging due to complex subsurface conditions and noise [151]. Raw DAS data, recorded as optical phase, are influenced by the interference pattern of backscattered light at two adjacent observation points, defining the gauge length [150,152,155]. This optical phase exhibits a linear relationship with the strain rate, allowing predictions of strain rate variations from numerical simulations to be compared with DAS signals [150,156]. Liu [151] and Tan [157] have compared numerical simulations with field DAS measurements to interpret fracture geometry and optimize completion designs. However, these studies often focus on qualitative analyses and lack detailed interpretations of interference scenarios contributing to DAS signals. These studies employ the displacement discontinuity method (DDM) for fracture propagation simulation, which is known for computational efficiency but is limited in directly simulating flow within a heterogeneous matrix.

In view of this, Chen [158] proposed a sequentially coupled multiphase flow and geomechanical simulation approach to address these limitations and enhance the understanding of hydraulic fracture deformation and propagation based on the Discrete Fracture Model (DFM). They introduced a fracture propagation model implemented in MATLAB Reservoir Simulation Toolbox, combining flow and geomechanical computations. The model was validated against Khristianovich–Geertsma–Deklerk (KGD) analytical solutions, ensuring accuracy and simulated stress and strain features aligning with field DAS signals. Case studies demonstrate the model’s utility in analyzing fracture interference, closure, and stress shadowing effects, providing insights for field measurement interpretation and HF design optimization. The identification of fracture hits, recognition of multiple adjacent fractures, and definition of the antenna in DAS data contribute to a comprehensive understanding of hydraulic fracture behavior during monitoring.

Table 4. Numerous advancements in techniques for modeling DAS data.

Authors	Aspect	Focus	Findings/Contributions
[145,146]	Interrogator unit design	- Statistical Modeling for OTDR signals - Numerical Modeling for Phase-Sensitive - DAS	- Created a statistical model for analyzing phase-sensitive OTDR signals with coherent sources - Predicted spectral and autocorrelation characteristics of amplitude signals, which showed strong agreement with experimental results. - Analyzed phase signals critical for OTDR sensing. - Created a flexible numerical model for distributed optical fiber acoustic sensors - Highlighted the importance of parameter optimization for enhanced sensor performance - Incorporated dynamic perturbation analysis to overcome static model limitations.
[147]	Cable deployment	- Signal processing techniques for improved sensitivity or resolution - Effect of environmental factors on DAS performance	- Increased cable slack enhances coupling, improving DAS signal-to-noise ratios
[144,148]	Cable designs	- Developing efficient methods to extract source information from DAS-recorded microseismic data - Numerical modeling of a DAS system based on ultra-low loss-enhanced backscattering (ULEB) fibers	- DAS enables non-invasive monitoring of hydraulic fracturing by detecting strain from propagating wavefields. - DAS’s dense spatial sampling enables wide-aperture recording essential for estimating microseismic source parameters - ULEB fibers, with evenly spaced highly reflective points, enhance DAS measurement SNR and linearity compared to standard single-mode fibers. - Enhanced backscattering in ULEB fibers improves DAS performance, enabling high sensitivity over long sensing ranges.
[150,151,154,157,158,159,160,161,162,163,164,165,166]	DAS interpretation aided by modeling	- Displacement Discontinuity Method (DDM) - Forward Modeling for DAS Signals - Mixed-Mode DAS Signal Analysis - Scattered Wave Analysis - Full Waveform Inversion (FWI) with DAS	- Analyzes strain rate and DAS signals to interpret fractures and optimize designs - Efficient for fracture propagation but limited for heterogeneous flow modeling - Simulates fracture propagation by integrating multiphase flow and geomechanics, validated analytically for fracture interference, closure, and stress shadowing - Analyzes low-frequency DAS signals, defines fracture dimensions, and creates signal templates - Simulates anomalous DAS signals from fault reactivation using a simplified DDM model to assess deformation and seismic hazard - Uses PS waves and low-velocity modeling to estimate SRV height and HF closure time - FWI with DAS yielded results comparable to traditional 4D inversion but faced challenges like limited illumination from sparse fiber arrangements
[144,146]	Advancement in DAS modeling	- Numerical analysis of phase-sensitive DAS by modeling and integrating individual components to simulate the system - A more advanced DAS model integrating non-ideal optical components with their inherent noises and imperfections	- The numerical model was designed for an ideal system with perfect components. - An advanced DAS system accounted for laser phase noise, digitizer quantization, and detector and optical amplifier noise - The simulation results provided key insights into DAS system dynamics, bridging ideal models and practical implementations by addressing optical noise and imperfections

Table 4. Numerous advancements in techniques for modeling DAS data.

Authors	Aspect	Focus	Findings/Contributions
[145,146]	Interrogator unit design	- Statistical Modeling for OTDR signals - Numerical Modeling for Phase-Sensitive - DAS	- Created a statistical model for analyzing phase-sensitive OTDR signals with coherent sources - Predicted spectral and autocorrelation characteristics of amplitude signals, which showed strong agreement with experimental results. - Analyzed phase signals critical for OTDR sensing. - Created a flexible numerical model for distributed optical fiber acoustic sensors - Highlighted the importance of parameter optimization for enhanced sensor performance - Incorporated dynamic perturbation analysis to overcome static model limitations.
[147]	Cable deployment	- Signal processing techniques for improved sensitivity or resolution - Effect of environmental factors on DAS performance	- Increased cable slack enhances coupling, improving DAS signal-to-noise ratios
[144,148]	Cable designs	- Developing efficient methods to extract source information from DAS-recorded microseismic data - Numerical modeling of a DAS system based on ultra-low loss-enhanced backscattering (ULEB) fibers	- DAS enables non-invasive monitoring of hydraulic fracturing by detecting strain from propagating wavefields. - DAS’s dense spatial sampling enables wide-aperture recording essential for estimating microseismic source parameters - ULEB fibers, with evenly spaced highly reflective points, enhance DAS measurement SNR and linearity compared to standard single-mode fibers. - Enhanced backscattering in ULEB fibers improves DAS performance, enabling high sensitivity over long sensing ranges.
[150,151,154,157,158,159,160,161,162,163,164,165,166]	DAS interpretation aided by modeling	- Displacement Discontinuity Method (DDM) - Forward Modeling for DAS Signals - Mixed-Mode DAS Signal Analysis - Scattered Wave Analysis - Full Waveform Inversion (FWI) with DAS	- Analyzes strain rate and DAS signals to interpret fractures and optimize designs - Efficient for fracture propagation but limited for heterogeneous flow modeling - Simulates fracture propagation by integrating multiphase flow and geomechanics, validated analytically for fracture interference, closure, and stress shadowing - Analyzes low-frequency DAS signals, defines fracture dimensions, and creates signal templates - Simulates anomalous DAS signals from fault reactivation using a simplified DDM model to assess deformation and seismic hazard - Uses PS waves and low-velocity modeling to estimate SRV height and HF closure time - FWI with DAS yielded results comparable to traditional 4D inversion but faced challenges like limited illumination from sparse fiber arrangements
[144,146]	Advancement in DAS modeling	- Numerical analysis of phase-sensitive DAS by modeling and integrating individual components to simulate the system - A more advanced DAS model integrating non-ideal optical components with their inherent noises and imperfections	- The numerical model was designed for an ideal system with perfect components. - An advanced DAS system accounted for laser phase noise, digitizer quantization, and detector and optical amplifier noise - The simulation results provided key insights into DAS system dynamics, bridging ideal models and practical implementations by addressing optical noise and imperfections

Forward modeling is a valuable approach for characterizing and extracting insights from field-measured Low-Frequency DAS (LFDAS) signals. Various studies have successfully applied forward modeling to various scenarios, including characterizing frac-hits and strain changes during production shut-in and reopening [167,168], constraining hydraulic fracture lengths/heights [154], examining signals due to temperature changes [169], and generating templates for recognizing typical LFDAS signals [157]. The integration of forward modeling with analytic methods has further enabled inversion for fracture geometries. These advances contribute to a more comprehensive understanding of HF processes and optimize field development strategies. As mentioned earlier, most studies have employed the DDM for fracture propagation simulations. However, limited research uses DDM to study mixed-mode LFDAS signals indicating induced fault reactivation during HF. While less frequent, induced fault reactivation events are not uncommon in HF operations [170,171,172]. However, understanding LFDAS signals from such mixed-mode failures is challenging due to uncertainties in fault location, orientation, and reactivation schemes [173].

Most LFDAS signals, typically displaying characteristics of Mode-I hydraulic fractures, can reveal the presence of pre-existing natural fractures or faults leading to mixed-mode failure. This involves shear slip (Mode-II) and dilation, with poorly understood LFDAS signal characteristics [161]. An anomalous LFDAS signal during HF operations, distinct from typical signals, suggests mixed-mode reactivation on a preexisting fault [161,173]. Using a simplified numerical model based on the DDM, Wang [161,173] introduced an approach that approximates the first-order characteristics of the anomalous signal as initial tensile fault opening (Mode-I) followed by shear slip (Mode-II) on a fault. The comparison of simulations with a field-observed signal demonstrates a good fit, providing a useful tool for understanding and quantifying mixed-mode LFDAS signals and enhancing the comprehension of an induced slip during HF. Sensitivity analysis and simulation results contribute to the identification of subsurface deformation linked to potential seismic hazards [161,173]. Overall, their work improved the understanding and characterization of mixed-mode fault reactivation events during HF, offering valuable insights for operational optimization and hazard mitigation.

To optimize completion design and well spacing for HF operations, the characterization of the stimulated rock volume (SRV) is crucial [159]. Various geophysical tools, including microseismic monitoring [160], microdeformation measurements [162], and strain measurements [150], have been developed for SRV characterization. While microseismic and microdeformation methods can estimate the 3D geometry of HFs, they have limited spatial resolution. Fiber-optic-based crosswell strain measurements offer high spatial resolution (<5 m) but are confined to monitor well locations, as demonstrated by Jin and Roy [150]. Additionally, techniques like surface or borehole active sources with downhole receivers have been employed to characterize induced fractures through analyses of direct or transmitted waves [174,175]. However, the use of scattered waves, encompassing reflected, converted, and diffracted waves, presents an alternative for estimating HF properties, providing additional insights into the subsurface changes associated with HF. Various investigations employing methodologies such as ray-tracing, full-wavefield modeling, and time-lapse seismic responses show how scattered waves can be effectively employed in SRV characterization [176,177,178,179,180].

Scattered waves observed during an interstage DAS VSP survey [181] have been used to estimate the half-height of SRV by Titov [159]. The scattered waves were identified as converted PS waves through kinematic travel time analysis and full-wavefield modeling. The authors tested three models (i.e., single HF, low-velocity zone (LVZ), and tip diffractors) for fracture-induced velocity inhomogeneities; the LVZ model showed the best fit, representing an SRV. The authors proposed a novel approach utilizing PS-waves converted by the SRV to estimate the SRV half-height and HF closure time, offering insights into real-time, cost-effective monitoring and critical constraints for optimizing unconventional field development. The analysis, involving derived travel time equations and numerical modeling, provided a comprehensive understanding of the observed scattered events and their implications for fracture characterization. The estimated half-height of the SRV, derived from these events, ranged from 245 to 310 m, with each event lasting more than one day. The stacked amplitude decay time was linked to the fracture closure rate, providing valuable information about the permeability of the stimulated reservoir [159].

Recent studies demonstrate the integration of DAS with Full Waveform Inversion (FWI) for enhanced seismic imaging across various applications. Yust [165] emphasized the critical role of starting models in FWI, showing that while waveform misfit reductions were similar across models, deeper subsurface imaging (>10 m) showed significant discrepancies. Validations against borehole lithology logs confirmed higher reliability near-surface but highlighted the need for site-specific starting models. Feng [166] and Wang [163] explored DAS-FWI for imaging steam chambers in Steam-Assisted Gravity Drainage (SAGD) operations, demonstrating cost-effective and operationally efficient monitoring through tubing-deployed fiber-optic cables. While these studies showed results comparable to traditional 4D inversion methods, challenges such as limited illumination due to sparse fiber arrangements were noted, with surface trench fibers suggested to improve resolution. Pan [164] focused on DAS-FWI in CO₂ sequestration at the CaMI Field Research Station, showcasing DAS’s ability to record high-quality surface waves with low spatial aliasing and resolving near-surface S-wave velocity (V_s) and attenuation (Q_s) models with high resolution. Across all studies, DAS proves to be a promising, cost-effective tool for seismic monitoring, though challenges like illumination constraints and starting model selection remain critical for further advancements.

4.2.5. Advancement in DAS Modeling

The aforementioned discussions have highlighted the pivotal role that modeling plays in advancing DAS technology, with a particular emphasis on enhancing the acquisition process to achieve superior data quality. By employing sophisticated modeling techniques, researchers and practitioners can make significant strides in optimizing DAS technology, ultimately leading to more accurate and reliable data acquisition methodologies. The utilization of advanced models facilitates a deeper understanding of the intricate dynamics involved in DAS systems, enabling the identification and mitigation of potential challenges. This emphasis on modeling underscores its indispensable contribution to the continual improvement and evolution of DAS technology, fostering advancements that directly translate into enhanced data quality and overall system performance. In this subsection, the various advancements in modeling toward achieving more realistic models and improving the analysis of DAS data through modeling are discussed.

As discussed earlier under Section 4.2.1, Masoudi [146] conducted a numerical analysis of a phase-sensitive DAS, modeling individual components separately and integrating them to simulate the entire system. However, the numerical model created was tailored for an ideal system featuring perfect components. In view of this, van Putten [144] presented a more comprehensive DAS model integrating non-ideal optical components with their inherent noises and imperfections (Figure 18). These include factors such as laser phase noise, digitizer quantization, and noises from the detector and optical amplifier. The simulation results from this study offer profound insights into the dynamics of a realistic DAS system, bridging the gap between the ideal model and practical implementations by accounting for noises and imperfections in the optical components. The developed numerical model emerges as a valuable tool, enabling a systematic assessment of how inherent noises and imperfections in optical components impact the overall performance of DAS systems, thus contributing to the ongoing refinement and optimization of this cutting-edge technology [144].

4.3. Preprocessing

Preprocessing DAS data involves steps to enhance signal quality and prepare it for analysis. Key tasks include organizing data, extracting metadata, and performing quality checks to identify anomalies. Noise removal techniques such as bandpass filtering, spike suppression, and adaptive filtering help isolate relevant signals [182,183]. Signal conditioning adjusts sampling rates, normalizes amplitudes, and corrects for phase continuity. Alignment of data to cable geometry and time references ensures consistency. Depending on the application, such as seismic analysis, flow monitoring, or structural health monitoring, specialized preprocessing like strain-to-velocity conversion or event detection may be applied [184].

4.3.1. Initial Stacking

DAS records single-component strain rates with dense spatial sampling along fiber-optic cables, generating massive datasets during continuous recording [182]. To manage data size and enhance signal clarity, initial processing involves decimation (selective sampling) and vertical stacking (averaging traces) to suppress noise and reinforce coherent signals. In DAS Vertical Seismic Profile (VSP) processing, the workflow begins by defining the survey geometry (e.g., zero-offset VSP) to align the source and receiver, followed by identifying first breaks and the initial seismic wave arrivals. These first breaks are critical for flattening data, which separates upgoing and downgoing wavefields, facilitating interval velocity determination, time-depth conversion, and seismic event alignment [185]. To further improve the signal-to-noise ratio, stacking combines similar seismic traces to amplify coherent signals and attenuate random noise [184]. Accurate depth calibration of DAS channels is essential, as precise fiber positioning directly impacts velocity models and seismic imaging reliability [186]. Optical noise arising from fiber imperfections and external factors poses a challenge in DAS. Techniques like common-shot stacking address this by creating representative noise gathers, which are subtracted from the original DAS data to clean the signal. The cleaned data are then correlated with noise-free pilot signals generated by the seismic source, enhancing the precision of seismic imaging [187]. This systematic approach optimizes DAS data for advanced analysis and reliable subsurface characterization.

4.3.2. Conversion of DAS Strain Rate to Ground Motion

Conventional seismometers measure ground motion (e.g., particle velocity) in three dimensions, whereas DAS records ground strain (rate) along a single dimension, specifically along the fiber length. Therefore, the swift adoption of DAS for various seismological applications necessitates fundamental research to assess its performance in comparison to conventional seismometers for the accurate characterization and calibration of DAS systems [18,22,71,188,189,190,191,192]. Furthermore, the full integration of DAS for numerous fundamental seismological tasks remains to be established, and conventional seismic processing workflows rely on ground motion data, such as particle velocity, as the input [193,194,195]. These reasons necessitate the reliable conversion of the strain (rate) to ground motion and vice versa.

The relationship between DAS measurement (i.e., strain rate) and standard geophone measurements (particle velocity) can be shown by the equations [196]

$${\displaystyle \frac{\partial }{\partial t}}{\left({\displaystyle \frac{\partial u_z}{\partial z}}\right)}_{z_0,t_0}={\dot{\epsilon }}_{zz}\left(z_0,t_0\right),$$

(2)

and

$${\displaystyle \frac{\partial }{\partial z}}{\left({\displaystyle \frac{\partial u_z}{\partial t}}\right)}_{z_0,t_0}={\left({\displaystyle \frac{\partial v_z}{\partial z}}\right)}_{z_0,t_0},$$

(3)

where u (z₀, t₀) is the dynamic displacement of the fiber at the location z₀ and measurement time t₀, respectively, ε_zz is the axial strain, $v$ is the particle velocity at a fixed point, and the dot denotes its time derivative. Equation (2) converts the output of Silixa’s iDAS to the strain rate of the fiber, while Equation (3) is the spatial derivation of the dynamic displacement of the fiber.

Considering a primary seismic plane wave traveling in the z-direction with an apparent velocity c (or slowness s = 1/c) and an angular frequency ω. Subsequently, the displacement and velocity field of the seismic wave are expressed as

$$u_z\left(z,t\right)=u{e}^{-i\left(\omega t-k_{z}z\right)}=u{e}^{-i\omega t}{e}^{-{\displaystyle \frac{i\omega z}{c}}},$$

(4)

$$v_z\left(z,t\right)=\left(-i\omega \right)u{e}^{-i\left(\omega t-k_{z}z\right)}=\left(-i\omega \right)u{e}^{-i\omega t}{e}^{-{\displaystyle \frac{i\omega z}{c}}},$$

(5)

where u is the constant displacement equal to u_z(0, 0), and k_z is the vertical component of the wavenumber. From Equations (4) and (5), Bakku [197] expressed the relationship between strain rate and particle velocity as

$${\dot{\epsilon }}_{zz}={\displaystyle \frac{\partial v_z}{\partial z}}=i{k}_z{v}_z={\displaystyle \frac{i\omega cos\theta }{V}}{v}_z.$$

(6)

In this relationship, V is the inherent velocity in the medium. Hence, the strain rate amplitude exhibits a cosine dependence on the incident angle of the seismic wave in contrast to geophone data. On the other hand, Daley [71] related the dynamic strain to the particle velocity using the apparent velocity expressed as

$${\epsilon }_{zz}=-{\displaystyle \frac{v_z}{c}}$$

(7)

It can be deduced from Equation (7) that the amplitude of the strain rate is dependent on the apparent velocity in the z direction. The relationship between the strain (rate) and particle velocity has been used to convert data between these two quantities for various purposes, as summarized above. Table 4 summarizes the various works that involve conversion between strain (rate) and ground motion data. In the subsequent paragraphs under this section, we discuss the various conversion techniques that have been proposed over the years, their various limitations, and their advantages.

Daley [71] used Equation (7) to convert the strain rate data of a limited interval to particle velocity, assuming a constant apparent velocity. This relation implies that the magnitude would vary across layers characterized by distinct apparent velocities. Furthermore, a time integration step is required to either convert strain rate to strain or particle acceleration to particle velocity. Nakajima [196] further established a relationship between the strain rate and the overall amplitude of the particle velocity. They established that $\sin v_z$ component is directly related to $\left|v_r\right|\cos \theta$; hence, the relationship transforms to

$${\epsilon }_{zz}=-{\displaystyle \frac{\left|v_r\right|{\cos }^2\theta }{V}}$$

(8)

Bóna [192] introduced a more intricate conversion method between DAS and geophone data, which accounts for gauge length and pulse width effects. Their assessment involves a two-spatial averaging filter, where the amplitude ratio between DAS and geophone is influenced by the seismic field’s wavelength relative to the gauge length. However, for this conversion filter, regularizations are required to prevent division by zero, which may occur for certain wavenumbers. Another conversion filter with further improvements was presented by Egorov [194]. In this study, they utilized a conversion filter that represents the ratio of the vertical component of particle velocity to the box approximations of each pulse light on the same plane. To avoid potential division by zero, a regularization coefficient, a small positive number, was incorporated into the filter. This approach, akin to methods by Bakku [197] and Dean [72], except for regularization and accounting for pulse width, enhances data processing in the vertical wavenumber domain. The challenge in this DAS conversion resides in reconstructing events characterized by near-infinite apparent velocity.

Like the observation made in Equation (7), for horizontal DAS, variations in the apparent velocity along the cable are anticipated due to local factors such as sedimentary cover, seismic wave velocities, and the direction of propagation relative to the cable’s orientation [195]. Therefore, Wang [22] converted strain rate to particle velocity in the frequency-wavenumber (f-k) domain. The process begins with the conversion of raw DAS data into strain, followed by Fourier transformation into the f-k domain. Fourier coefficients are then adjusted by k/ω for integration and differentiation to derive particle velocity. The inverse transformation back to the time–space domain yields particle velocity for each channel. The results of the DAS-to-geophone waveform comparison were reasonably good for a few cycles after arrival but could be affected by coda waves associated with geological heterogeneity. Furthermore, the f-k rescaling method might face instability in scenarios where the seismic wavefield is characterized by k_x values approaching 0. This situation can pose challenges, particularly when teleseismic body waves have a near-vertical incidence on horizontal DAS arrays or when surface waves arrive from a broadside back azimuth on a linear DAS array [18]. In view of this, Lindsey [18] remediated this small number division instability in the f-k rescaling technique by numerically adjusting the frequency and wavenumber values by the same water level parameter (η). Nevertheless, the water level parameter may have a strong effect on the phase velocity and may, therefore, yield an unphysical modification [191].

The variations in seismic phase velocities and directions necessitate diverse apparent velocities for converting strain rate to ground motions. A fixed value may introduce bias, especially considering velocity fluctuations along the fiber due to local structures and orientation changes. Robust conversion of DAS records to ground motions, accommodating varying phase velocities, is crucial for various seismological applications. Estimating phase velocities through the f-k-based techniques can be intricate. Achieving suitable temporal and spatial resolutions demands lengthy cable segments and time intervals, with interpretation complexities. A more suitable method to retrieve time-dependent phase velocities is warranted. In view of this, Lindsey [195] proposed a technique for continuous estimation of apparent phase velocity using a semblance-based local slant-stack transform. This method proved to improve the reliability of conversion from strain rate to ground motion compared to the f-k-based method.

Nodal seismometer data can be converted into the strain rate [22,188]. This conversion compromises waveform coherence, emphasizing slow phases like surface waves in beamforming. Applying the same logic, the reverse is anticipated to hold true [189]. Therefore, to mitigate the impacts of heterogeneities and slow surface waves, van den Ende and Ampuero [189] proposed conversion from strain rates to particle velocities through spatial integration if a seismometer is co-located with the fiber, especially along straight sections with consistent coupling. In this approach, no assumptions regarding the apparent propagation speed of the signals are required. However, integration errors will accumulate as the strain rate is integrated away from the seismometer due to deviations from assumptions like a straight cable, uniform coupling, and instrument noise. A least squares inversion scheme has also been proposed to obtain particle velocity along the fiber from strain rate in a DAS system [193]. This approach is developed based on DAS principles; however, regularization is crucial to prevent amplification of noise before first arrivals.

Finally, another approach to deal with uncertainties associated with the estimation of apparent velocity in the physics-based conversion has been proposed by Fu [198]. Authors devised a machine-learning-based methodology to transform DAS data into geophone data. This approach stands out for its precision and resilience in capturing the intricate details of the wavefield. Noteworthy for its ability to augment spatial resolution and coverage, this method seamlessly integrates into established seismic processing, imaging, and inversion workflows.

4.3.3. Denoising

Traditional seismic noise attenuation techniques can be classified into four major categories, each designed to address specific aspects of signal enhancement. Predictive filtering algorithms are used to model and predict the desired signals in either the time or frequency domain. Techniques such as f-x deconvolution, median filtering, and nonstationary predictive filtering have been extensively utilized to achieve this goal [199,200,201]. These methods are particularly effective in identifying and suppressing coherent noise while preserving seismic events. Mode decomposition algorithms, another approach, work by breaking down the noisy seismic data into various components. This allows for isolating and retaining only the meaningful signal components, with methods like empirical mode decomposition, variational mode decomposition, and singular-value decomposition being widely adopted [202,203,204,205,206]. Additionally, sparse transformation algorithms leverage transformations such as Fourier, seislet, or wavelet to map data into sparse domains. Noise can then be effectively separated from signals using thresholding techniques, ensuring the integrity of the original signal [207,208,209,210]. Finally, rank-reduction algorithms utilize methods like multichannel singular spectrum analysis and damped rank reduction to isolate signals by reducing the rank of the noise matrix, improving clarity and focus on the seismic data of interest [93,211,212,213]. Each of these methodologies plays a critical role in seismic data processing, addressing challenges posed by noisy environments and enabling accurate interpretation of subsurface features.

Researchers have explored diverse strategies to reduce noise in DAS data, improving its utility for seismic applications. Median and low-pass filters effectively attenuate noise, as demonstrated by Lellouch [214] at the FORGE site, while band-pass filters suppress high-frequency noise but may inadvertently remove low-amplitude signals [215]. To address complex noise patterns like checkerboard artifacts and directional interference, median and dip filters have been useful [216]. Sparse transformation algorithms mitigate coupled noise caused by instrument vibrations [216]. Advanced denoising methods from reflection seismology include the wavelet method for separating coherent signals [217], the curvelet method [218,219], and principal component analysis (PCA) for reducing noise while retaining signal structure [220]. Yang [221] proposed a comprehensive denoising framework combining band-pass filtering, SOMF for erratic noise suppression, and dip filtering in the f-k domain. Basic approaches like weighted-mean stacking [222], linear filtering [223], and median trace subtraction [224] also effectively addressed background noise. Additional techniques, such as subtracting synthetic noise models [91] and dip filtering, specifically target horizontal noise. Bagheri [225] combined frequency-offset deconvolution (FXD) and decision-based median (DBM) filters to reduce Gaussian and impulsive noise. These strategies provide both practical and advanced solutions to mitigate diverse DAS noise challenges.

Conventional denoising methods, while effective, face significant limitations. Independent variable analysis, which processes data using adjacent traces, risks destroying effective signals by misinterpreting noise as a signal, leading to information loss [226]. Wavelet transform methods combined with Independent Component Analysis (ICA) are constrained to single-direction transformations, making them ineffective for seismic signals with multi-directional variations in complex geological environments [226]. Sparse transformation-based algorithms struggle due to their reliance on precise threshold selection, which is challenging without prior information, while band-pass filtering often suppresses valuable signal components along with noise [221]. The spatial variability of DAS data complicates parameter selection, with manual adjustments risking improper settings and false event identification [201]. Traditional approaches, like time-frequency peak filtering (TFPF), variational mode decomposition (VMD), and robust principal component analysis (RPCA), depend heavily on human expertise for parameter tuning, which can introduce inaccuracies [227]. Additionally, linear models used in conventional methods fail to effectively address nonlinear noise components, further limiting their performance [228].

Deep learning techniques can independently identify complex, multivariate, and nonlinear features from large datasets without prior knowledge [229,230]. For DAS, various denoising methods have been developed. Zhao [231] integrated Variational Mode Decomposition (VMD) with Convolutional Neural Networks (CNNs) to denoise seismic data. Wang and Nealon [232] employed CNNs to enhance 3D seismic data quality, while Zhong [233] proposed RCEN, a residual encoder-decoder network with iterative memory and channel aggregation blocks, outperforming other methods for DAS-VSP denoising. Attention-based CNNs were used for rapid response denoising in field scenarios [234], and Gao et al. [235] developed an unsupervised method using signal reconstruction. Calvarons [236] improved the Noise2Noise (N2N) method to function with fewer noisy data pairs, while Feng and Li [237] combined Singular Spectrum Analysis (SSA) with deep learning for enhanced signal denoising. Van den Ende [238] introduced a self-supervised neural network for blind denoising by leveraging DAS spatial density. Additionally, Dong [239] applied data augmentation to reduce supervised learning’s reliance on real noise data. Further advancements include unsupervised random denoising [240] and real-time CNN-based microseismic detection for low signal-to-noise ratio data [241]. Li [242] integrated Convolutional Attention (CA) with Multi-Scale Residual Networks (MSRNet) to enhance weak signal recovery. Tian [243] developed an Iterative PA-MRNet, capable of suppressing various noise types while recovering critical subsurface signals. Ma [244] introduced adversarial learning and domain adaptation to denoise without paired data. Attention-guided CNNs by Wang [245] and multi-scale approaches like MSDC-Net [246] effectively reduce noise while preserving key details. Methods like CP-SANet [227] and MSAACNN [247] utilize transformer frameworks and dilated convolutions for broader feature extraction. Zhu [248] employed diffusion models for probabilistic denoising, while Gu [249] compared supervised learning with Noise2Noise, which uses noisy data pairs for training. High-low feature fusion models [250] combine convolutional and transformer strengths to recover signals. Techniques such as Distributing-Local-Attention Expansion (DLAE) [251] and Nonlocal Selective Attention mechanisms [228] further optimize weak signal extraction. Self-supervised methods, including DAS-N2N [252] and Noise2Self extensions [253], eliminate reliance on labeled data. Li [254] enhanced seismic quality using spatial–temporal correlations, and Cui et al. [255] applied attention-based deep image priors for denoising without pretraining.

4.4. Processing and Imaging of DAS Seismic Data

The processing and imaging workflow for DAS seismic data involves a systematic sequence of steps to handle the unique characteristics of DAS measurements and produce high-resolution subsurface images. DAS records strain-rate data along an optical fiber, requiring specialized pre-processing to address signal fidelity challenges. The workflow begins with pre-processing, including noise removal, trace editing, and correction for DC bias or drift caused by optical fiber imperfections [11,84]. A band-pass filter is applied to isolate the desired frequency range and suppress low-frequency noise or high-frequency artifacts [54]. Due to DAS’s sensitivity to axial strain along the fiber, additional polarization corrections or directional filtering may be implemented to account for the fiber’s directional sensitivity relative to seismic wave propagation [130]. Signal enhancement techniques such as stacking, trace normalization, and cross-correlation are applied to improve the signal-to-noise ratio and align DAS data with geophone-equivalent outputs [256,257].

Following pre-processing, DAS strain-rate measurements are often integrated into the frequency domain to convert them to particle velocity or displacement, aligning with conventional seismic workflows [195]. Special attention is given to spatial resampling, as DAS produces densely sampled data along the fiber, which can require decimation or interpolation for computational efficiency [258]. Once pre-processed, the DAS data undergoes imaging using standard seismic imaging techniques, such as vertical seismic profile (VSP) imaging, Kirchhoff migration, or reverse time migration (RTM). The densely sampled DAS data allow for more robust wavefield reconstruction and migration, enhancing the resolution of the final subsurface images [259,260]. Quality control steps involve comparing the processed DAS data with co-located geophone datasets to validate results and assess amplitude fidelity, arrival times, and wavelet character. DAS’s unique advantage is dense spatial sampling, which enables finer resolution of seismic wave propagation and improves imaging of subsurface structures. This workflow ensures that DAS data can complement or, in some cases, replace traditional geophones in seismic applications for reservoir monitoring, borehole characterization, and carbon storage monitoring [54,71].

Earthquake Source Parameter Detection

Earthquake source parameter detection using DAS involves extracting key earthquake characteristics, such as hypocenter location, magnitude, fault orientation, and rupture dynamics, using dynamic strain or strain-rate data measured along fiber-optic cables. DAS transforms existing fiber infrastructure into dense seismic arrays, providing continuous, spatially distributed measurements with resolutions of meters to tens of meters [24,261]. The arrival times of seismic waves recorded by DAS can be inverted to locate the earthquake hypocenter, while amplitude-based inversion helps estimate the magnitude and seismic moment. DAS’s dense coverage further enables moment tensor inversion, revealing fault orientation and slip direction [195]. The high spatial resolution allows for the study of rupture dynamics, including the initiation and propagation of fault ruptures, as well as the detection of smaller earthquakes, such as foreshocks and aftershocks, that traditional seismometers might miss [121,262]. Additionally, DAS is particularly effective for near-fault imaging, capturing seismic signals close to fault zones where conventional instrumentation struggles. While DAS generates vast data volumes requiring efficient processing workflows, recent advances, such as improved strain-to-velocity conversion approaches [195], have enhanced the accuracy of source parameter detection. The use of ambient noise interferometry and real-time monitoring also makes DAS a promising tool for seismic studies in remote and inaccessible areas [12,263]. Despite challenges such as environmental noise interference, DAS’s cost-effectiveness and ability to leverage existing telecommunication fibers make it a transformative technology for earthquake detection and fault characterization [24,261].

5. Applications of DAS in Seismology

Seismology is a broad and interdisciplinary field that extends beyond earthquake studies, encompassing various specialized branches. For this study, the applications were classified into five primary categories: Earthquake Seismology, Exploration Seismology, Engineering Seismology, Environmental and Hydro-Seismology, and Volcanic Seismology. A comprehensive review of approximately 474 scholarly articles on seismological applications reveals that DAS is predominantly employed in Exploration Seismology, representing approximately 56.1% of the studies. This is followed by Earthquake Seismology at 21.9%, while Engineering Seismology, Environmental and Hydro-Seismology, and Volcanic Seismology account for 10.3%, 9.9%, and 1.7%, respectively (Figure 19). A selection of these applications is discussed in the following sections.

5.1. Earthquake Seismology

Wang [22] compared DAS with conventional seismometers for assessing ground motion response to earthquakes, demonstrating its high spatial resolution but noting differences in amplitude and phase. Lindsey [261] and Nayak [264] explored the repurposing of dark fiber for seismic monitoring, highlighting its cost-effectiveness and ability to detect seismic activity over vast areas, while Karrenbach [265] applied DAS to monitor aftershocks from the 2019 Ridgecrest earthquake, showcasing its potential in real-time seismic hazard assessment. Studies by Yetik et al. [29] and Farghal [266] further validated DAS’s capability for earthquake epicenter localization and early warning applications, emphasizing its scalability and efficiency. Li et al. [267] demonstrated the effectiveness of DAS in converting fiber-optic cables into extensive seismic arrays for high-frequency earthquake imaging, with Baba [268] applying it to offshore monitoring of slow-slip events in the Nankai Trough. Ajo-Franklin [13] and Matias [269] expanded DAS applications to subsurface characterization and tsunami monitoring using submarine cables, proving its viability for oceanographic research. Biondi [270] and Lior [271] examined DAS’s potential for urban seismic monitoring and underwater acoustic detection, demonstrating its versatility in various environments. Additional studies, including those by Yetik [29], Zeng [272], and Lior [273], underscored DAS’s ability to enhance earthquake early warning by providing rapid magnitude estimation and ground motion forecasts. Tsuji [274] introduced a Portable Active Seismic Source (PASS) system for geological imaging, which, when combined with DAS, could improve geological monitoring. Yin [275] demonstrated offshore DAS applications for rapid earthquake localization, estimating magnitude directly from strain rate amplitudes, while Zhai [276] showcased DAS’s potential in detecting previously unrecorded seismic events, reinforcing its role in advancing earthquake hazard assessment. Further research by Cheng et al. [277] explored its role in submarine tectonic studies, while Yu et al. [23] and Williams et al. [27] showed its effectiveness in detecting teleseismic events in deep-sea environments, expanding seismic monitoring to previously under-covered regions. Another key application of DAS is in volcanic seismology. Nishimura et al. [278] showed how DAS can track magma movement, pinpoint earthquake origins, and detect volcanic tremors, making it valuable for real-time hazard assessments. Brenguier et al. [279] demonstrated its role in monitoring fault zones, using background noise from sources like passing trains to study fault dynamics.

5.2. Exploration Seismology

Several studies have demonstrated the growing significance of DAS technology in oil and gas exploration and production, particularly for monitoring subsurface activities. Correa [280] and Cheng [281] integrated DAS with Surface Orbital Vibrators (SOVs) and permanent orbital sources to achieve real-time, continuous monitoring of CO₂ injection (Otway Project) and unconventional reservoirs (Eagle Ford Shale), showing high spatial–temporal resolution and repeatable seismic signals. Also, Rebel [110] tested Distributed Acoustic Sensing (DAS) on seabed and subsea fiber-optic cables using active and passive seismic methods, demonstrating effective seismic signal capture and subsurface imaging comparable to conventional techniques. Atterholt [282] advanced fault zone imaging by converting DAS strain measurements into detailed seismic wave analyses. Applications in shallow subsurface imaging were explored by Jiang [283] and Yuan [262] using urban noise and surface wave data, enhancing near-surface characterization. Similarly, Beilecke [284] and Harris [285] applied DAS for CO₂ storage monitoring, effectively identifying fault structures and reservoir changes. Studies by Dou [86] and Abukrat [263] underscored DAS’s potential when combined with natural seismic sources but stressed the need for improved signal processing. Finally, Daley [71] integrated DAS with geophones for borehole monitoring, while Wang and Stewart [286] and Abbas [287] showcased DAS’s capabilities in imaging geothermal systems and complex geological features. Asfha [288] provided an in-depth review of sand production mechanisms, current prediction techniques, and the role of emerging technologies such as fiber-optic sensing and machine learning in monitoring and managing sand production.

5.3. Engineering Seismology

Several studies have demonstrated the versatility of DAS technology using fiber-optic cables for monitoring infrastructure and urban activities. Wang [289] utilized fiber-optic networks in Pasadena, California, to monitor traffic changes during COVID-19 by transforming underground cables into seismic sensors. Similarly, Lindsey [290] leveraged dark fiber DAS to analyze shifts in human and transportation activity caused by pandemic lockdowns. Kowarik [291] applied DAS to detect train movements, capturing real-time data on location, speed, and axle counts. Rivet [292] used DAS on telecom cables to monitor ships, providing insights into ship trajectories, speeds, and types, showcasing its value for maritime safety and environmental surveillance. Spica [293] demonstrated DAS’s role in urban seismic site characterization at Stanford University, utilizing techniques like HVSR and Rayleigh wave analysis to map subsurface structures. Finally, Fang [102] highlighted DAS’s ability to acquire high-resolution seismic data, aiding in the analysis of subsurface characteristics, seismic risks, and infrastructure stability. Quinn [119] found DAS effective for soil vibration monitoring, with better performance in newer segments. Zhang [294] demonstrated DAS’s value in real-time tunnel monitoring by detecting structural disturbances. Kishida [295] applied DFOS to a high-speed railway bridge, tracking deformations and seismic impacts, with DAS capturing dynamic train responses.

5.4. Environmental and Hydro-Seismology

Walter [296] and Booth [32] demonstrated the potential of DAS in environmental monitoring, particularly for understanding glacial dynamics. Xie et al. [297] deployed a DAS array on a frozen lake, utilizing water vibrations and AI techniques to detect icequakes and flexural-gravity waves, offering insights into ice plate properties and disintegration mechanisms. Lin [298] utilized submarine optical-fiber cables during Typhoon Muifa to measure ocean current speed and direction by analyzing microseismic noise induced by surface gravity waves, demonstrating the feasibility of real-time current monitoring in extreme weather. Xie [299] highlighted how DAS, using fiber-optic cables to capture seismic vibrations, enables precise rockfall detection and volume assessment, aiding hazard mitigation. Similarly, Piatz [300] leveraged DAS to measure strain rates and detect ground deformations caused by avalanches, providing detailed spatial and temporal data that enhance early warning systems. Expanding its application further, Tribaldos [301] employed DAS to analyze ambient seismic noise over a five-month period along a 23 km fiber-optic cable in California’s Sacramento Valley, revealing correlations between seismic velocity variations, precipitation events, and river stage fluctuations, thereby demonstrating DAS’s potential in groundwater monitoring.

5.5. Volcanic Seismology

Biagioli et al. [302] used DAS technology at Stromboli volcano, Italy, with a 1 km fiber-optic cable to analyze strain signals from explosions and tremors, providing insights into volcanic activity. Nakano et al. [303] tested DAS on Tonga’s undersea telecom cable, showing it as a cost-effective tool for monitoring underwater volcanic activity. Caudron [304] used DAS at Laacher See volcano, Germany, detecting and classifying underwater gas emissions, proving its potential for real-time degassing monitoring. Jousset [16] deployed a 1.3 km cable near Mount Etna, capturing high-resolution strain data to detect volcanic events and subsurface processes. Currenti [305] tested DAS at Mount Etna with a 1.5 km cable and 26 seismometers, confirming its accuracy in mapping local strain changes. Currenti [306] monitored low-frequency volcanic activity between Vulcano Island and Sicily, detecting 1488 events and showing how DAS, combined with machine learning, enhances volcanic hazard assessment.

6. Challenges and Future Directions

DAS systems exhibit lower sensitivity compared to traditional broadband seismometers, limiting their ability to detect weak seismic signals such as small earthquakes or subtle earth vibrations. This challenge is compounded in urban or shallow environments where high noise levels, including instrumental noise from interrogators and environmental noise from traffic, wind, and human activities, obscure seismic signals [12,249]. Spatial resolution in DAS is influenced by gauge length and pulse spacing, with improvements often resulting in trade-offs, such as reduced signal-to-noise ratio or range, which limits the ability to detect small-scale seismic features or localized events [84,263]. Directional sensitivity further complicates DAS measurements, as strain rate is measured only along the cable axis, making it difficult to resolve vertical and transverse ground motions accurately [22,307]. DAS systems also generate vast datasets due to dense spatial sampling and high temporal resolution, placing significant demands on storage, transfer, and computational processing, which necessitate advanced machine learning and signal analysis techniques [249,308]. Finally, DAS deployment is constrained by the availability of fiber-optic cables, which may not be ideally located for seismic studies, leading to coverage gaps. Installing new cables, particularly in remote or offshore regions, remains logistically challenging and cost-intensive [84,263,309].

To address the sensitivity and noise issues of DAS, advancements in interrogator technology, such as improved laser stability, higher signal-to-noise ratios, and optimized signal processing algorithms, can enhance the detection of weak seismic signals [12,102]. Noise reduction techniques, including machine-learning-based filtering and adaptive noise cancellation, can help isolate seismic signals from environmental and instrumental noise [249]. For spatial resolution constraints, optimizing gauge lengths and implementing dynamic signal processing approaches can balance resolution and sensitivity while extending the operational range [84,263]. To mitigate directional sensitivity, hybrid approaches that integrate DAS with traditional three-component seismometers can provide complementary measurements to resolve full ground motion [307]. Data management challenges can be addressed through cloud-based storage solutions, advanced compression techniques, and machine-learning algorithms that enable real-time processing and anomaly detection [249,308]. For deployment limitations, leveraging existing fiber-optic infrastructure, such as telecommunications and subsea cables, along with advancements in low-cost fiber-laying technologies, can improve coverage in remote and offshore areas [84,309]. Combined, these innovations will significantly enhance DAS capabilities for large-scale, cost-effective seismic monitoring.

Future research in DAS should focus on improving sensitivity and noise reduction through advancements in interrogator technology, signal processing algorithms, and machine-learning-based noise filtering to enhance the detection of weak seismic signals. Developing methods to optimize gauge lengths and pulse spacing will help balance spatial resolution and range, enabling improved detection of small-scale seismic features. Research into hybrid systems that integrate DAS with traditional three-component seismometers can address directional sensitivity and allow full wavefield reconstruction. Additionally, innovative data management strategies, such as real-time processing, data compression, and cloud-based analytics, are needed to handle the vast datasets generated by DAS. Finally, expanding the global deployment of DAS through the utilization of existing fiber-optic infrastructure and the development of cost-effective cable-laying solutions will be critical to improving coverage, particularly in remote and offshore regions. These research areas will significantly enhance the performance, reliability, and scalability of DAS technology for seismic monitoring and earth science applications.

Integration of Communication and Fiber Sensing Techniques

Deploying new fiber-optic cables in remote areas, especially across oceans, is costly. However, with more than 1.4 million kilometers of fiber already installed underwater [310], integrating fiber-optic sensing with existing communication infrastructure offers a practical and cost-effective solution. Traditional DAS relies on backscattered Rayleigh light, which often requires extra hardware and, in some cases, new fiber deployment. Additionally, its single-span sensing range is limited to approximately 50 km without amplification, making it incompatible with optical amplifiers used in communication networks, which are typically spaced 90 km apart [311,312,313].

In contrast, forward transmission-based fiber-optic sensing is an emerging technique that uses forward-propagating light, enhancing the signal-to-noise ratio and extending sensing distances up to 200 km [314]. This method is fully compatible with communication networks as it utilizes the same optical fiber, signal, and demodulation equipment [315] while also ensuring that it does not interfere with data transmission or consume additional bandwidth. Despite these advantages, it presents challenges in pinpointing the exact locations of disturbances and differentiating multiple vibration sources with similar frequencies. Researchers have successfully demonstrated ultra-long multi-span sensing, reaching distances of thousands of kilometers using existing submarine fiber-optic cables [25,316,317]. Various demodulation techniques, such as intensity [318], phase [311,319], and polarization [320], are being explored to enhance sensing accuracy. Ultimately, forward transmission-based sensing is emerging as a promising alternative to DAS, offering extended sensing reach and seamless integration into existing fiber-optic networks without requiring additional hardware.

7. Conclusions

DAS has emerged as a transformative tool in seismological research and hazard mitigation. By leveraging existing fiber-optic infrastructure, DAS offers unprecedented spatial coverage, enabling high-resolution detection and monitoring of seismic events in regions that were previously inaccessible or under-monitored. Its cost-effectiveness, scalability, and ability to collect dense, real-time seismic data make it a compelling complement to traditional seismometers. Recent advancements, such as improved signal processing, machine learning integration, and hybrid approaches with conventional sensors, have significantly enhanced its sensitivity and accuracy. Applications extend from earthquake monitoring and fault imaging to volcanic hazard assessments and early tsunami warnings, highlighting DAS’s versatility in addressing pressing geophysical challenges. As research continues to address limitations in noise reduction, data management, and deployment, DAS is poised to revolutionize seismic monitoring, strengthen early warning systems, and advance our understanding of Earth’s dynamic processes, ultimately contributing to global hazard mitigation efforts. The future of DAS research lies in improving signal-to-noise ratio, enhancing spatial resolution, and integrating hybrid systems for full ground motion detection. Efficient data management through machine learning and real-time processing is essential to handle large datasets. Expanding deployment in remote and marine regions using existing infrastructure and cost-effective solutions will enhance global coverage. Further advancements in AI-driven signal analysis and exploration of marine and planetary applications will solidify DAS as a transformative tool for seismic monitoring and hazard mitigation. The use of forward transmission-based fiber-optic sensing methods should also be explored. In conclusion, while DAS is promising, more research is needed to improve its ability to measure seismic wave characteristics, determine earthquake magnitudes, and pinpoint locations accurately.