Marine seismic exploration is traditionally conducted using towed streamers to investigate the geological structure of sea shelves and identify mineral deposits. Conventional streamers typically use piezoelectric hydrophones or fiber-optic interferometric hydrophones, which are complex, costly, and challenging to manufacture. In this study, we introduced a fiber-optic marine towed streamer seismic acquisition system based on distributed acoustic sensing technology. This system features a simplified design by removing the need for optical components within the streamer, thereby streamlining system architecture and manufacturing. The system’s effectiveness was validated through a sea trial conducted in the slope zone of a basin, with water depths ranging from 500 to 2000 m. Notably, this study represents the first successful application of distributed fiber-optic towed streamers for marine seismic exploration, enabling the effective detection of complex sedimentary structures in the surveyed area. The results underscore the significant potential of distributed fiber-optic towed streamers for seismic exploration, paving the way for advancements in marine seismic technologies.
A marine seismic survey is a highly effective technique for exploring oceanic energy resources, such as oil [1] and natural gas hydrates [2,3]. This method typically involves the use of towed streamer cables equipped with numerous hydrophones to capture subsurface reflections of acoustic energy generated by seismic sources, such as air guns or sparkers. Marine towed streamers serve as receivers for these sound signals, and the extensive collection of recordings enables the generation of detailed subsurface images in the exploration area [4]. These streamers are widely utilized in geophysical research and the seismic exploration of mineral resources in continental shelf regions [5].
Due to the high cost of equipment development and field experiments in seismic exploration, technological advancements in this field are highly specialized and not widely accessible. The Seal 428, manufactured by Sercel in France, is a well-known marine seismic acquisition system extensively used worldwide [6]. This system is a high-capacity, high-resolution seismic data acquisition tool specifically designed for marine towed streamers. Featuring advanced electronics and workstation technologies, it has proven highly effective for seismic exploration. In one study, a 1200 m, 96-channel Sercel Seal 428 streamer was employed to conduct a high-resolution two-dimensional (2D) seismic survey of Middle Pleistocene glaciogenic sediments in the Central Graben, North Sea [7]. Additionally, in the Arctic Mega Project—a geological and multidisciplinary investigation of the Arctic Ocean led by the Russian government—a Sercel Seal system was utilized to perform multi-channel seismic surveys as part of the initiative [8].
In 2019, a new deep-towed streamer for sub-bottom profiling, capable of operating at water depths of up to 2000 m, was developed [9]. To withstand high pressures, the electrical components within the acoustic module that acquire seismic data were immersed in silicon oil, and a specialized anti-pressure treatment was applied to the electrical components in the main devices. The streamer comprised several catenated single modules, with recorded acoustic data transmitted to the control computer via the Ethernet protocol. This configuration allowed for the catenation of up to 24 channels in the streamer.
In 2022, a high-resolution deep-towed seismic system named the Kuiyang-ST2000 was developed [10]. Unlike other equipment, the Kuiyang-ST2000 utilized a deep-towed sparker instead of a resonator source, enabling it to operate at a water depth of 2000 m for the first time.
Compared to traditional electronic technologies, fiber-optic towed streamers provide several advantages, including high sensitivity, resilience to harsh environments, and immunity to electromagnetic interference [11]. Fiber-optic technology has been widely applied in hydroacoustics and geophysics [12,13]. In 2017, Beijing Appsoft Technology Co., Ltd. (China) developed a fiber-optic hydrophone array that was employed for marine seismic exploration in the northern South China Sea (19°N-21°N, 114°E-116°E) [14,15]. The streamer cable, comprising 1024 fiber-optic hydrophones, exhibited significant advantages in shallow-resolution imaging compared to traditional electronic towed streamers.
In 2021, a method to compensate for the flow noise in a fiber-optic streamer using an additional interferometer was proposed. This method demonstrated that flow noise could be reduced by up to 70% in the low-frequency range, up to 40 Hz [16]. Hydrophones in fiber-optic streamers typically adopt Michelson or Mach-Zehnder interferometer configurations [17]. These large-scale multiplexed fiber-optic hydrophone arrays rely on numerous optical components to achieve time-division and wavelength-division multiplexing. Precise control of the splitting ratio in the optical configuration is required, increasing the system’s complexity and manufacturing costs.
This study presents a fiber-optic towed streamer seismic acquisition system that employs distributed acoustic sensing (DAS) technology. The system featured a unique design in which a single optical fiber was helically wound on a continuous acoustic-sensitive material, forming a fully distributed sensing towed streamer cable. Additionally, two discrete-sensor towed streamer cables were designed and manufactured. These cables incorporated optical fibers wound on multiple hydrophone skeletons, enabling comparative analysis. A key advantage of these towed streamer cables was the elimination of optical components, which simplified both the system and the manufacturing process.
During the sea trial, all of the towed streamer cables effectively captured the reflection and structural information of formation interfaces. They provided valuable insights into the complex sedimentary processes occurring in the slope zone of the basin, at water depths ranging from 500 to 2000 m. This technology demonstrated significant potential for future geological surveying applications, offering promising advancements in the field.
2. Material and methods
2.1. Heterodyne demodulation distributed acoustic sensing system
DAS is based on phase-sensitive optical time-domain reflectometry (Φ-OTDR), in which a narrow linewidth laser is injected into an optical fiber to stimulate Rayleigh back-scattered light waves. Using advanced signal processing techniques, acoustic-induced light phase changes can be accurately retrieved [[18], [19], [20]]. In recent years, numerous studies and trials have focused on DAS, particularly in geophysics [[21], [22], [23], [24], [25], [26], [27], [28]].
To improve upon single-pulse DAS, a dual-pulse heterodyne demodulation DAS (HD-DAS) system was proposed, enabling the distributed retrieval of acoustic-induced phase changes [29]. In this system, a pair of pulses with a fixed frequency difference functioned as a moving interferometer along the sensing fiber, with phase information demodulated using a heterodyne detection algorithm. This design achieved an extremely low noise floor and a high signal-to-noise ratio, as the pulse pair shared noise properties while propagating along the same fiber [30].
The gauge length (G) of the DAS system was determined by the pulse width [31]. Since the HD-DAS system employed two pulses in the interrogator, its gauge length was calculated as:
In this calculation, τp = 20 ns is the pulse width, and τd = 30 ns is the time interval between the two pulses. c = 3 × 108 m·s−1 is the speed of light in vacuum, and m ≈ 1.5 is the refractive index of the optical fiber.
The spatial sampling interval (Δz) of the HD-DAS system was determined by the sampling rate of the data acquisition card and was calculated as:
where fADC = 100 MS·s−1 is the sampling rate of the data acquisition card. MS stands for mega samples (106 samples).
The sampling rate of the HD-DAS interrogator for each trace was 4 kHz (0.25 ms). During the calibration process, data were recorded at this rate to verify the system’s full-bandwidth performance. In the sea trial, seismic raw data were initially recorded at 4 kHz. A low-pass filter with a cutoff frequency of 500 Hz was then applied to the data to prevent aliasing. Following this, the data were down-sampled to 1 kHz (1 ms) to reduce the data volume for subsequent processing.
2.2. Distributed fiber-optic towed streamers
A distributed fiber-optic towed streamer cable was developed using HD-DAS technology. This technology enabled the detection of acoustic fields in a continuous and distributed manner using a single optical fiber as the sensing unit [32]. The optical fiber could function directly as the acoustic sensor or be helically wound around a specially designed sensitizing mandrel to enhance sensitivity [33]. Consequently, two types of distributed fiber-optic towed streamer cables were developed: the helically wound fiber distributed sensing towed streamer cable (HW-TSC) and the discrete-sensor towed streamer cable (DS-TSC).
Fig. 1(a) shows the schematic illustration of the HW-TSC. The mandrel, made of polyethylene, had an outer diameter of 15 mm and was used to wind a single-mode optical fiber at a helix angle of 25°. For a cable length of 1 m, the internal optical fiber was 1.1 m long, resulting in a physical gauge length of 7.27 m and a physical spatial sampling interval of 0.91 m for the HW-TSC. The inner sheath layer, made of polyurethane with an outer diameter of 22 mm, was followed by an aramid braided layer that enhanced load-bearing capacity and could withstand a pulling force of up to 1 t. Finally, an outer sheath layer made of polyethylene with an outer diameter of 35 mm provided additional protection.
We calibrated the acoustic pressure sensitivity of the HW-TSC in an anechoic tank. A 20 m segment of the cable was helically fixed to a cylindrical shelf with a diameter of 60 cm, as shown in Fig. 1(b). The head end of the cable was connected to the HD-DAS interrogator to collect the distributed acoustic signal. The shelf was then immersed in an anechoic pool to perform the calibration. Both the cable and acoustic source were positioned 5 m underwater, with the cable located 4 m away from the acoustic source.
During the calibration procedure, a sinusoidal acoustic signal was generated by the acoustic source, and the HD-DAS interrogator sampled data at a rate of 4 kHz. The calibration frequencies were set at 1/3-octave intervals [34], including 100, 125, 160, 200, 250, 315, 400, 500, 630, 800, 1000, 1250, 1600, and 2000 Hz. To avoid reverberation in the anechoic tank, a minimum calibration frequency of 100 Hz was established. A standard piezoelectric hydrophone was positioned along the connection line between the cable and acoustic source at a distance of 0.8 m from the acoustic source. The acoustic signal at 630 Hz collected by the cable was displayed in a distributed manner in Fig. 1(c). Fig. 1(d) presents a single-frequency (1600 Hz) pulse signal collected by the HW-TSC (blue curve) and standard piezoelectric hydrophone (red curve). These two signals exhibited significant consistency, except for the noticeable tailing in the acoustic signal of the cable. Furthermore, the tailing effect became more pronounced as the frequency of the acoustic wave decreased. This phenomenon may be attributed to the cable’s structure, which acted as a one-dimensional continuous medium causing acoustic waves to oscillate back and forth, resulting in a tail. Lower-frequency acoustic waves exhibited more pronounced tailing due to their lower damping. The calibration results for the acoustic pressure sensitivity of the HW-TSC are shown in Fig. 1(e). The average sensitivity of the HW-TSC was approximately −143.1 dB re rad·μPa−1 (referenced to 1 rad·μPa−1), with a maximum fluctuation of 12.6 dB within the measurement frequency band.
2.2.2. Discrete-sensor towed streamer cable
To address the issues of tailing and inconsistent sensitivity in the HW-TSC, we developed a solution in the form of the DS-TSC. The DS-TSC comprised several components, including connectors, aramid ropes, support skeletons, hydrophones, fillers, and polyurethane tubing, as illustrated in Fig. 2(a). To ensure load-bearing capacity, four aramid ropes attached to support skeletons were evenly arranged along the circumference, capable of withstanding up to 4 t of pulling force. The DS-TSC improved the consistency of acoustic pressure sensitivity by winding the optical fiber around separate elastic cylinders to create discrete hydrophones, as indicated by the black dashed ellipses in Fig. 2(a). The optical fiber between the two hydrophones remained in a free state. Unlike the fiber-optic interferometer hydrophone [17], the proposed hydrophone in the DS-TSC required no optical couplers or mirrors, simplifying the towed streamer cable structure and enabling cost-effective large-scale production.
We calibrated the acoustic pressure sensitivity of the hydrophone array in a standing wave tube. Each hydrophone was wound with 5 m of optical fiber, with 4 m of free fiber between adjacent hydrophones. This configuration allowed the fabrication of an array containing four hydrophones (H1 to H4) as shown in Fig. 2(a) using a single optical fiber, without the need for fusion-splicing to avoid fusion losses. The hydrophone array was connected to the HD-DAS interrogator via a 20 m long guiding fiber to retrieve the acoustic-induced phase signal.
We placed the hydrophone array in a standing wave tube to calibrate its acoustic pressure sensitivity. During the calibration, sinusoidal vibrations were applied to the standing wave tube, and the HD-DAS interrogator sampled data at a rate of 4 kHz. The calibration frequencies were set at 1/3-octave intervals [34], including 20, 25, 31.5, 40, 63, 80, 100, 125, 160, 200, 250, 315, 400, 500, 630, 800, and 1000 Hz. A standard piezoelectric hydrophone was also positioned in the standing wave tube to record the reference vibration signals.
The acoustic signal collected by the hydrophone array was shown in a distributed manner in Fig. 2(b), with the positions of the four hydrophones indicated by the blue arrow lines. The calibration results for the acoustic pressure sensitivity of the hydrophone array are presented in Fig. 2(c). The average sensitivity of the hydrophone array was approximately −157.8 dB re rad·μPa−1, with a maximum fluctuation of 4.5 dB within the measurement frequency band.
To achieve zero buoyancy of the DS-TSC, different fillers were employed to fill the polyurethane tube. Specifically, two types of DS-TSC were produced using distinct fillers: the gel-filled DS-TSC and the oil-filled DS-TSC. The outer diameter of the gel-filled DS-TSC was 45 mm, with polyurethane gel and foams used as the filler to achieve a density of 1.027 g·cm−3. Each hydrophone in the gel-filled DS-TSC was wound with 5 m of optical fiber. The spatial distance between two hydrophones in the cable was 6 m, while the length of the free optical fiber between two hydrophones was 7 m, ensuring that the optical fiber remained in a free state.
The gel-filled DS-TSC had a length of 160 m and consisted of 27 hydrophones, resulting in a channel spacing of 6 m and a total of 27 channels. Similarly, the outer diameter of the oil-filled DS-TSC was 45 mm, with light wax oil and foams used as the filler to achieve a density of 1.027 g·cm−3. Each hydrophone in the oil-filled DS-TSC was wound with 5 m of optical fiber. The spatial distance between two hydrophones in the cable was 2.5 m, while the length of the free optical fiber between two hydrophones was 2.85 m. The oil-filled DS-TSC had a length of 40 m and included a total of 16 hydrophones, resulting in a channel spacing of 2.5 m and a total of 16 channels.
3. Results and discussion
3.1. Sea trial for marine seismic exploration
We employed the distributed fiber-optic towed streamer cable system to conduct a sea trial in the gas hydrate province of the Qiongdongnan Basin, located on the northern continental slope of the South China Sea (18°N-19°N, 111°E-112°E). As shown in Fig. 3(a), the towed streamer consisted of six sections: guide cable, front vibration isolation cable, gel-filled DS-TSC, oil-filled DS-TSC, HW-TSC, and end vibration isolation cable. Detailed structural parameters of the towed streamer are provided in the table in Fig. 3(a). The guide cable was constructed by encasing an optical cable with a braided hair layer, which facilitated smooth water flow and reduced flow noise. The vibration isolation cable had a structure similar to the DS-TSC but lacked hydrophones, with the optical fiber maintained in a free state. Its polyurethane tubing was filled with light wax oil and foams, resulting in a density of 1.027 g·cm−3. This cable effectively isolated vibration noise transmitted along the towed streamer. The gel-filled DS-TSC, oil-filled DS-TSC, and HW-TSC were interconnected to form the sensing section, allowing the reception of acoustic wave signals reflected by seabed strata.
Flow noise significantly impacted the performance of towed streamers, primarily originating from two sources: flow-induced cable array vibrations and fluctuating pressure in the turbulent boundary layer. Cable vibrations, known as cable strumming, resulted from vortex shedding from the cable, which exerted a vibrating force on the towed streamer. Similarly, wake instability led to tail vibrations, commonly referred to as tail wagging [35]. These vibrations were partially dampened by employing vibration isolation cables. Furthermore, the impact of vessel noise on the towed streamer was minimized by extending the length of the guide cable.
During the towing process, the guide cable was released over a distance of 200 m, positioning the head of the sensing section 300 m away from the vessel’s stern. The sensing section was submerged underwater. A 2540 in3 (1 in3 ≈ 0.0164 L) Bolt air gun (Teledyne Marine, USA) served as the acoustic source, deployed at a sinking depth of 8 m and a distance of 92 m. The sea trial took place in the South China Sea along a survey line approximately 30 km in length. During the trial, the air gun was fired every 20 m, and the HD-DAS interrogator collected 6 s seismic signals for each shot.
The raw seismic data was initially recorded at a sampling rate of 4 kHz. A low-pass filter with a cutoff frequency of 500 Hz was applied to the data to prevent aliasing. After low-pass filtering, the data was down-sampled to 1 kHz to reduce the data volume for subsequent processing.
To establish the connection between the six cables shown in Fig. 3(a), their internal optical fibers were sequentially connected and linked to the HD-DAS interrogator to retrieve acoustic signals. The distributed sensing characteristic enabled the differentiation of the various cables in the spatial domain. Fig. 3(b) displays the acoustic signals collected by the HD-DAS interrogator at the towed streamer’s sensing section. According to the table in Fig. 3(a), the total length of the optical fiber in the sensing section was 900 m. The spatial sampling interval of the interrogator was 1 m (Eq. (2)), resulting in 900 traces within the sensing section, as shown in Fig. 3(b).
The acoustic signals retrieved from the gel-filled DS-TSC are represented by traces 1-335, while traces 336-460 correspond to the acoustic signals retrieved from the oil-filled DS-TSC. Traces 461-900 represent the acoustic signals from the HW-TSC. Notably, each hydrophone in the DS-TSC is wound with 5 m of optical fiber, with the black diamonds in Fig. 3(b) indicating the hydrophone positions. Each hydrophone corresponds to 5 traces. The fiber between two hydrophones remains free, exhibiting low sensitivity and producing weak acoustic signals. Therefore, for the DS-TSC, only the acoustic signals detected by the hydrophones were extracted.
3.2. Extract and optimize the distributed acoustic signals
As shown in Fig. 3(b), the HD-DAS interrogator captured distributed signals along the optical fiber. To retrieve the relevant data, the physical structure of the towed streamer cable was considered. Fig. 4(a) displays five traces collected by a hydrophone, showing spike-like noises (indicated by the red arrows), which represent phase fading noise within the DAS system [[36], [37], [38]]. Fading noise presents a significant challenge in DAS systems, as it severely impairs their ability to detect signals effectively.
To address this issue, the average value of the five traces was calculated in an initial attempt, as shown in Fig. 4(b). However, as depicted, fading noise remained prevalent. To mitigate this, the sort and average over trace (SAOT) algorithm was recently proposed to remove fading noise from distributed raw data [39]. In this study, the SAOT algorithm was employed to extract and optimize the acoustic signals.
First, the data from the five traces of a specific hydrophone is sorted in ascending order at a certain time:
where Hn(tr,t) represents the raw data of trace tr (tr = 1, 2, 3, 4, and 5) for the nth hydrophone at time t. Ĥn(tr,t) denotes the sorted data sort{·}, satisfying the condition Ĥn(1,t) ≤ Ĥn(2,t) ≤ Ĥn(3,t) ≤ Ĥn(4,t) ≤ Ĥn(5,t).
Next, the extracted signal for the nth hydrophone was determined as the mid-value Ĥn(3,t). Fig. 4(c) shows the recovered signals obtained after applying the phase fading noise elimination algorithm, which effectively eliminated the phase fading noises. Figs. 4(d) and (e) display the signals recovered from the gel-filled DS-TSC and oil-filled DS-TSC, respectively. Each trace in Figs. 4(d) and (e) corresponds to a hydrophone, resulting in a channel spacing of 6 m with 27 channels in Fig. 4(d) and a channel spacing of 2.5 m with 16 channels in Fig. 4(e).
The HW-TSC operated by continuous sensing in the spatial domain, in contrast to the DS-TSC. To process the acoustic signals from the HW-TSC, a sliding window with a window length of five traces and a step size of one trace was employed. The raw data were processed using the method described in Eq. (3). Fig. 4(f) illustrates the recovered signals obtained after applying the phase fading noise elimination algorithm, which effectively mitigated the phase fading noise. The HW-TSC had a channel spacing of approximately 0.91 m and encompassed a total of 435 channels.
Finally, Fig. 4(g) shows the power spectral density (PSD) of the seismic signals for the different streamer cables. The frequency bandwidth for all three streamer cables primarily ranged from 20 to 150 Hz. Within this range, slight variations were observed between the gel-filled DS-TSC and oil-filled DS-TSC, with the gel-filled DS-TSC exhibiting higher noise in the low-frequency band. In contrast, the HW-TSC demonstrated a weak PSD within the 20-150 Hz range due to its higher low-frequency noise.
3.3. Flow noise for different towed streamers
Firstly, the influence of vessel speed on flow noise was investigated. It is commonly observed that as vessel speed increases, flow noise gradually amplifies [35]. In the experiment, all streamer cables exhibited the lowest flow noise when the vessel’s speed was set at 4 kn (1 kn = 0.514444 m·s−1), compared to speeds of 3 and 5 kn, as shown in Fig. 5(a). Previous studies by Elboth et al.[40] identified an inclination threshold of 6° to 15° for hydrodynamic fluid noise. When the angle between the towed streamer and the water flow direction falls below this threshold, the fluid remains relatively stable, leading to reduced flow noise [40]. Based on these findings, it was determined that if the vessel’s speed is too low, certain factors contribute to increased flow noise. Firstly, the towed streamer lacks the necessary controllers, such as water birds. Additionally, the HW-TSC is heavier than seawater. These factors combine to create an excessively large inclination angle of the towed streamer in the water, ultimately increasing flow noise. Therefore, during the sea trial, the towing vessel maintained a speed of 4 kn.
Figs. 5(b)-(d) show the time-domain flow noise of the gel-filled DS-TSC, oil-filled DS-TSC, and HW-TSC, respectively, when the acoustic source was in a quiet state. The oil-filled DS-TSC exhibited the cleanest signal, indicating the lowest flow noise. Flow noise comprises various types, including swell noise, mechanical jerks caused by the irregular speed of the towing vessel, and oscillations resulting from the unstable attitude of the towed streamer in the water [16,41]. Fig. 5(a) illustrates that flow noise primarily resided in the low-frequency range below 100 Hz, particularly under 10 Hz. Dowling’s research [42] demonstrated that incorporating porous foam into liquid-filled streamers effectively attenuated flow noise, which explained why the oil-filled DS-TSC exhibited the lowest noise level. In contrast, the HW-TSC displayed the highest flow noise, approximately 30 dB louder than the oil-filled DS-TSC.
Furthermore, it exhibited pronounced noise power within the 1 to 10 Hz bandwidth. According to the table in Fig. 3(a), the density of the HW-TSC was greater than that of seawater, and no water bird was installed on the streamer cable. Consequently, the HW-TSC’s stability in the water was compromised, leading to significant swings during the towing process. In this case, the inclination angle of the towed streamer frequently exceeded the threshold range proposed by Elboth et al. [40], resulting in substantial flow noise.
Fig. 6 illustrates the frequency panels of a shot gather acquired from different streamer cables. The analysis revealed that the predominant noise was concentrated in the low-frequency range (< 20 Hz). For all streamer cables, clear reflected signals were observed in the frequency range from 20 to 150 Hz, while signals nearly disappeared at frequencies exceeding 150 Hz. The oil-filled DS-TSC exhibited lower noise levels and a higher signal-to-noise ratio across the full-frequency panels. In contrast, the HW-TSC showed linear noise in the low-frequency range, particularly within the 5-20 Hz band. A frequency-wavenumber (f-k) analysis of the HW-TSC was subsequently conducted (Fig. 7).
The analysis revealed two distinct speeds of linear noise: 1200 and 400 m·s−1. The 1200 m·s−1 noise corresponded to strumming or tugging noise generated by the longitudinal vibration of the streamer [43]. These longitudinal waves typically propagated along the streamer, with their propagation speed dependent on the material properties of the streamer. The 400 m·s−1 noise represented extensional waves within the hose wall [35]. Extensional waves caused pressure fluctuations in the inner material through Poisson coupling between longitudinal and radial displacements. The linear noises could be attenuated using the methods proposed in Ref. [43].
3.4. Stacked profile for different towed streamers
Table 1 presents the survey specifications. A standard processing workflow was employed to construct a stacked profile [44], including geometry definition, data sorting, noise attenuation, predictive deconvolution, velocity analysis, normal moveout (NMO) correction, horizontal stacking, and migration. Due to the varying channel spacing of the three towed streamers, three distinct 2D geometries were defined accordingly. As shown in Figs. 4(g) and 6, the effective reflected seismic signals predominantly fall within the frequency range of 20-150 Hz. Therefore, a band-pass filter (20-150 Hz) was applied to attenuate low-frequency flow noise and high-frequency interference.
Figs. 8(a)-(c) show the stacked profiles derived from the gel-filled DS-TSC, oil-filled DS-TSC, and HW-TSC, respectively. All three towed streamer cables effectively captured reflection and structural information of formation interfaces, revealing complex sedimentary processes in the basin’s slope zone, spanning water depths from 500 to 2000 m.
Notably, the stacked profile of the oil-filled DS-TSC provided the most impressive imaging results, despite its shorter length of only 40 m. It demonstrated resolution, frequency bandwidth, and signal-to-noise ratios comparable to conventional 2D towed streamer cable profiles in shallow seabed areas (< 500 ms). This meets the fundamental requirements for shallow seismic interpretation and highlights excellent cost-effectiveness in offshore natural gas hydrate exploration [3]. However, in deeper seabed areas (> 500 ms), the oil-filled DS-TSC’s profile exhibited limitations due to the shorter length of the towed cable. These limitations resulted in reduced shot-receiver offsets, weaker seismic energy, and a lower signal-to-noise ratio.
However, it still performed admirably by clearly identifying major reflection interfaces and enabling imaging of the subsurface down to 2000 ms beneath the seabed. Conversely, despite exhibiting higher flow noise, the HW-TSC demonstrated effective imaging of the shallow seabed (< 500 ms) due to its small channel spacing and large number of sensing channels. Future improvements in seismic profiles with the HW-TSC could be achieved by controlling the cable’s posture in the water to maintain stability and reduce flow noise. Fig. 8(d) shows the stacked profile obtained in a previous trial using a traditional electronic towed streamer with a length of 900 m and channel spacing of 12.5 m. The survey line in Fig. 8(d) is offset by approximately 300 m compared to the survey line conducted in this study. For comparison, Figs. 8(e)-(g) show results from the distributed fiber-optic towed streamer cables, which demonstrated superior resolution.
A comprehensive comparison of the three towed streamer cables is presented in Table 2. As shown in the table, the oil-filled DS-TSC, despite having a shorter length of 40 m and only 16 sensing channels, demonstrated superior performance. It successfully identified major reflection interfaces and enabled imaging of subsurface structures to a depth of 2000 ms beneath the seabed. Increasing the cable length and number of sensing channels is anticipated to improve seismic profiles and allow for greater penetration depths.
The structure of the gel-filled DS-TSC is similar to that of the oil-filled DS-TSC but exhibits significantly higher flow noise. Dowling’s study [42] conducted detailed numerical calculations based on Lighthill’s theory to characterize underwater flow noise. Simple algebraic expressions were derived to compare the performance of liquid-filled and visco-elastic-filled streamers. Incorporating porous foam into a liquid streamer was identified as an effective method for attenuating low wave-number disturbances [42]. This finding explains why the oil-filled DS-TSC demonstrates lower flow noise.
The HW-TSC exhibited poor performance due to its higher density than seawater, which led to an unstable posture during the towing process. However, the HW-TSC effectively imaged the shallow seabed (< 500 ms) due to its small channel spacing and a large number of sensing channels. This feature allowed for increased stacking times to generate the seismic profile. Furthermore, the HW-TSC’s simpler structure enabled manufacturing through an industrialized optical cable process, significantly reducing fabrication costs. Moreover, the HW-TSC facilitated fully distributed detection and provided a vast number of sensing channels, highlighting its potential as a promising option.
Several strategies have been proposed to improve the performance of the HW-TSC. Firstly, according to Dowling’s study [42], an oil-filled layer can be added outside the HW-TSC to reduce flow noise. Furthermore, setting the HW-TSC to zero buoyancy and incorporating water birds to control the cable’s posture in seawater have been suggested. These improvements can significantly enhance the streamer’s performance, meeting the requirements of marine seismic exploration.
4. Conclusions
This study demonstrated a marine towed streamer seismic acquisition system utilizing distributed acoustic sensing technology. The system development included three types of distributed fiber-optic towed streamer cables: gel-filled DS-TSC, oil-filled DS-TSC, and HW-TSC. These cables successfully captured the reflection and structural information of formation interfaces. During the sea trial conducted in the slope zone of the basin, with water depths ranging from 500 to 2000 m, all cables effectively captured complex sedimentary structures. The marine towed streamer seismic acquisition system presented in this study showed significant potential for fine marine seismic exploration, such as the exploration of natural gas hydrate reservoirs.
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This work was supported by the National Natural Science Foundation of China (62105007), the Key Program of Marine Economy Development Special Foundation of the Department of Natural Resources of Guangdong Province (GDNRC [2020] 045), and the Financial Support from China Geological Survey (DD20221703). We also acknowledge the Guangzhou Marine Geological Survey (GMGS) for providing the survey ship for the sea trials.
L.A.Levin, B.J.Bett, A.R.Gates, P.Heimbach, B.M.Howe, F.Janssen, et al. Global observing needs in the deep ocean. Front Mar Sci, 6 (2019), p. 241.
[2]
Y.Konno, T.Fujii, A.Sato, K.Akamine, M.Naiki, Y.Masuda, et al. Key findings of the World’s first offshore methane hydrate production test off the coast of Japan: toward future commercial production. Energy Fuels, 31 (3) (2017), pp. 2607-2616.
[3]
J.Ye, X.Qin, W.Xie, H.Lu, B.Ma, H.Qiu, et al. The second natural gas hydrate production test in the South China Sea. China Geol, 3 (2) (2020), pp. 197-209.
A.A.Vlasov, M.Y.Plotnikov, V.S.Lavrov, S.S.Kiselev, A.S.Aleinik. The influence of a method of bracing a fiber-optical seismic streamer during towing on the parameters of its output signal. Instrum Exp Tech, 63 (4) (2020), pp. 577-582.
[6]
Sercel. Seal 428—new generation marine seismic recorder [Internet]. Nantes: Sercel; [cited 2025 Feb12]. Available from: https://www.sercel.com/en/products/seal-428.
[7]
D.P.Vaughan-Hirsch, E.R.Phillips. Mid-Pleistocene thin-skinned glaciotectonic thrusting of the Aberdeen ground formation, central Graben Region, central North Sea. J Quat Sci, 32 (2) (2017), pp. 196-212.
[8]
A.M.Nikishin, E.I.Petrov, S.Cloetingh, A.V.Korniychuk, A.F.Morozov, O.V.Petrov, et al. Arctic ocean mega project: paper 1—data collection. Earth Sci Rev, 217 (2021), Article 103559.
[9]
T.Yamazu, K.Haraguchi, T.Inoue, K.Arai. Development of a highly flexible and high-resolution deep-towed streamer. J Mar Sci Eng, 7 (8) (2019), p. 254.
[10]
Y.Pei, M.Wen, L.Zhang, K.Yu, G.Kan, L.Zong, et al. Development of a high-resolution deep-towed multi-channel seismic exploration system: Kuiyang-ST2000. J Appl Geophys, 198 (2022), Article 104575.
[11]
F.Liu, S.Xie, M.Zhang, C.Tang, B.Xie, X.He, et al. Downhole microseismic monitoring using FOSS and its field test comparison with moving-coil geophone. IEEE Trans Geosci Remote Sens, 60 (2022), pp. 1-14.
NashPJ, CranchGA, HillDJ. Large-scale multiplexed fiberoptic arrays for geophysical applications. In: Proceedings of the Conference on Industrial Sensing Systems 2000; 2000 Nov 5-6; Boston, MA, USA. Bellingham: Society of Photo-Optical Instrumentation Engineers (SPIE); 2000. p. 55-65.
[14]
X.Hao, H.Zhang, C.Wei, C.Wang, H.Zhang. Sea trial for fiberoptic hydrophone array used in marine geophysical exploration. J Trop Oceanogr, 37 (2018), pp. 93-98.
[15]
H.Qian, X.Wang, X.Chen, Z.Yang. Research on noise suppression technology of marine optical fiber towed streamer seismic data based on ResUNet. Energies, 15 (9) (2022), p. 3362.
[16]
P.Y.Dmitrashchenko, M.Y.Plotnikov, V.S.Lavrov, A.V.Volkov, I.A.Sharkov, A.S.Godovova. Compensation of the towing noise of a fiber-optic streamer using an additional interferometer. J Opt Technol, 88 (9) (2021), pp. 532-535.
[17]
C.K.Kirkendall, A.Dandridge. Overview of high performance fibre-optic sensing. J Phys D, 37 (18) (2004), pp. R197-R216.
[18]
A.Masoudi, T.P.Newson. High spatial resolution distributed optical fiber dynamic strain sensor with enhanced frequency and strain resolution. Opt Lett, 42 (2) (2017), pp. 290-293.
H.M.Ogden, M.J.Murray, J.B.Murray, C.Kirkendall, B.Redding. Frequency multiplexed coherent ϕ-OTDR. Sci Rep, 11 (1) (2021), p. 17921.
[21]
T.M.Daley, B.M.Freifeld, J.Ajo-Franklin, S.Dou, R.Pevzner, V.Shulakova, et al. Field testing of fiber-optic distributed acoustic sensing (DAS) for subsurface seismic monitoring. Leading Edge, 32 (6) (2013), pp. 699-706.
[22]
T.M.Daley, D.E.Miller, K.Dodds, P.Cook, B.M.Freifeld. Field testing of modular borehole monitoring with simultaneous distributed acoustic sensing and geophone vertical seismic profiles at Citronelle, Alabama. Geophys Prospect, 64 (5) (2016), pp. 1318-1334.
[23]
P.Jousset, T.Reinsch, T.Ryberg, H.Blanck, A.Clarke, R.Aghayev, et al. Dynamic strain determination using fibre-optic cables allows imaging of seismological and structural features. Nat Commun, 9 (1) (2018), p. 2509.
[24]
A.Sladen, D.Rivet, J.P.Ampuero, L. DeBarros, Y.Hello, G.Calbris, et al. Distributed sensing of earthquakes and ocean-solid earth interactions on seafloor telecom cables. Nat Commun, 10 (1) (2019), p. 5777.
[25]
Y.Abukrat, M.Reshef. Distributed acoustic sensing for shallow seismic investigations and void detection. Geophysics, 86 (3) (2021), pp. WA89-WA97.
[26]
J.Li, W.Zhu, E.Biondi, Z.Zhan. Earthquake focal mechanisms with distributed acoustic sensing. Nat Commun, 14 (1) (2023), p. 4181.
[27]
Y.Abukrat, P.Sinitsyn, M.Reshef, A.Lellouch. Applications and limitations of distributed acoustic sensing in shallow seismic surveys and monitoring. Geophysics, 88 (6) (2023). WC1-12.
[28]
E.Cunningham, N.Lord, D.Fratta, A.Chavarria, C.Thurber, H.Wang. Three-dimensional distributed acoustic sensing at the sanford underground research facility. Geophysics, 88 (6) (2023), pp. WC209-WC220.
[29]
X.He, S.Xie, F.Liu, S.Cao, L.Gu, X.Zheng, et al. Multi-event waveform-retrieved distributed optical fiber acoustic sensor using dual-pulse heterodyne phase-sensitive OTDR. Opt Lett, 42 (3) (2017), pp. 442-445.
[30]
X.He, M.Zhang, L.Gu, S.Xie, F.Liu, H.Lu. Performance improvement of dual-pulse heterodyne distributed acoustic sensor for sound detection. Sensors, 20 (4) (2020), p. 999.
[31]
T.Dean, T.Cuny, A.H.Hartog. The effect of gauge length on axially incident P-waves measured using fibre optic distributed vibration sensing. Geophys Prospect, 65 (1) (2017), pp. 184-193.
[32]
B.G.Gorshkov, K.Yüksel, A.A.Fotiadi, M.Wuilpart, D.A.Korobko, A.A.Zhirnov, et al. Scientific applications of distributed acoustic sensing: state-of-the-art review and perspective. Sensors, 22 (3) (2022), p. 1033.
[33]
M.Li, H.Wang, G.Tao. Current and future applications of distributed acoustic sensing as a new reservoir geophysics tool. Open Petrol Eng J, 8 (1) (2015), pp. 272-281.
[34]
A.E.Isaev. The sensitivity and characteristic size of a hydrophone in the frequency band when there are sources of reflections. Meas Tech, 48 (12) (2005), pp. 1227-1234.
[35]
BeerensP, IjsselmuideS, VolwerkC, TrouvéE, DoisyY. Flow noise analysis of towed sonar arrays, Nice, France. In: UDT Europe 1999—Conference Proceedings Undersea Defence Technology; 1999 Jun 29-Jul 1; Nice, France. Swanley: TNO innovation for life; 1999. p. 392-7.
[36]
A.H.Hartog, L.B.Liokumovich, N.A.Ushakov, O.I.Kotov, T.Dean, T.Cuny, et al. The use of multi-frequency acquisition to significantly improve the quality of fibre-optic-distributed vibration sensing. Geophys Prospect, 66 (S1) (2018), pp. 192-202.
[37]
H.He, L.Yan, H.Qian, Y.Zhou, X.Zhang, B.Luo, et al. Suppression of the interference fading in phase-sensitive OTDR with phase-shift transform. J Lightwave Technol, 39 (1) (2021), pp. 295-302.
[38]
X.He, X.Xu, M.Zhang, S.Xie, F.Liu, L.Gu, et al. On the phase fading effect in the dual-pulse heterodyne demodulated distributed acoustic sensing system. Opt Express, 28 (22) (2020), pp. 33433-33447.
[39]
X.He, Z.Cao, P.Ji, L.Gu, S.Wei, B.Fan, et al. Eliminating the fading noise in distributed acoustic sensing data. IEEE Trans Geosci Remote Sens, 61 (2023), pp. 1-10.
[40]
T.Elboth, B.A.Reif, Ø.Andreassen. Flow and swell noise in marine seismic data. Geophysics, 74 (2) (2009), pp. Q17-Q25.
[41]
BevcD, NedorubO, HlebnikovV, ElbothT, VinjeV, GeliusLJ. Onboard de-noise processing for improving towed marine seismic acquisition efficiency. In: Proceedings of the SEG Technical Program Expanded Abstracts 2019; 2019 Sep 15-20; San Antonio, TX, USA. Tulsa: Society of Exploration Geophysicists; 2019. p. 47-51.
V.Hlebnikov, T.Elboth, V.Vinje, L.J.Gelius. Noise types and their attenuation in towed marine seismic: a tutorial. Geophysics, 86 (2) (2021), pp. W1-W19.
[44]
ChowdhuryKR. Seismic data acquisition and processing. In: GuptaHK, Encyclopedia of solid earthgeophysics, Cham: Springer International Publishing; 2020, p. 1-22.
PDF
(6906KB)
