The origin of nonvolcanic tremors has been linked to different ways the crust slips on a fault in the presence of fluids. We collected the tremors recorded by a seismic array in Nanao, Taiwan, and found a class of high-frequency tremors among ambient tremors and triggered tremors. The f-k analysis results and the large moveout implied that the triggered tremors originated from moving sources beneath the array. With seismogeodetic inversion, the triggered tremors occurred when dilation was larger than 10-8, similar to the deep low-frequency fluid-related seismic tremors in the Nankai trough. Based on previous tomographic studies, the partial melting due to the dehydration of the subducting Ryukyu slab is distributed along the slab edge and migrates upward near Nanao. Such a partial melting at shallow depth can potentially elevate the pore-fluid temperature and the pressure. According to these findings, we proposed a scenario analogous to fluid-related acoustic emission lab experiments. Dehydration near the edge of a subducting slab provides fluid and heat, and the dilational strains from the teleseismic waves cause decompression and therefore the fluid motions. Teleseismically perturbed fluids helped create at least some of the triggered tremors. This could be among the first field examples of dynamically triggered tremors associated with moving fluids from a slab and its edge.

Submarine landslides reshape seafloor geomorphology, transport sediment and carbon to the deep-sea, and can trigger tsunamis or damage valuable seabed infrastructure. Yet, submarine landslides have never been directly observed in action. Here, we present the most detailed measurements yet of an ongoing submarine landslide using high-frequency (1–10 Hz) ground motions captured by an ocean-bottom seismometer placed in the vicinity of the slide’s source region. We tracked a ~ 0.00423 km³ flank collapse in the Congo Canyon, and reveal its triggers and chronology of movement, thereby testing fundamental landslide models. A powerful turbidity current undercut the canyon wall, and landslide failure initiated at the base-of-slope. The landslide retrogressed 1.3 km upslope for 15 minutes in 1–2 min pulses, interspersed with brief downslope movement. Retrogression speeds ranged from 1.6 to 5.8 m/s, and downslope movement speeds from 0.8 to 3.3 m/s. Once the failure reached its upslope limit, the landslide transitioned into 1.5 hours of pulsed sediment transport, which fed a turbidity current along the canyon-axis. These observations show seismic monitoring can produce major advances in the understanding of submarine landslides. They challenge simple “toe-backward” and “head-forward” collapse models, revealing a hybrid, bottom-up retrogressive process that only released sediment after climbing far upslope.

Ubiquitous bottom simulating reflections (BSRs) on seismic profiles indicate the presence of gas hydrates and free gases in both the active and passive margins offshore SW Taiwan. Detailed seafloor temperature measurements and modeling were conducted in offshore SW Taiwan gas hydrate provinces to investigate the thermal structure under dynamic gas hydrate systems. We performed seismic analysis, direct temperature measurements, and BSRbased thermal modeling to understand the seafloor thermal structures and subsurface fluid flow systems. First, seismic interpretation was applied to understand the BSR distribution and structural features in the gas hydrate provinces along the convergent plate boundary offshore SW Taiwan. Then, we compiled the 159 in-situ temperature measurements in marine sediments, and a regional heat flow map was proposed. After that, we use the sub-bottom depths of the widespread BSRs to derive seafloor thermal structures offshore SW Taiwan. Finally, by comparing the measured and BSR-based temperature fields, several interesting geological processes affecting the seafloor thermal structures are revealed. The distinct heating effects near the lower slope of the accretionary prism indicate active fluid flow along the thrust and decollement systems. In the upper slope, the seafloor thermal structure shows an overall low with local higher heat flow near the diapiric structures, implying that the diapirs may serve as active fluid conduits. Although few large-scale fault systems exist in the passive South China Sea (SCS) Slope, the upward fluid migration along normal faults, dipping strata, and gas chimneys contribute to intense local-scale seafloor heating. Apparent thermal nonequilibrium observed near the deformation front reveals a recent mass transport deposit (MTD) event with a volume of similar to 0.8 km(3). Our results present the geologically controlled seafloor thermal structures under the dynamic gas hydrate systems, which may give insights into the subsurface fluid flow, gas hydrate systems, and submarine geohazards.

Abstract Turbidity currents carve Earth’s deepest canyons, form Earth’s largest sediment deposits, and break seabed telecommunications cables. Directly measuring turbidity currents is notoriously challenging due to their destructive impact on instruments within their path. This is especially the case for canyon-flushing flows that can travel >1000 km at >5 m/s, whose dynamics are poorly understood. We deployed ocean-bottom seismometers safely outside turbidity currents, and used emitted seismic signals to remotely monitor canyon-flushing events. By analyzing seismic power variations with distance and signal polarization, we distinguish signals generated by turbulence and sediment transport and document the evolving internal speed and structure of flows. Flow-fronts have dense near-bed layers comprising multiple surges with 5-to-30-minute durations, continuing for many hours. Fastest surges occur 30–60 minutes behind the flow-front, providing momentum that sustains flow-fronts for >1000 km. Our results highlight surging within dense near-bed layers as a key driver of turbidity currents’ long-distance runout.

Aquifer volumetric changes induce ground deformation. In hydrogeology, a constant-rate pumping test is used to characterize an aquifer system and its flow regimes. However, data only from one or two wells in a pumping site may lack detailed spatial information for a heterogeneous site. We propose that ground tilts detected by seismometers may provide additional spatial coverage. Here, we installed nine closely spaced broadband seismometers around a pumping well within 3-38 m to measure hydraulic-induced ground tilts during a 24-h constant rate pumping conducted in northeastern Taiwan. The tilts are overall consistent with an inverted cone shape analytical solution of ground vertical displacement due to water pressure perturbation. We found that ground subsidence, water table drawdown, vertical displacement, and ground tilt are linear with each other. However, inconsistent tilt directions show that local heterogeneities at each station affect the deformation pattern. The tilts at the stations near the pumping well were approximately oriented in a northwest-southeast direction, parallel to the dip directions of the fractures and the bedding plane. Tilt-estimated flow dimensions at seismic stations are spatially consistent with the drawdown-estimated flow dimensions at two wells. Preliminary hydrogeological surveys and seismic analyses show that this groundwater system combines sub-radial and spherical flow and has impermeable boundaries. The network is restricted by impermeable strata at greater depth, and the steep dipping fractures and the bedding plane. We demonstrate that the time series of tilts can be used to estimate flow dimensions at the tested site. Multiple closely spaced seismometers help to characterize details of the fractured groundwater network for constraining flow regimes and hydrogeological conditions.

We deployed nine broadband seismometers within 32 m of a well where a four-hour step-drawdown pumping test was conducted at a site near Shuangxi, Taiwan in August 2018. The whole seismic array simultaneously recorded hour-long seismic signals that we interpret as ground tilt evolution during and after the pumping. Here we document data from the six seismic stations within 10 m of the boreholes to show the ground deformation closest to the pump site. Instead of ground subsidence due to groundwater over-draw as documented in many long-term and even short-term GPS and tiltmeter studies, four out of the six seismometers show ground bulge at the pumping site with radial directional tilts up to 1.0 x 10-5 rad for a drawdown of 10 m. We interpret the bulge as the dominant transient crustal response to the unloading of discharged groundwater, outpacing the minor subsidence due to reduced effective stress from reduced pore fluid head. After the pumping stopped, the ground tilt signals started to level back, though not simultaneously: some signals recorded delays up to an hour, possibly related to replenishment of groundwater from far-field or different aquifers and shallow depth hydrological heterogeneity including minor faults and fractures. We conducted seismic reflection surveys and found mostly layered structures without large fault offsets. Using an analytical solution that calculates the ground tilt due to surface load changes, we calculated the effects of unloading on ground deformation. The predicted tilt signals are in the range of 10-5 rad, the same order as the seismically derived tilts from this study. For 10 m of drawdown we estimate 10-5 to10-4 m of vertical ground uplift for this particular site. Our study demonstrates that dense broadband seismic arrays may offer a relatively less expensive and non-invasive tool to study temporal and spatial evolution of ground motions, and thus the ground water flow patterns, during pumping tests.

Quantification of gas hydrates in marine sediments is crucial for understanding gas hydrate systems. By empirical relationships or effective medium modelling, gas hydrate concentrations can be derived from velocity and/or conductivity logs. However, these approaches do not take the co-occurrence of free gas and gas hydrate into account leading to large uncertainties in the calculated free gas and gas hydrate concentrations. To overcome this issue we adopt a joint elastic and electric self-consistent/differential effective medium model as the basis for a new joint inversion scheme that distinguishes between both phases. We apply this scheme to p-wave velocity and electric induction data measured by downhole-logging of boreholes at Formosa Ridge off Taiwan - a known hydrate province with an active gas conduit. Gaussian Mixture Modeling separates the background signal of the host medium from anomalies and allows to determine a background porosity as a probability density function of depth. We use this derived porosity to jointly invert electrical conductivity and velocity data for hydrate and free gas concentrations. At Formosa Ridge, we find two resistive anomalies, one in the shallow and another in the deep part of the borehole. Only the deep anomaly in conductivity coincides with a high-velocity anomaly. This is consistent with ∼30% hydrate with ∼1% free gas concentration. For the shallow anomaly, increased velocities due to hydrate concentrations of ∼15% are compensated by a decrease in velocity due to ∼1% of free gas. The method reconciles the different sensitivities of the two data types and yields hydrate and free gas concentrations that are largely consistent with geochemically derived values.

The presence of gas hydrates is well known in the marine realm southwest of Taiwan due to the widespread distribution of bottom simulating reflectors in seismic records. To learn more about gas hydrate systems and their dynamics at passive and active margins, we drilled boreholes with MARUM-MeBo200 seafloor drill rig at areas where geophysical indications for gas hydrates have been detected in the past. Several gas hydrate proxies like negative chloride anomalies in the pore water, cold spots detected by infrared thermal scans on cores, increased resistivity, and lithological parameters clearly showed the presence of hydrates in the drilled sections. However, gas hydrate was not recovered by MeBo most likely because of small crystal sizes which dissociated during re-covery from the seafloor. Three holes were drilled at southern summit of Formosa Ridge down to 126 m below seafloor (mbsf) and recovered sediments from which in situ hydrate saturation values were estimated between 1 and 10% at 15-42 mbsf and even higher values of up to 38% below 100 mbsf. The latter are probably related to the sealing effect of carbonate precipitation which occur at 85-95 mbsf directly above the hydrate-enriched layer. Four holes were drilled at Four-Way Closure Ridge where a maximum drilling depth of 143.90 mbsf was reached. Hydrate presence starts in 65 mbsf continuing down-core with a range of 1-29% gas hydrate saturation in fine-grained homogenous clay. An abrupt change to higher gas hydrate saturation values of up to 80% occurs below 109 mbsf where silty and sandy turbidite layers are often intercalated. Such high gas hydrate contents only occur in the sand layers and not in the fine-grained sediments intercalated to the sand deposits.

Abstract Tremors are a type of slow earthquake with long‐duration signals compared to ordinary earthquakes. The long signals have been considered to solely reflect their long source process. However, here, we provide evidence suggesting that the source processes of tremors are not always long. We refer to these observations as short‐duration tremors. They were recorded by ocean‐bottom seismometers placed very close to the source. Although these tremors exhibit a short‐duration signal when recorded near the source, they exhibit a typical long‐duration signal elsewhere. Our numerical simulations demonstrate that the features can be captured by considering a strongly scattering medium around their source. One such structure could be small low‐velocity inclusions distributed around the seismic source. The inclusions may represent the seismic expression of geologically detected aquifers in tremor source regions. Furthermore, this medium could be embedded along the slow earthquake fault zone and play a critical role in their source process.

The Formosa Ridge cold seep is among the first documented active seeps on the northern South China Sea passive margin slope. Although this system has been the focus of scientific studies for decades, the geological factors controlling gas release are not well understood due to a lack of constraints of the subsurface structure and seepage history. Here, we use high‐resolution 3D seismic data to image stratigraphic and structural relationships associated with fluid expulsion, which provide spatio‐temporal constraints on the gas hydrate system at depth and methane seepage at modern and paleo seafloors. Gas has accumulated beneath the base of gas hydrate stability to a critical thickness, causing hydraulic fracturing, propagation of a vertical gas conduit, and morphological features (mounds) at paleo‐seafloor horizons. These mounds record multiple distinct gas migration episodes between 300,000 and 127,000 years ago, separated by periods of dormancy. Episodic seepage still seems to occur at the present day, as evidenced by two separate fronts of ascending gas imaged within the conduit. We propose that episodic seepage is associated with enhanced seafloor sedimentation. The increasing overburden leads to an increase in effective horizontal stress that exceeds the gas pressure at the top of the gas reservoir. As a result, the conduit closes off until the gas reservoir is replenished to a new (greater) critical thickness to reopen hydraulic fractures. Our results provide intricate detail of long‐term methane flux through sub‐seabed seep systems, which is important for assessing its impact on seafloor and ocean biogeochemistry.

Supercritical flow bedforms are important elements of sedimentary environments, but their internal three-dimensional structure has been elusive due to seismic imaging limitations. This article presents high-resolution three-dimensional seismic reflection data from Formosa Ridge - a ridge between two canyons that incise into the north-eastern South China Sea margin. The ridge consists of 300 m thick submarine deposits including sediment waves that are manifested as crescentic depressions surrounded by elevated walls on the palaeo-seafloor. Cross-sectional profiles display scour fills that turn into step-like sediment waves further downstream. These bedforms are 470 to 1370 m long and 30 to 140 m high. The three-dimensional seismic data clearly show the step-like bedforms that may be misinterpreted as faults or slumps on data with lower resolution. Despite exhibiting negative palaeo-seafloor relief, they are overall depositional structures and have constructed at least part of the Formosa Ridge. The bedforms' morphology and upslope migration suggest that they are the continuum of partially depositional to fully depositional cyclic steps formed by bottom currents travelling, based on a series of simple calculations, at least 2 m s(-1). These currents are able to transport sediments with grain size up to coarse sand, and such dynamic processes might impact seafloor infrastructure safety, oil and gas reservoir systems, and the functioning of benthic ecosystems in similar settings.

Seafloor heat flow provides information about the thermal evolution of the lithosphere, the magnitude and timing of volcanic activity, and hydrothermal circulation patterns. In the central Gulf of California, the Guaymas Basin is part of a young marginal spreading rift system that experiences high sedimentation (1–5 km/Myr) and widespread magmatic intrusions in the axial troughs and the off‐axis regions. Heat flow variations record magmatic and sedimentary processes affecting the thermal evolution of the basin. Here, we present new seismic evidence of a widespread bottom‐simulating reflection (BSR) in the northwestern Guaymas Basin. Using the BSR depths and thermal conductivity measurements, we determine geothermal gradient and surface heat flow variations. The BSR‐derived heat flow values are less than the conductive lithospheric heat flow predictions for mid‐oceanic ridges. They suggest that high sedimentation (0.3–1 km/Myr) suppresses the lithospheric heat flow. In the central and southeastern regions of the basin, the BSR‐derived geothermal gradient increases as the intruded magmatic units reach shallower subsurface depths. Thermal modeling shows that recent (<5,000 years) igneous intrusions (<500 m below the seafloor) and associated fluid flow elevate the surface heat flow up to five times. BSR‐derived geothermal gradients correlate little with the depth of the shallowest magmatic emplacements to the north, where the intrusions have already cooled for some time, and the associated hydrothermal activity is about to shut down.

Abstract Characterizing properties of marine subsurface sediment helps with siting for offshore infrastructure. Shear‐wave velocity (Vs) provides information on the geotechnical properties of the seabed. We present our initial efforts to obtain a detailed two‐dimensional model of Vs for a large‐offset multi‐channel seismic (MCS) transect collected in shallow waters across the Taiwan Strait using surface waves excited by a large volume airgun. We derived the dispersion curves of the Scholte waves along the 37.5‐km‐long transect using the phase‐shift method and then conducted multimodal inversion to obtain a Vs model down to a depth of 150 m. To estimate the dynamic Poisson's ratio across the transect, we combined the Vs model with a compressional wave velocity model derived from the traditional MCS semblance velocity analysis. Lastly, we approximated the seismic attenuation of the profile. Our results show a large lateral variation in shear‐wave velocity. In the north, a low‐velocity zone with shear‐wave velocities of about 150 m/s was identified, while in the south, the shear‐wave velocity was found to be 300 m/s. With synthetic data, several sensitivity tests were performed to derive optimal parameters for offshore large‐offset streamer data. We particularly focused on the depth of the streamer and source and the water depth in combination with different seabed properties. Our results show that we can robustly derive the shear‐wave velocity, along with the Poisson's ratio, using large‐offset streamer data elsewhere based on the criteria we have tested using field and synthetic data sets.

Turbulent mixing in the deep ocean is not well understood. The breaking of internal waves on sloped seafloor topography can generate deep-sea turbulence. However, it is difficult to measure turbulence comprehensively due to its multi-scale processes, in addition to flow–flow and flow–topography interactions. Dense, high-resolution spatiotemporal coverage of observations may help shed light on turbulence evolution. Here, we present turbulence observations from four broadband ocean bottom seismometers (OBSs) and a 200-m vertical thermistor string (T-string) in a footprint of 1 × 1 km to characterize turbulence induced by internal waves at a depth of 3000 m on a Pacific continental slope. Correlating the OBS-calculated time derivative of kinetic energy and the T-string-calculated turbulent kinetic energy dissipation rate, we propose that the OBS-detected signals were induced by near-seafloor turbulence. Strong disturbances were detected during a typhoon period, suggesting large-scale inertial waves breaking with upslope transport speeds of 0.2–0.5 m s−1. Disturbances were mostly excited on the downslope side of the array where the internal waves from the Pacific Ocean broke initially and the turbulence oscillated between < 1 km small-scale ridges. Such small-scale topography caused varying turbulence-induced signals due to localized waves breaking. Arrayed OBSs can provide complementary observations to characterize deep-sea turbulence.

Earth and Space Science Open Archive This preprint has been submitted to and is under consideration at Journal of Geophysical Research - Solid Earth. ESSOAr is a venue for early communication or feedback before peer review. Data may be preliminary.Learn more about preprints preprintOpen AccessYou are viewing the latest version by default [v1]Dynamical triggering of tremors near Nanao, Taiwan, by the 2011 Tohoku earthquake: Slab-related fluid-induced seismicityAuthorsYi-HsuanWuiDWu-ChengChiiDLiamChaiiDChin-JenLiniDSee all authors Yi-Hsuan WuiDCorresponding Author• Submitting AuthorInstitute of Earth Sciences, Academia SinicaiDhttps://orcid.org/0000-0002-8151-0592view email addressThe email was not providedcopy email addressWu-Cheng ChiiDAcademia SinicaiDhttps://orcid.org/0000-0002-6111-6379view email addressThe email was not providedcopy email addressLiam ChaiiDUniversity of ChicagoiDhttps://orcid.org/0000-0003-3195-6980view email addressThe email was not providedcopy email addressChin-Jen LiniDInstitute of Earth Sciences, Academia Sinica, TaiwaniDhttps://orcid.org/0000-0003-4374-4256view email addressThe email was not providedcopy email address

Temperature is used to trace ocean density variations, and reveals internal waves and turbulent motions in the deep ocean, called ‘internal motions.’ Ambient temperature detected by geophysical differential pressure gauges (DPGs) may provide year-long, complementary observations. Here, we use data from four DPGs fixed on the ocean bottom and a high-resolution temperature sensor (T-sensor) 13 m above the seafloor as a square-kilometer array deployed offshore ~ 50 km east of Taiwan facing the open Pacific Ocean to examine the impact of temperature on DPG signals related to internal motions. The DPG signals correlate with T-sensor temperature variations between 0.002 and 0.1 mHz, but have time shifts partially caused by slow thermal conduction from the ambient seafloor to the DPG chamber and partially by internal motion propagation time across the array. Applying beamforming-frequency-wavenumber analysis and linear regression to the arrayed T-sensor and DPG data, we estimate the propagating slowness of the internal motions to be between 0.5 and 7.4 s m −1 from the northwest and northeast quadrants of the array. The thermal relaxation time of the DPGs is within 10 3 –10 4 s. This work shows that a systematic scan of DPG data at frequencies < 0.1 mHz may help shed light on patterns of internal wave propagation in the deep ocean, especially in multi-scale arrays.

Estimates of the sub‐seabed fluid flow rates are important for understanding hydrological budgets, biogeochemical cycles, and physical properties of the sediments. Fluid flow rates and directions, however, are difficult to measure, particularly beneath the seafloor. We developed a rapid method to estimate regional fluid migration rates using an extensive database of seismic reflection profiles taken offshore SW Taiwan. We observe bottom‐simulating reflector (BSR) that deflects toward the seafloor near thrust faults that indicate localized heat flow variations. At these sites, advecting warm pore fluids transport heat to shallower depths and force the BSR shallower. Our 2D steady‐state numerical method quantifies the fluid flow rates required to cause such thermal anomalies. We found that fluid flow rates near the trench of the accretionary wedge range between 0.1 and 16 m3 yr−1 m−1, with slower and faster rates generally associated with slope basin discontinuities and faults, respectively. To evaluate the fluid pattern evolution from subduction to collision, we studied three transects: one along the Manila subduction zone in the south and two in Taiwan's initial collision zone in the north. We quantified the fluid budget and partitioning of fluid flow between focused discharge through faults and diffusive flow through the wedge. Faults in Taiwan's accretionary wedge capture on average 25% of the total dewatering flux in the younger subduction zone and 38.5% in the tectonically mature collision zone. Our method provides estimates of fluid migration rates along convergent plate boundaries, and contributes to our understanding of focused fluid flow processes in many other regions.

Developing new ways to observe tsunami contributes to tsunami research. Tidal and deep-ocean gauges are typically used for coastal and offshore observations. Recently, tsunami-induced ground tilts offer a new possibility. The ground tilt signal accompanied by 2010 M-w 8.8 Chilean earthquake were observed at a tiltmeter network in Japan. However, tiltmeter stations are usually not as widely installed as broadband seismometers in other countries. Here, we studied broadband seismic records from Japan's F-net and found ground tilt signals consistent with previously published tilt meter dataset for this particular tsunamic event. Similar waveforms can also be found in broadband seismic networks in other countries, such as Taiwan, as well as an ocean bottom seismometer. We documented a consistent time sequence of evolving back-azimuth directions of the tsunami waves at different stages of tsunami propagation through beamforming-frequency-wavenumber analysis and particle-motion analysis; the outcomes are consistent with the tsunami propagation model provided by the Pacific Tsunami Warning Center. These results shown that dense broadband seismic networks can provide a useful complementary dataset, in addition to tiltmeter arrays and other networks, to study or even monitor tsunami propagation using arrayed methods.