Marine Geophysical Research

Online ISSN: 1573-0581
Recent publications
Article
The purpose of this study is to look into the effects of borehole washouts on log measurements and the resulting error in predicting gas hydrate saturation using well logs. We employ logging while drilling (LWD) data from the Indian National Gas Hydrate Program's second expedition (NGHP-02) in 2015. The NGHP-02 expedition discovered a significant amount of gas hydrate in coarse grain sediments in the Krishna Godavari (KG) Basin while drilling, coring, and logging. Borehole collapse or washout at particular depths in the presence of loose sediments impacted downhole log data at a few sites. We chose Holes NGHP-02-22A and NGHP-02-23A drilled in Area B of the KG Basin for our investigation and attempted to compensate washout effects in density-derived porosity, sonic and resistivity measurements, and the corresponding effects in estimating gas hydrate saturation. We use the sand-shale porosity model to remove the washout effects from density-derived porosity at washed-out depths. The corrected porosities and washout parameters are then used in rock physics theory to remove the washout effects from resistivity and velocity measurements by assuming washed-out zones as vertical fractures filled with seawater. We also estimate gas hydrate saturations from resistivity and velocity logs, taking into account both pore-filling and fracture-filling distributions. Analyzing velocity and resistivity logs jointly, we obtain fracture-filled porosity as 7.5% and 8% at Hole 02-22A and 02-23A respectively. Estimated saturation compared with that of the pressure core measurements show good correlation.

Article
We use unsupervised machine learning techniques aided by the Gaussian mixture model (GMM) for clustering downhole data of gas hydrate reservoir in the Andaman Sea, where drilling and coring were done at Site 17 in 2006 under the first expedition of the Indian National Gas Hydrate Programme (NGHP-01). Six different logging data (namely; density, neutron porosity, gamma-ray, resistivity, P- and S-wave velocity) are used in this study. We obtain six clusters using Davies-Bouldin index, Calinski-Harabasz index, Dunn index, dendrogram and self-organizing map, which are verified by high silhouette values. Data are then clustered using k-means, principal component analysis (PCA) and GMM. We notice that the k-means with random initialization gets biased towards the dominant principal component (gamma-ray), whereas, PCA shows each log has optimal weightage. Based on statistical analysis using 100 runs, GMM with the k-means initialization provides better results than GMM with random initialization. However, it provides three possible configurations of six clusters, which become stable when a combination of six logs is used as another input. Six clusters are interpreted in terms of lithology by histogram analysis of the corresponding log values. Lithology is found clay-dominated sediments with little silt and sand, as well as scattered volcanic ash, carbonate ooze, and pyrite, which is consistent with the lithology determined by smear slide and sieve data. Except at a few depths with higher concentrations (20–50%) in volcanic glass and carbonate ooze, gas hydrate occupies about 10% of the pore space in silty-clay sediments with sand and volcanic ash.

Article
Coral rock of biological origin exhibits distinctive mechanical behavior compared to that of conventional mineral rocks. Quasi-static and impact loading tests were conducted to investigate the mechanical properties of coral rock using a servo-hydraulic device and a split Hopkinson pressure bar (SHPB) apparatus. The dependence of the dynamic compressive strength, elastic modulus and energy absorption were quantitatively analyzed with respect to loading strain rate. The coral rock is characterized by a high porosity of 50.99 (± 5.60)% and a low bulk density of 1.31 (± 0.07) g cm⁻³ because of biological sedimentary. Both the quasi-static and dynamic stress–strain curves contain more pronounced compaction portion than dense rocks due to its special porous skeleton structure. The dynamic increase factor (DIF) of strength increases from 1.88 to 3.76 with the strain rate increasing from 117 to 334 s⁻¹, which could be expressed with a parabolic function. The dynamic elastic modulus exhibits an overall linear increasing trend with loading strain rate. The ratio of reflected energy to incident energy is positively related with the strain rate while the absorbed ratio is roughly negatively correlated with strain rate with a value less than 0.08. The specific energy absorption (SEA) is found to be proportional to the logarithm of the loading strain rate. It seems that the failure mode of coral rock under both quasi-static and dynamic compression is more dependent on inherent flaws like pre-existing pores, growth lines, weak bonding areas than loading strain rate. Nevertheless, the dominant failure mode in microscopic scale for coral rock at higher strain rate is still trans-granular fracture by means of SEM.

Article
Parameters viable for characterization of gas hydrates are mostly obtained from individual logs based on rock physics-elastic theories or electrical empirical relations. We demonstrate an approach for prediction of gas hydrate saturations using joint inversion of resistivity and sonic logs at Sites 22 and 23 of the second expedition of Indian National Gas Hydrate Program (NGHP-02), located in Area B of the Krishna-Godavari (KG) Basin. The strategy is based on a gradient descent optimization algorithm, where it combines the rock physics based elastic theories with the resistivity-based Archie’s empirical relations to ascertain the ambiguities associated with the evaluation of gas hydrate concentrations. It also assumes isotropic behaviour of gas hydrate reserves and incorporates saturation ranging from 0 to 100%, as a prior information to constrain the model. Results at two sites show a considerable match between the estimated (~ 30% at Site-02-22 and ~ 35% at Site-02-23) and pressure core derived gas hydrate saturations with a slight mismatch at certain depth intervals. The observed mismatch could be due to washout and anisotropic effects in the reservoir or uncertainty associated with the model parameters.

Article
A palynological analysis was carried out for the first time on sediments from Hupo Basin, East Sea, offshore Korea, to locate the Pliocene–Pleistocene boundary and thus determine the depositional age of this stratigraphic unit. Core 19ESDP-101, taken from Hupo Basin, yielded diverse, abundant to common pollen and dinocysts. Age-diagnostic palynomorphs were present in certain core intervals (Zone I, depths 120–63.96 mbsf). However, those age indicators were dark brown, heavily broken representatives that appeared together with poorly sorted, opaque, dark phytoclasts in the lower part of Zone II (63.96–38.76 mbsf), and they were considered to have been recycled from reworked late Pliocene strata due to contour currents during the transgression. Biostratigraphically meaningful taxa were the pollen Carya, Liquidambar, and Fagus and the dinocysts Filisphaera filifera subsp. pilosa and Spiniferites pachyderma. The latest stratigraphic occurrence of these pollen taxa in northeast Asia is the Late Pliocene, and that of the dinocysts is the Late Pliocene across a wide range of aquatic areas, especially in the Pacific. The last appearance datum of the age indicators in this study suggested 63.96 mbsf (top boundary of Zone I) as the Pliocene–Pleistocene boundary in core 19ESDP-101.

Article
Time- and depth-domain processing procedures were applied to seismic data acquired from shallow coastal areas using limited source-receiver offset ranges. A signal-to-noise ratio improvement strategy was applied in time-domain processing, and migration velocity analysis (MVA) with tomographic inversion was performed at every common image gather domain. Although the source-receiver offset was limited, the seismic image generated by the velocity from MVA represents more transparent reflections compared to the image from applied time processing only, which is likely because the shallow water depth of the survey area might yield a certain range of offsets. The processing procedure applied in this case also focused on the improvement of high-frequency content and removing as many multiples as possible in the time domain. Several small-scale faults and horizons of sedimentary units were well defined by this method. The improvement of the image with an accurate velocity by our method was demonstrated by representing the flattened events in the common image gather and velocity differences between time and depth processing. The final prestack depth migration image generated by our approaches provides refined information to interpreters that reduces uncertainties in geohazard assessments.

Article
Sediment geotechnical properties are an important component of geohazard analyses and studies on submarine slope failures on continental margins. The northern Cascadia margin is a region of repeated subduction earthquakes and related geohazards including devastating tsunamis. Despite several drilling campaigns at this margin mostly for gas hydrate studies, sediment geotechnical properties are lacking to date. We therefore investigated a set of 20 half-round samples from four drill sites of Integrated Ocean Drilling Program (IODP) Expedition 311 along a margin-perpendicular transect. Samples were analysed for Atterberg properties (liquid limit, plasticity limit, plasticity index) and consolidation state (pre-consolidation stress, over-consolidation ratio, initial void ratio and compression index). Results of these parameters are in good agreement with similar studies at the southern Cascadia margin from Ocean Drilling Program (ODP) Leg 204 and are regarded robust, despite the long time-span between sample recovery in 2005 and our analyses. Individual results are interpreted at each of the drill sites and reflect the overall tectonic state of the location within the accretionary prism. An interplay of site-specific tectonic forcing, including thrust faulting, uplift, folding, and erosion appears to govern the depth-dependent consolidation state of the sediments.

Article
The present study deals with mapping of the crustal structure over the Central Indian Ridge (CIR) covering 3°S–11°S using SGG-UGM-2 Free-air gravity data. We have discussed the importance of mantle temperature correction along with Bouguer, sediment thickness, and curvature corrections to the gravity data. We have discussed the influence of mantle temperature correction on gravimetric Moho determination along the ridge axis. We found an erroneous impression of Moho along the ridge axis from the gravity data without thermal correction. The Moho depth near the vicinity of CIR (3°S–16°S) is calculated from the residual mantle Bouguer gravity data using the Parker–Oldenburg algorithm. The inversion-based Moho shows an RMS error of 2.39 km with respect to the CRUST1.0 global Moho model. The isostatic Moho has been calculated using Airy-Heiskanen’s (AH’s) hypothesis to study compensation characteristics. The isostatic Moho exhibit a significant dissimilarity with the inversion-based Moho. The crustal distribution based on the inversion and isostatic-based Moho models were computed by removing the water column and sediment layer. The derived compensation percentage from these two crustal models indicates that the Somalian and Arabian basins are the zone of over-compensation, while an under-compensation zone characterizes the Madagascar basin. We categorized these crustal thicknesses into thin (< 6 km), normal (6–9 km), and thick (> 9 km) crust. Both thin and thick crusts show symmetrical and asymmetrical patterns around the CIR ridge axis. The fracture zone (FZ) and non-transform discontinuities (NTDs) are associated with a thin crust. It is inferred that FZ with a thick crust is perhaps resulted due to the process of serpentinization. We found that the Oceanic core complexes (OCCs) are associated with a thin crust. The OCCs associated with thick crust are perhaps related to the redevelopment of old OCCs near a slow-spreading ridge.

Article
The seismic stratigraphy and sedimentary facies of the Cretaceous and Paleogene sedimentary successions in north-eastern offshore part of Taranaki Basin, New Zealand have been investigated in order to unravel their depositional evolution and identify the potential hydrocarbon plays. Interpretation of regional seismic lines covering the entire shelf-slope and deepwater regions as well as integrating seismic and sedimentary facies allows the identification of several seismic-stratigraphic sequences within the studied successions. Early Cretaceous syn-rift successions (C1 sequence) were deposited in the structural lows near the present-day slope as swamp and marsh facies changing basinwards into turbidites and marine shales. The post-rift Cretaceous sequences (C2, C3 sequences) started with the progradation of the Taranaki delta (C2A-C2D units) followed by sedimentation of the transgressive C3A-E facies accumulated in response to thermal subsidence and high-rates of clastic supply. Extensive shelf conditions prevailed during the deposition of C2 and C3 Cretaceous sequences continued during the Paleogene with deposition of uniform marine shales throughout the basin. Rates of sediment supply substantially decreased during Oligocene prompting the deposition of marine micrite-rich carbonate. Variation in fossil content confirms the occurrence of several cycles of sea level fluctuations and episodic variations in terrigenous input. Organic-rich facies could be associated with the prograding clinoforms of the C2B unit and probably contain, basinward, large amount of gas and oil prone kerogen. The transgressive facies of C2D unit may also contain organic-rich shales basinward and coal near the present day Taranaki shelf. Potential reservoirs are hosted on the C3 transgressive estuarine sandstones. Paleogene mudstones are excellent regional seals for the hydrocarbons generated and trapped in the underlying Late Cretaceous facies.

Article
2D seismic reflection data from the western Barents Sea are used to map and date the Plio–Pleistocene shelf edge clinoforms. The mapped clinoforms are qualitative and quantitatively analyzed to estimate the water depths during their deposition (herein referred as paleo-water depth = PWD). Three different estimates are presented based on trajectory analysis, and relative water depth reconstructions at the upper clinoforms (PWD wd) and lower roll-over points (PWD Hh). A new quantitative PWD reconstruction approach is suggested which is based on relationships between clinoform parameters obtained from global analysis of clinoform systems. The results indicate PWD wd of ca. 680 m to 1160 m and PWD Hh between 1640 and 3200 m. A comparison of all results revealed discrepancies between the relative PWD and the trajectory analysis results. These discrepancies can be related to the complex interplay between thermo-tectonic driven differential subsidence and climate driven glacial/interglacial changes in sediment supply. This is most evident for the Pleistocene clinothems P5, P6, and P7 when the relief along the north western continental margin of the Basin was modified. From clinoforms P5 to S, ice extended onto the continental shelf and drove margin progradation with an enhanced sediment supply from the continental hinterland. Thus, waxing and waning of the Northern Hemisphere ice sheets and associated changes in the sediment source to sink behavior contribute significantly to the present day preserved clinoform shapes. If further tested, the reconstruction scheme presented in this study might help to assess, analyze and understand the PWD of continental margin clinoforms, worldwide.

Article
Some parts of the Yangsan Fault, a prominent mature intraplate fault on the Korean Peninsula, are still active. However, structural and paleoseismic investigations are limited because a large portion of the fault zone is covered by Quaternary sediments. To characterize the northern Yangsan Fault (NYF) and its paleoseismic features, we conducted topographic analyses, geological mapping, electrical resistivity surveys, borehole drilling, SHRIMP U–Pb age dating, a trench survey, and optically-stimulated-luminescence age dating (OSL). This multidisciplinary approach shows that the NYF is expressed as a nearly straight incised valley and a ridge-disrupting topographic lineament with anastomosing multiple core zones at outcrops. The NYF in Yeonghae area exhibits a NNE-striking eastern strand and a NNW- to N–S-striking western strand. Between these, a Jurassic granite (> 300-m-wide) is distributed as an enclosed lens, bounded by Precambrian metamorphic rocks to the east and Cretaceous sedimentary rocks to the west. The eastern strand likely passes S–N through the offshore area (East Sea) to onshore Pyeonghae area. A trench survey identified faults transecting Quaternary strata, providing the first paleoseismic record along the NYF. Stratigraphic features and OSL ages show that the most recent rupture occurred after 97 ± 7 ka, with the rupture along the western boundary of the mature fault core. Although older structures are prominent, paleoseismic records are few—a limitation for our onshore investigations. To reveal YF neotectonic activities under the East Sea, we need further information about off-fault damage (landslide, turbidite, tsunami records) within marine deposits as well as on-fault damage.

Article
Relocation of the ocean-bottom seismometer (OBS) is a basic step for subsequent inversion of stratigraphic structure. Obtaining sufficient and accurate position information is crucial to the final velocity model. We propose a new OBS relocation method using direct-wave arrival time information and accurate bathymetric data to provide a better starting of seismic velocity tomography. The new method consists of three steps. The first step is to determine the projection position of OBS on the shooting line according to the symmetry of the time-distance curve under the same yaw distance (the distance between ship’s position and the shooting line to the left or right). Next, the depth node closest to the observation time is found in the depth profile as the initial position of OBS in the direction perpendicular to the shooting line, based on the projection position. Finally, the final position of the OBS is determined by gradient grid search with the depth node as the center. In this paper, the feasibility of the proposed method is verified by experiments on model data. Then, the method is used to relocate 8 OBSs in the Southwest Pacific Ocean. The results show that the offset between the deployment position and relocation position is 50–500 m, with an average of 206 m.

Article
This paper presents interpretation examples of synthetic marine frequency domain CSEM responses from multiple reservoirs of known and realistic geologic models. The electric field responses were decomposed into constituent responses to examine individual reservoirs’ contributions to the EM field. Several realistic geological models of the reservoirs were created with all the structural complexities which are synonymous with the real-life scenario in the Niger Delta. The model has three reservoir layers, tagged 1st, 2nd and 3rd reservoir layers, in 2.5 km water depth at 0.25 Hz frequency. The result shows a clear Normalized Amplitude Ratio (NAR) for each resistive layer, with different detectability strengths of the layers. The most near-seafloor resistive layer, dubbed 1st reservoir layer, is marked by high NAR while the deeper resistive layer shows lower detectability. The value of NAR for each decomposed resistive layer reflects their depth of burial while the sum of individual NAR values correlates with the NAR obtained for the model containing all the resistors. It implies that the magnitude of NAR could be used to deduced the relative depth of burial of a resistive hydrocarbon layer as well as to detect the presence of multiple resistors. The decomposition of the resistive layers could be of help during real-time exploration surveys and suggest possible multiple hydrocarbon layers. The study also allows the understanding that the resultant response of two vertical reservoir layers is a summation of the strength of the individual layer, which differs from the seismic response of the same layers.

Article
In recent years, sparker source has gradually been applied for high-resolution seismic surveys. But the air gun is still the most commonly used seismic sources in marine seismic exploration. The seismic data frequency range of the air gun sources is below 200 Hz. On the contrary, the seismic data frequency range of the sparker sources is about 50–500 Hz. The low and high frequency components of the seismic data are both important for high resolution seismic exploration. Usually the energy produced by the air gun sources is stronger than that of the sparker sources, so the exploration depth of the air gun sources is bigger than that of the sparker sources. How to make full use of the two kinds of source to carry out high-resolution seismic exploration is a particularly meaningful work. Here the combined processing idea of the two kinds of sources’ towed streamer seismic data were presented and we designed a complete data processing workflow. The amplitude matching of air gun and sparker source seismic data is a very key technique in this combined processing. The main processing steps include conventional processing such as noise attenuation, amplitude compensation, wavelet processing, velocity analysis, pre-stack time migration and so on. But also, there are some special processing techniques such as the residual static corrections, CDP trim statics corrections and post-stack predictive deconvolution which are applied to the sparker source data. The results show that the migration section of the combined processing is better than those of the separately processing.

Article
The complex underwater environment and sonar parameters make the captured acoustic side scan sonar imagery to suffer from depleted contrast, low brightness, speckle noise, and deteriorated contour. Though the electromagnetic waves are highly absorbed in water and sonar is exemplary considered, these issues will affect the performance of the imaging Side Scan Sonar (SSS). Hence, these images need effective enhancement to achieve a privileged visual effect. The paper proposes the Retinex based Contrast-Enhanced Edge Preserved (RCEEP) technique to enhance the low-quality SSS image. Initially, the degraded image is convolved with a smoothing filter to obtain an illumination map. After the noise suppression, the reflectance map is computed and the brightness factor is interpolated. To rid of the blurred edges, the amended unsharp mask filter is applied to obtain the sharp-contour and smoothens the speckle noise. Finally, the contrast factor is weighted with a masked image to retain the contrast-enhanced sharpened image. The qualitative and quantitative analysis is carried out on the acoustic imagery. To evaluate each of the image attributes, the considered quantitative parameters are Root Mean Square Error (RMSE), Peak Signal to Noise Ratio (PSNR), Contrast Enhancement based contrast-changed Image Quality (CEIQ), Natural Scene Statistics (NSS), and Perceptual Sharpness Index (PSI). It is observed that the proposed RCEEP methodology enhances even the features in the dark region and outperforms the other state-of-the-art enhancement techniques.

Article
In estimating the geohazards posed by submarine landslides, the rheological properties of marine sediments are of significant importance for their postfailure dynamics. We report an experimental study of the rheological behavior of marine sediments taken from the Ulleung Basin, East Sea and their influence on numerical simulations of debris flow runout. Marine sediments exhibit a typical yield stress behavior, such as that of low-activity clays. For the materials examined, different yield stresses are observed depending on the shearing methods. Steady-state and oscillatory shear tests were conducted for different volumetric concentrations of sediment. According to the test results, the Bingham yield stresses under controlled shear stress and shear rate range from approximately 100 Pa to 1500 Pa, but the yield stresses under oscillatory shear loads range from approximately 25 Pa to 3500 Pa for a given sediment concentration. In the latter cases, the value obtained in the elastic region is approximately doubled. Experiments under steady-state and oscillatory shear loads can be helpful in determining the yield points in the elastic and viscous regions and in explaining changes in the structure of the soil sample due to shear. We apply the range of measured yield strength values to numerical simulations of debris flow runout using a Herschel-Bulkley model and find that only the lowest values of yield strength, despite the low sediment concentrations, could account for the observed runout and thickness distribution. We infer that significant wetting must occur during debris flow motion to attain the observed runout.

Article
Back-arc rifting at the middle eastern continental margin of the Korean Peninsula took place in the early stage of separation of the SW Japan Arc from mainland Asia in the Cenozoic. As a result, the margin preserves structural elements of rifting seawards that include a narrow rift (Hupo Basin), a major syn-rift border fault (Hupo Fault), and an uplifted rift flank (Hupo Bank). We correlated current seismicity at the margin with the rifting structures and estimated the potential seismic hazards. The Hupo Basin, in its overall shape, is a half-graben created by back-tilting of the hanging wall toward the Hupo Fault. The trace of the Hupo Fault is composed of multiple segments arcuate seawards, suggesting nucleation of rift faults at multiple places. The Hupo Fault is vertical at a shallow depth, but hypocentral locations of earthquakes indicate that the Hupo Fault is listric down to the upper and lower crustal boundary. Stochastic estimations of the maximum possible magnitude of earthquakes at the margin encompassing the Hupo Basin and the Hupo Bank average to ML 5.86. The estimated b-value of 0.63 is lower than that in the Korean Peninsula, which may suggest the dominance of thrust faulting with a strike-slip (or oblique) component on the listric and lower dipping portions of the faults at the margin.

Article
The southwestern continental margin of India reveals the presence of an anomalous bathymetric high feature located in the mid-continental slope region off Trivandrum. Based on the analysis of the available geophysical data and plate tectonic reconstruction, this feature was interpreted as a scar of India-Madagascar separation, with its conjugate identified on the Northern Madagascar Ridge. Although the conjugate nature of the Alleppey-Trivandrum Terrace Complex and the Northern Madagascar Ridge was postulated, this inference has not yet been evaluated by comparing their geophysical signatures and crustal structure. The present study is aimed to derive and compare the crustal configuration of these two features using an up-to-date compilation of the bathymetry, gravity and magnetic data, and by employing integrated forward modelling of gravity and magnetic anomalies. Our derived crustal models for the Alleppey-Trivandrum Terrace Complex and the Northern Madagascar Ridge suggest that both these features can be explained in terms of thinned continental crust intermingled with volcanic intrusives. The crustal thicknesses of the Northern Madagascar Ridge and Alleppey-Trivandrum Terrace Complex at their conjugate continent-ocean boundaries are ~17 km and both these features are associated with high amplitude magnetic anomalies whose genesis is attributed to the volcanism caused by the Marion hotspot activity. Therefore, based on our integrated interpretation of the geophysical data, we support the earlier interpretation on conjugate nature of the Northern Madagascar Ridge and the Alleppey-Trivandrum Terrace Complex that was proposed based on the fitting of shape and size of the bathymetric notch observed in the southeastern continental margin of Madagascar with a bathymetric protrusion observed in the southwestern continental margin of India in the India-Madagascar pre-drift scenario. These features remained as a single unit prior to ~88 Ma and subsequently got separated during the India-Madagascar breakup.

Article
Three-dimensional (3D) sub-bottom profilers (SBPs), which extend the data acquisition range of traditional single-channel SBPs, have been developed and utilized in various forms. However, owing to the fixed arrangement of the source and receivers, there are limitations on changing the survey specifications. In this study, we developed and field-tested a 3D SBP system that can be efficiently used in various fields, including the identification of buried objects and imaging of geological structures. This modular 3D SBP can arbitrarily adjust the type of seismic source and the number and interval of receiving channels, as demonstrated using a field-test in the waters near Ulsan, Korea. Using CHIRP as a seismic source, raw data with a sweep signal for a 1420 × 320 m² area was successfully acquired; a 3D seismic cube was created for the same area through subsequent 3D data processing based on 50 × 50 × 4.5 cm³ volumetric resolution. Using the 3D seismic cube, we could identify various geological features, including irregular seafloor, seafloor channels, sub-parallel stratification, and the top surface of free gas. The results confirm that this method can be utilized in geological study for various engineering purposes.

Article
Vertical cable seismic (VCS) is a reflection seismic exploration technique, which is mainly used for structural imaging in high dip angle areas. Because the source and receiver points are usually not in the same plane, it is not possible to use conventional velocity analysis to obtain the velocity field. In most cases, the velocity field of the streamer data is processed in the same survey area as the VCS. Seismic interferometry is to obtain new seismic signals by cross-correlation or convolution operation of seismic signals received by different receivers. Therefore, we propose to apply seismic interferometry to VCS exploration. Compared with conventional VCS data processing, this method does not need towed streamer data and improves exploration efficiency. In this paper, the method is applied to model data and actual data of South China Sea respectively to obtain the stacking profile. The results show that this method is applied to VCS data, and the stacking profile is continuous in phase axis and clear in structure.

Article
Continental shelves around the globe are hosts to vast reservoirs of offshore freshened groundwater. These systems show considerable complexity, often as a function of the geological heterogeneity. Data needed to characterise these systems are often sparse, and numerical models rely on generalized simplifications of the geological environment. In order to improve our understanding of these systems, it is necessary to implement modelling approaches that can produce large-scale geologically representative models using sparse data. We present an interdisciplinary stochastic modelling workflow incorporating borehole data, 2D depth-migrated seismic profiles, seismic attributes, and prior knowledge of the depositional setting. We generate a conditioned Gaussian field of porosity on the New Jersey shelf. We also perform a petrophysical conversion to a corresponding permeability distribution. The model dimensions are 134 km × 69 km × 1.7 km, with an adjustable resolution that can be adapted for process-based models of flow and solute transport. The integrated approach successfully translates small-scale porosity variations to a shelf-scale model that honors key characteristics of the New Jersey shelf wave-dominated depositional environment. The model was generated using open-source packages. All data and code to reproduce the complete workflow are provided along with this study so the model can be reproduced at any resolution for further studies of continental shelf processes offshore New Jersey.

Article
We studied the Quaternary incised fills drilled at the northern Yangsan Fault having multiple deformation histories since Late Cretaceous or Paleogene to determine tectonic influence on development of incised valley and its sedimentation. Incised valley fills were deposited during and after the Last Glacial Maximum and are composed of fluvial lag, debris flow deposits interbedded with fluvial sediments, shallow marine sandy deposits, and fluvial sediments from bottom to top. These fills show lateral changes in sediment thickness from 44 to 11.5 m over a short distance of 230 m, implying sediment stacking in a deep and steeply inclined valley. Fluvial lag and debris flow deposits are common in the thalweg of a valley. Despite small drainage basin (195.9 km²), the development of deep incised valley is interpreted to have resulted from fluvial downcutting on erodible basement during sea level fall as a consequence of dense development and fault and fracture networks in the pre-Quaternary rocks caused by multiple movements of Yangsan Fault. With steep gradient, the damaged rocks led to frequent slope failure and forceful accumulation of debris flow deposits on the valley’s axis at the time. In addition, stacking of debris flow deposits resulted in decrease of longitudinal gradient of incised valley, promoting rapid transgression during sea level rise (9 to 7 ka). This resulted in insufficient time for the central basin mud to be accumulated, which explains why the studied fills lack central estuarine mud that is common in incised valleys fills deposited during transgression.

Article
Analysis of high-resolution seismic profiles from the SE continental shelf of Korea reveals that the shelf sequence consists of four sedimentary units, namely, Pliocene deposits (III), Lower Pleistocene deposits (IIB), Upper Pleistocene deposits (IIA), and Holocene deposits (I), which were mainly controlled by regional tectonic activity and sea-level fluctuations. Vertically, with the exception of unit I, the units form a series of superimposed prograding wedges that thicken seaward. A structural high, including several faults and folds, occurs north of Tsushima Island, extending northeastward to the Dolgorae Thrust Belt at the southern margin of the Ulleung Basin. In this study, we deduced that the deformed zone was formed due to compressional deformation associated with back-arc closure that occurred after the extension of the East Sea. Most of the faults and folds, oriented in NE–SW or NNE–SSW trends, mainly developed in the lower two units (III and IIB). The lower two units were also significantly deformed, whereas the two overlying units remained relatively undeformed. This indicated that, during the Pliocene and Lower Pleistocene (units III and IIB), sedimentation was mainly controlled by tectonic activities, whereas sedimentation after the Lower Pleistocene (units IIA and I) was mainly influenced by periodically repeated sea-level changes rather than tectonic controls.

Article
Acoustic data collected by multibeam echosounders (MBES) are increasingly used for high resolution seabed mapping. The relationships between substrate properties and the acoustic response of the seafloor depends on the acoustic angle of incidence and the operating frequency of the sonar, and these dependencies can be analysed for discrimination of benthic substrates or habitats. An outstanding challenge for angular MBES mapping at a high spatial resolution is discontinuity; acoustic data are seldom represented at a full range of incidence angles across an entire survey area, hindering continuous spatial mapping. Given quantifiable relationships between MBES data at various incidence angles and frequencies, we propose to use multiple imputation to achieve complete estimates of angular MBES data over full survey extents at a high spatial resolution for seabed mapping. The primary goals of this study are (i) to evaluate the effectiveness of multiple imputation for producing accurate estimates of angular backscatter intensity and substrate penetration information, and (ii) to evaluate the usefulness of imputed angular data for benthic habitat and substrate mapping at a high spatial resolution. Using a multi-frequency case study, acoustic soundings were first aggregated to homogenous seabed units at a high spatial resolution via image segmentation. The effectiveness and limitations of imputation were explored in this context by simulating various amounts of missing angular data, and results suggested that a substantial proportion of missing measurements (> 40%) could be imputed with little error using Multiple Imputation by Chained Equations (MICE). The usefulness of imputed angular data for seabed mapping was then evaluated empirically by using MICE to generate multiple stochastic versions of a dataset with missing angular measurements. The complete, imputed datasets were used to model the distribution of substrate properties observed from ground-truth samples using Random Forest and neural networks. Model results were pooled for continuous spatial prediction and estimates of confidence were derived to reflect uncertainty resulting from multiple imputations. In addition to enabling continuous spatial prediction, the high-resolution imputed angular models performed favourably compared to broader segmentations or non-angular data.

Article
In ocean-bottom node (OBN) seismic exploration, a ghost is a common interference wave that affects the accuracy of seismic data interpretation. Receiver de-ghosting can be achieved using dual-sensor summation technology, which employs a hydrophone and geophone to collect seismic signals. The differences between the two receivers cause the polarities of the ghost wave signals to be opposite; therefore, the ghost waves can be eliminated by adding these receivers. However, there are differences between the actual data obtained from the hydrophone and geophone with regard to frequency, phase, and amplitude, thereby preventing them from being directly summated. Therefore, the frequency, phase and amplitude of both data records must be matched for consistency before dual-sensor summation can be conducted. In addition, some noise and ghosts will remain during data processing, resulting in a reduction in the signal-to-noise ratio of the data, making it necessary to adopt noise and residual ghost suppression methods. In this study, a wavelet analysis was newly introduced to the dual-sensor summation process. Specifically, the wavelet spectrum whitening method was proposed for the frequency matching of dual-sensor data, and the nonlinear wavelet transform threshold method of the wavelet denoising method was applied to suppress the noise and residual ghost. On this basis, a new dual-sensor process flow in OBN seismic exploration was developed. The feasibility and effectiveness of the method were verified using actual data. The method proposed in this study will help to improve the accuracy of future data processing.

Article
The study investigates sea-level measurements observed from 12 tide gauge sites over the Indian coastal area during the last two decades. Initially, the rise of sea-level (slope) and acceleration is estimated by fitting the linear and parabolic equations in the recorded data from tide gauge measurements. The estimated results show the abrupt change of relative sea-level measurements at the distinct site depends upon the tide gauge site locations. To accurately analyse the regional coastal sea level pattern, the tide gauge time series data are decomposed in sine and cosine functions at different frequencies followed by the spectral analysis. Analysis of the results confirm that the tide gauges peaks do not occur at the fixed period at each tide gauge site, instead they repeat with varying time periods. The spectrum peaks width fluctuate at distinct sites and the general pattern of frequency spectrum does not follow a unique model. Such type of characteristic variation with the time is possibly because of the effective variables, which affects the steadiness of sea-level changes. The study concludes that the experimental results from the Indian coastal region must be included during the comparison of global data sets and other contemporary oceanic models.

Article
Seismic feature characterization in the time domain remains a challenge for geoscientists, thus motivating transformation to the frequency domain. This study exposes an unreported buildup through a frequency domain transformation of seismic sections using constrained least squares spectral analysis for spectral decomposition. The imaged buildup appears at an anomalous low frequency of 5 Hz and vanishes at frequencies higher than 10 Hz and is traced in time domain. The geological context and feature morphology suggest that the buildup seems to be a reef. Further, quantitative scoring using regional, geophysical and geological criteria suggest a possible isolated carbonate buildup. This suggests a future exploration prospect and opens the scope for further geophysical investigation, developing conceptual models and practical evaluation.

Article
Analysis of anisotropy from velocity data is essential for improving the hydrocarbon reservoir characterization. The anisotropy of a medium is affected by the mechanical strength, presence of fracture, mineral distribution of the rock, and its degree affects the seismic velocity. We attempted to characterize the anisotropy of the gas hydrate bearing sediments in the offshore Mahanadi basin using three wells. Initially, the presence of anisotropy was investigated by estimating the stiffness coefficients and Thomsen’s parameters (epsilon, gamma and delta) assuming a horizontal transversely medium using dipole S-wave (upper and lower) velocities. The natural fractures were identified from the formation image data. The strong anisotropy is associated with the presence of natural fractures and lower values of the elastic modulus. Most of the strong and weak anisotropy zones are oriented in the NW to W direction of the study area. Our study suggests that the anisotropy in gas hydrate bearing sediment is stress-induced due to the presence of pore filling fractures, and the change of mechanical behavior. The higher positive values of epsilon and delta with gamma represent either dry solid gas hydrate or free gas filled in the fracture of the sediments as observed in the crossplot analysis. Finally, we modeled P-wave and S-wave velocities by incorporating the Thomsen’s parameters. S-wave velocity is less effective than P-wave velocity at 90° angle of fracture relative to the symmetry axis and the modeled P-wave velocity increases upto 2.8% in the gas hydrate bearing sediments.

Article
Coastal dynamics are the result of several processes controlling the balance between sediment input and output over time. The beach system is not always able to maintain a neutral coastal balance due to natural and anthropogenic causes. We present an integrated marine geology, geomorphological and sea-level rise analysis in the coastal sector between Torre delle Ciavole and Capo Calavà (North-Eastern Sicily, Italy).This sector is characterized by high uplift rates and frequent seismicity (mainly generated by the very active Vulcano-Tindari Fault System), promoting the development of mass-wasting processes in the coastal and offshore sectors. A main erosive feature observed in the area is the head of the Gioiosa Marea submarine canyon, located at some meters of depth, few hundred meters far the coastline. The main morphological features of the canyon were reconstructed through the analysis of high-resolution multibeam data, indicating that the canyon is active, as also testified by the comparison of time-lapse aerial photos. Due to this active setting, the study area is exposed to multiple geohazards, among which we deal with: (1) retrogressive instability at the head of the Gioiosa Marea submarine canyon, (2) coastal erosion favored by the downlope funnelling of littoral drift at the canyon head, (3) flooding scenario at 2100 using the IPCC (Intergovernmental Panel on Climate Change) and Rahmstorf sea-level projections. The consequences associated with these geohazards are amplified by the strong anthropization pressures occurring along in this sector. Our results provide key insights regarding the future scenarios of this coastal sector, revealing the effects of the retrogressive activity associated with the canyon head on the coastal strip. We also present the first management tool for the application of forecasting studies by local administrations.

Article
Surface Related Multiple Elimination (SRME) usually suffers the issue of either over-attenuation that damages the primaries or under-attenuation that leaves strong residual multiples. This dilemma happens commonly when SRME is combined with least-squares subtraction. Here we introduce a more sophisticated subtraction approach that facilitates better separation of multiples from primaries. Curvelet-domain subtraction transforms both the data and the multiple model into the curvelet domain, where different frequency bands (scales) and event directions (orientations) are represented by a finite number of curvelet coefficients. When combined with adaptive subtraction in the time–space domain, this method can handle model prediction errors to achieve effective subtraction. We demonstrate this method on two 2D surveys from the TAiwan Integrated GEodynamics Research (TAIGER) project. With a careful parameter determination flow, our result shows curvelet-domain subtraction outperforms least-squares subtraction in all geological settings. We also present one failed case where specific geological condition hinders proper multiple subtraction. We further demonstrate that even for data acquired with short cables, curvelet-domain subtraction can still provide better results than least-squares subtraction. We recommend this method as the standard processing flow for multi-channel seismic data.

Article
CryoSat-2 was launched successfully in April 2010 and thus has long exceeded its design life, so it is necessary to assess the reliability of gravity anomalies derived from CryoSat-2 altimeter data beyond its nominal lifetime. CryoSat-2 altimeter data in four different periods is used to derive 1′ × 1′ gravity anomalies within the Bay of Bengal (BOB) and its adjoining areas (80°–100°E, 0°–23°N). The accuracies of gravity anomalies in different periods are assessed in comparison with processed ship-borne gravity anomalies, revealing that the gravity anomalies derived from CryoSat-2 altimetry data in different periods have similar accuracies. The amount of geodetic mission data acquired by the CryoSat-2 altimeter exceeds that recorded by all other altimeters. Thus, we determine the gravity anomalies within the BOB and adjoining areas by using data from January 2011 to May 2020 to evaluate the gravity anomaly derivation performance of all available CryoSat-2 altimeter data, and establish the 1′ × 1′ BOBGA1.0 marine gravity model. In this work, the performance of BOBGA1.0 at different distances from the coast and at different water depths is evaluated by comparisons with ship-borne gravity anomalies and the SIO V28.1 marine gravity model. The role of CryoSat-2 in deriving gravity anomalies is further assessed by a comparison with other satellite altimeters. We conclude that CryoSat-2 altimetry data can reliably be used to derive gravity anomalies and that high-resolution and high-accuracy CryoSat-2 altimetry data play a major role in gravity anomaly derivation worldwide.

Article
Increases of water turbidity and suspended sediment transport in submarine canyons have been associated with high-energy events such as storms, river floods and dense shelf water cascading (DSWC), and occasionally with bottom trawling along canyon flanks and rims. To assess the variations on the water column turbidity and sediment transport in the Palamós Canyon linked to both natural and trawling-induced processes, an autonomous hydrographic profiler, as well as a near-bottom current meter and a turbidimeter were deployed in the canyon axis (929 m depth) from February to June 2017, covering a trawling closure (February) and trawling activities (March-June). Periods of enhanced water turbidity during the trawling closure were mostly associated with storms and DSWC events, transporting turbid dense waters into the canyon. In absence of such events, the water column displayed low suspended sediment concentrations (~ 0.3 mg L− 1) until the trawling season began, when particulate matter detachments, ranging between > 1 mg L− 1 and 3.8 mg L− 1, were observed at the water depths where the trawling grounds are found. During the trawling closure, high near-bottom suspended sediment fluxes (35–44 g m− 2 s− 1) were sporadically registered at ~ 920 m depth associated with a major storm and DSWC event. Smaller but more frequent increases of near-bottom suspended sediment fluxes (0.1–1.4 g m− 2 s− 1) were recorded during trawling activities. Despite these smaller trawling-induced suspended sediment fluxes, 30 days of continuous bottom trawling activity transported a total amount of 40 kg m− 2, of similar magnitude to that generated by a major DSWC event (50 kg m− 2). Since bottom trawling in Palamós Canyon is practiced on a daily basis throughout the year, a much larger contribution of anthropogenically derived water turbidity and suspended sediment transport can be expected.

Article
We approach the problem of identifying the stratigraphic markers using seismic data in automated mode. This automation would be achieved through the application of correlation exponent and its double derivative for the first time. Prominent contrasts derived from correlation exponents from seismic trace are associated with the different stratigraphic layers that help to isolate these layers. Further explicit analysis of double derivative of correlation exponent aids to segregate the un-compacted and compacted sediments. The feasibility and reliability of the proposed technique were tested over the seismic traces of Taranaki Basin. 48 seismic traces traversing the various geological features identified by earlier researchers in the field were selected. The combined interpretation of the anomalous value of correlation exponent and their double derivative for the seismic traces indicate the prominent geological features: progradation surfaces, bright spots, sedimentary packages and faults. This study presents a tool for automated seismic interpretation and reservoir characterization. The overall success in identification of various features by correlation exponent is about 77.3%.

Article
Regional seismic data with deep wells information has been used to interpret the depositional setting of Miocene sediments in the Offshore Indus Basin, Pakistan. The current study has implications for understanding the complex subsurface geology of the Offshore Indus Basin, Pakistan. Two seismic profiles were selected to interpret the sedimentary packages within the Miocene sediments. The interpreted seismic data showed four different recognizable system tracts in the Miocene sediments. These include Lowstand, Highstand and Falling Stage System Tracts. The sedimentary packages differ in regional setup depending on the location within the offshore basin. The eastern part (as highlighted by seismic Profile-2 and Profile-3) exhibits more complex sedimentary geometries in comparison to western part (as highlighted by seismic Profile-1). The earlier deposition is controlled by lowstand settings when the sea-level dropped below the shelf margin leading to incision over the shelf. This led to the deposition of Lowstand fan systems. This was followed by a rise in sea-level leading the way for the carbonate deposition over the shelf. An intermittent unconformity is identified by another set of lowstand fans and truncation patterns. This unconformity halted the carbonate deposition over the shelf and transported the sediments to the basin floor. Later, rise in sea-level resumed the carbonate deposition over the shelf. Another sea-level drop is noticed and sediments are interpreted to be moving towards the basin leading to a Falling Stage System Tract and easily identified by regressive onlaps. This unconformity is more pronounced towards east where it eroded the carbonates. Downlaps seen above this unconformity confirm the deepening conditions leading the way for the deposition of finer sediments. Further drop in sea-level is seen leading the way for coarse sediment deposition. We believe that this study would provide strong support for researching global passive margins from economic perspective as well as understanding complex depositional framework.

Article
The seismic data interpretation reveals properties of the subsurface structures, appropriate for further geological or exploration investigations. The buried channels are desired exploratory targets for petroleum reservoir exploration, drilling concerns or geological investigations. Therefore, it is required to precisely identify the geometry of the buried channels through the signal image analysis and the geological interpretation of the seismic data. The continuous wavelet transform (CWT) is a spectral discrete transform method for non-stationary signal analysis, useful to enhance the resolution of the seismic data both in time and frequency domains. In the presented study, we propose a strategy for a precise identification of buried channels, using a vast number of time–frequency seismic attributes on 3D seismic data. The spectral decomposition by the CWT method has been performed on 20, 25, 30, 35, and 40 Hz frequency sections. The seismic attributes have been extracted, including the energy, the instantaneous amplitude, the average root mean square, the Prewitt and Grubbs filters, the similarity, the semblance, and the grey-level covariance matrix. The most favorable results have been selected to be integrated, so providing new attributes through RGB composition. This integration has been performed both for single frequency and multi-frequency seismic sections. As a result of applying the proposed strategy on the selected synthetic and field dataset, the boundaries of the buried channel as well as its inner structure, have been individuated. The final model of the seismic attributes has revealed that the proposed strategy represents an alternative to the conventional procedure of identification the buried channel.

Article
The study investigates the real-time sea-level measurement from six tide gauge stations at the coastal area of Japan during the period 2017 to 2019. Initially, the six co-located global navigation satellite station (GNSS) sites vertical velocity has been observed followed by estimation of the relative tide gauge velocities. The results showed that the estimated relative sea-level changes abruptly at different sites depending on the location of tide gauge stations. To accurately analyse the regional tide gauge patterns at each site, a continuous wavelet transforms (CWT) based on Morse wavelet as a default transformation has been applied. Analysis showed that the peaks of tide gauges have not occurred at a static time at each site, rather they were repeating with the variation of time. The width of the spectrum peaks varied with different sites and the overall spectrum frequency was not unique. These characteristics variation with time may be due to the variables that affect the stability of the sea level measurements. Finally, the kernel extreme learning machine (KELM) approach based on variational mode decomposition is used to reconstruct the tide gauge measurements. The correlation coefficients between observed and reconstructed data sets are very high varying from 0.91 to 0.96. The estimated root mean square error between two data sets is varying from 0.04 m to 0.06 m at the entire region. This means that the KELM reconstructed tide gauge showed good agreement with the observed tidal waveforms. Thus, it can be concluded that the proposed techniques can be implemented for effective real-time sea-level monitoring changes with the available data.

Article
The utilization of sea areas adjacent to land has rapidly increased in recent years due to the expansion of economic activities from the land to the ocean. To secure the economic feasibility and safety of coastal development, technologies are required that can be used to image the seafloor and shallow subsurface structures at high resolution. For example, ultra-high-resolution (UHR) seismic surveys have improved the spatiotemporal resolution of seismic surveys. Since 2014, the Korea Institute of Geoscience and Mineral Resources has been working on the development of a three-dimensional (3D) UHR seismic survey system that can be operated with a small 10-ton vessel (EOS-Streamer). In this study, the composition of the EOS-Streamer and its individual components are described in detail. To validate the system, the first field demonstration was conducted offshore Pohang in southeastern Korea in 2019. A 9.77-ton small vessel and two 10 in³ air guns were used for the acquisition of field data and a 3D cube with a bin size of 2.5 m was obtained at the postprocessing stage. Continuity was maintained in the 3D cube in both the inline and crossline directions. Various geological structures and fault planes were observed close to the seafloor. The results of this study confirm that economical 3D seismic surveys in coastal areas, which are difficult to conduct using large vessels, can be carried out with the EOS-Streamer system.

Article
Seafloor acoustic reflectivity is fundamental for both underwater detection and ocean acoustic field prediction, especially in shallow-water offshore regions. This paper presents a procedure to estimate the seafloor reflectivity in shallow water based on the direct wave and seafloor reflection data from single-channel seismic records of sparker sources. We apply the procedure to a seismic line acquired in the western part of the Taiwan Strait. To resolve the uncertainty in the inversion results, we implement strict quality control on the data obtained from seismic records and utilized for the inversion calculation. According to the preprocessing results, it is recommended to exclude the bubble pulses and their surface reflections, which are unstable and act as noise, from the estimated source wavelet applied to the inversion calculation. The calculation results show a significant level of directional variation in the acoustic energy radiated from the sparker source, and this variation should be considered in the calculation of seafloor reflectivity. In this work, the seafloor reflectivity is calibrated with relatively stable calculated values that correspond to known types of sediment, and the directivity constant of a sparker source is estimated (≈ 0.2), which is defined as the amplitude ratio between horizontally propagating seismic waves and vertically propagating seismic waves. The resulting relationship between the seafloor reflectivity and the sediment type is consistent with those of previous research, indicating the feasibility of our procedure.

Article
We have studied the electric E→\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\vec {E}$$\end{document} and magnetic B→\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\vec {B}$$\end{document} fields by displaying the arrival directions through a 2D image named spectral-directionalogram (SD-gram). This technique maps electromagnetic field directions in time and frequency space. We tested the method through modeling experiments. An exclusive case is explained theoretically. In addition, method is applied on the two field data sets. One was acquired on the land and other in the ocean. The land data, is studied for the 50 Hz and harmonics. The frequencies are continuous in the direction during a time range, while splittings are observed during the other. In the second example, we have used marine electromagnetic data. The spectrogram suggests two anomalies, one close to 1 Hz and the other having a broad spectral range between 1 to 0.08 Hz. We hypothesize two possible causative sources, microseisms and degassing of a mud volcano. Out of these two choices, one can be easily falsified using the SD-gram. These examples highlight the usefulness of the technique in the data analysis. Moreover, the electromagnetic noise caused by the ocean waves have different spectral damping characteristics compared to those of plane waves. Therefore, we are proposing a new damping relation for the ocean, where dispersion is a dominant case.

Article
Seamount ages are important for understanding crust-mantle interactions and exploring sea bottom ore resources. Rock sampling and laboratory measurements are time-consuming and expensive, and are currently the main methods for dating seamounts. Thus, the ages of many seamounts in the Pacific Ocean are still unknown. To address this problem, gravity anomalies, magnetic anomalies, oceanic crustal ages, sediment thicknesses and seamount heights are chosen as the input parameters for seamount age prediction based on the potential relationship between geophysical observables and seamount ages. A Back-propagation (BP) neural network is constructed using the currently known seamount ages in the Pacific Ocean. Then, the weight and threshold of the BP network are optimized by a Genetic algorithm (GA); finally, the GA-BP model for seamount age production is derived. In addition, Convolutional neural network (CNN), BP model, and Support vector regression (SVR) methods are also used to predict seamount ages. The uncertainties in the prediction results decrease in the order of the GA-BP model, CNN, BP model, and SVR methods. The RMSE of the GA-BP prediction results is 10.22 Ma, and R2 is 0.90.

Article
Geothermal heat flow, as a parameter characterizing the Earth's thermal state, records deep thermodynamic processes. However, measurements of heat flow (HF) in the oceanic crust are relatively sparse and susceptible to surface activity such as hydrothermal circulation. We propose a machine learning approach to predict marine HF (MHF). We apply the random forest (RF) regression algorithm to train predictors capable of mapping multiple geological and geophysical features to MHF, thus enabling HF prediction in the global oceanic crust. We generate three data sets with different qualities of HF measurements for training. The best predictor has an accuracy of ~ 0.13 (normalized root mean squared error) or ~ 0.07 (normalized mean absolute error), and the predicted global oceanic crust HF map reflects the basic pattern of the MHF distribution. We find by comparison that the quality of HF measurement affects the prediction results. Then, we use a cross-prediction scheme to screen out the "underestimated" measured HF cases, which are mostly located in tectonic environments such as mid-ocean ridges and back-arc basins and show high spatial correlation with hydrothermal circulation. Furthermore, we conduct experimental calculations for the extent and proportion of underestimated cases in the oceanic part of a new global heat flow (NGHF) database; for example, the proportion of records with the degree of underestimation greater than 50% is approximately 30.8%. These calculations can provide reference information for the selection and application of MHF records in the NGHF database.

Article
This paper presents an investigation describing the potential future earthquake areas and the subsurface structures in the Marmara region (NW Turkey) by using aeromagnetic data. The region locates in the entrance of the North Anatolian Fault Zone (NAFZ) into the Marmara Sea and it has a crucial importance due to its seismological past. Firstly, the aeromagnetic dataset has been processed by using the power spectrum analysis. This analysis shows that the average depths of causative sources in the region range from 0.39 to 18.52 km. Secondly, to emphasize the magnetic anomalies caused by shallow sources, high-pass filter has been applied to the data. Finally, the Second vertical derivative method has been implemented to high-pass filtered data for delineating the exact location and trends of the subsurface structures. This final map clearly shows the boundaries of the causative sources. Anomalies align in the E-W direction from Izmit to Sapanca while the other anomalies located around Duzce elongate in the NE-SW direction. Moreover, the second vertical derivative map has shown that horsts in the region have no magnetization, and the magnetic anomalies become weak along the E-W direction in the north of the Central Ridge Horst. This can be associated with the existence of the horsts that are older than the magnetic material surrounding them. In addition, these horsts act as a barrier to the NAFZ and the NAFZ cannot pass through the region easily and the NAFZ has been separated into sub-branches towards the west in this region. As a result, this study provides valuable insights about the subsurface structures in the region and it plays a key role to define future earthquake areas.

Article
Seismic reflection data were used to investigate the multiple stages of headward erosion, incision, and sedimentation on the initiation and evolution of the Central Canyon, a Late Miocene-Pliocene continental slope-parallel canyon on the northwestern margin of the South China Sea. Secondary channels of the early stage generally display deep, V-shaped cross-sectional morphologies, indicating that vertical erosion was dominant in the early stage of canyon formation, while later-phase secondary channels have shallow, U-shaped cross sections, indicating that sedimentation dominated in the late stage. Influenced by gradual sea-level rise, headward erosion shifted the Central Canyon head from the Qiongdongnan Basin to the southeastern margin of the Yinggehai Basin. The length of headward erosion is up to 140 km, equivalent to one-fourth of the total length of the Central Canyon (525 km). The thalweg depth, height, and width of the Central Canyon increases down canyon from 2723 m, 164 m, and 2 km to 5318 m, 1023 m, and 16 km, respectively. This increase is gradual on gently-sloping palaeoseafloor, and more abrupt on steep-gradient palaeoseafloor. Canyon segments on gently-sloping palaeoseafloor are also characterised by a good correlation between canyon height and width. The sediments fill in the Central Canyon differ across the canyon’s segments. Turbidites and debrites infill the upstream segment of the Central Canyon. Turbidites and Mass-transport deposits are widespread in the middle reaches of the Central Canyon. The typical feature of the Central Canyon’s downstream segment is under-compensation, and canyon wall collapse are common in the upper fill of the downstream segment. We reconstruct the multistage evolution of the Central Canyon as comprising primary erosion, burial, renewed erosion, infilling, and abandonment stages. This multistage evolution was mainly controlled by relative sea-level fluctuations of the Qiongdongnan Basin.

Article
Diatremes and volcanoes have been described largely from rock exposures in magma-rich sedimentary basins. A growing number of research works however describe volcanoes and similar structures (hydrothermal vent complexes and fluid-escape structures) from seismic reflection data. Nevertheless, the detailed timing, composition and dimensions of diatremes and volcanoes from seismic reflection data are sparsely known owing to paucity of direct measurements from wellbores, cores, and limited resolution power of seismic reflection data. In this work, high quality 3-D seismic reflection data and seismic attributes are used to interpret the origin and configuration of diatremes buried beneath 40 paleo-volcanoes. The diatremes are mostly reflected as faulted zones of low amplitude and distorted seismic signal that are depicted as composite columnar, conical, fault-controlled, and bifurcated structures. Above the diatremes are cone- and crater-type volcanoes, which are infilled with syn-eruptive deposits. Volcanoes in the study area, are common in two main domains, representing the boundaries of isolated centres of (a) eruptive volcanoes, and (b) strong material fragmentation and dispersion, common with diatreme volcanoes on the paleo-seafloor. The interaction of magma and fluids at deeper depths caused piercement of the overburden rocks, creating the diatremes, which later acted as conduits for the migration of magma and fluids onto the Palaeocene–Eocene seafloor. Magma and fluid plumbing through the diatremes were principally aided by the complex interactions of pre-existing faults and diatremes. The ensued eruptions predominantly produced small-volume, monogenetic volcanoes (< 1 km3) which are dispersed along the flanks of a NE–SW oriented volcanic field where large-volume, polygenetic volcanoes (> 1 km3) are prevalent. The complex internal configuration of subsurface diatremes and their volcanoes as demonstrated in this work further reinforces the diversities and the caveats in their seismic interpretation along many magma-rich margins.

Article
We analyzed original and published seismic data to investigate the distribution and characteristics of subsurface fluid flow features in the gas hydrate provinces along the northeastern South China Sea (SCS) continental slope, including Jiulong Ridge, Horseshoe Ridge, Pointer Ridge, and Formosa Ridge. Numerous features indicating the presence/migration of hydrocarbons, such as bottom simulating reflections (BSRs), bright spots, gas chimneys, and fluid flow through faults, are identified. The results reveal that the hydrofracturing-induced gas chimneys act as the primary conduits for the overall focused fluid flow in the study area, though fluid flow along permeable faults is also observed at Pointer Ridge. Thirty-three gas chimneys identified are categorized into two types depending on their capability for focused fluid flow migrating into the gas hydrate stability zone (GHSZ). The type-I chimneys that transport fluids into the GHSZ contribute to gas hydrate formation and may even lead to seafloor seepage. Buried by a considerable thickness of sediments, the type-II chimneys cannot directly contribute to generating gas hydrate and surface seepage. Our results suggest that the sedimentary processes and fluid accumulation significantly control the development of gas chimneys in the study area. Since the focused fluid flow conduits that indicate overpressured fluids are critical pathways that feed gas into the GHSZ and form gas hydrates, utmost attention should be paid to them during hydrocarbon exploration.

Article
Depressurization technique is considered as one of the most promising techniques in dissociating gas hydrates. However, since dissociation of hydrates is an endothermic process. Dissociation alone through depressurization is not a feasible technique due to limited heat transfer. The reduced heat transfer results in rapid cooling, thereby causing reduced permeability due to ice formation and re-formation of hydrates. The objective of the current study is to investigate the viability of depressurization under worst case scenario of suppressed heat transfer. The worst case scenario is simulated by employing Newman boundary of no heat flux from the surroundings. The novelty of the present work lies in investigating the gas production behavior using depressurization in a worst case scenario. For this purpose, a 2D model is applied for a 150 m × 150 m system. A production well is placed at the center of the domain. The depressurization is performed by the withdrawal of fluids from the production well. In order to determine the suitable depressurization rate, the withdrawal of fluids is carried out within a range of 0.01–0.6 kg/s. The overall cumulative production at the well (mass of CH4) is determined. In this study, we demonstrate that three major causes, namely ice formation, secondary hydrates and reservoir achieving steady state are responsible for stopping of gas production. Insights into the dissociation behavior of the cases analysed are obtained from the contours of gas, water, hydrate, pressure, equilibrium pressure, temperature, relative gas permeability, and relative water permeability.

Article
The particle breakage of calcareous sands plays a key role in determining the particle shape and shear strength. This process has been analysed in the current study by employing the ring shear tests on calcareous sands sampled from the South China Sea. In this study, a set of parameters, e.g. roundness and aspect ratio, have been employed to quantify the geometrical properties of calcareous sands pre- and after the breakage. These two parameters have been found to correlate well with the vertical loading stress level, shear strain and shear strength. The particle breakage leads to the change of micro-structure during the shear test, as illustrated by the Scanning Electron Microscopy (SEM) images. It is found that during the shearing deformation, particle breakages can effectively reduce, while particle rearrangements can increase the shear strength of the calcareous sands.

Article
The northern South China Sea (SCS) plate between Taiwan and the Philippines experienced complex underplating and overriding. Ocean-bottom seismometer (OBS) data along the southern profile (20.5 °N) with 18 OBSs, the NE-SW profile with 11 OBSs and the northern profile (21.5 °N) with 12 OBSs were collected from a large-offset active-source seismic survey during the 2009 Taiwan Integrated Geodynamics Research. Here we re-analyzed the OBS data to image the transitional continental crust (TCC) and the subducted slab across the northern SCS, the Manila Trench, the deformation front (DF) and the accretionary wedge off southwestern Taiwan. By picking both refracted and reflected arrivals from OBS data and adopting irregular grids during modeling, we have derived P-wave velocities and interfaces of the crustal structures to identify the characteristics of the crustal structures and the collision-subduction transitional zone in the northern SCS. Based on the lateral velocity variation (2.5–5 km/s) of the accretionary wedge along three OBS profiles, the geometry of the splay faults on top of a décollement was imaged. Brittle deformation due to the hardened sediment with a relatively high velocity (3.5–5.5 km/s) from the collision was also found between a mid-crust detachment and the décollement along the NE-SW profile. Our study indicates the deeper Moho in the south (15–18 km deep) but shallower in the north (14–17 km deep), and the Moho is generally shallower than that derived from the previous studies. West of the trench or the DF, the TCC in the northern SCS has been identified in shallow crustal layers while underplated by the thick oceanic crust or the oceanic crust from the SCS in the middle and lower crustal layers. East of the trench or the DF, the TCC in the upper and middle crust has subducted beneath the accretionary wedge. We found a failed rift zone west of the DF along the northern profile. One seismically active region with large lateral velocity change (2.5–4.5 km/s) in the accretionary wedge was imaged. Another seismically active region is near the southern end of the Kaoping Canyon where the crustal thickness from the basement to the Moho is extremely thin (8 km).

Top-cited authors
• University of Lisbon
• +Atlantic CoLAB
• University of Lisbon