Zijian Zhang

Conventional high-resolution marine seismic surveys often face challenges in detecting small features in the shallow subsurface because acquisition parameters involve trade-offs among spatial resolution, penetration, signal-to-noise ratio (SNR), and operational efficiency. This study presents an intelligent marine seismic acquisition framework that integrates selectable acquisition hardware with modified ML Real domain adaptation, deep learning reconstruction, quantitative quality-control (QC) evaluation, and a human-supervised adaptive agent. Rather than directly predicting acquisition parameters with the neural network, the deep learning model reconstructs high-resolution seismic images from degraded acquisitions, which are subsequently evaluated using QC metrics including reflector continuity, spectral bandwidth, SNR, and depth penetration. The adaptive agent uses these metrics to rank discrete, hardware-feasible configurations of source type, source energy, dominant frequency, and shot interval and recommends the configuration for the subsequent acquisition cycle. Synthetic benchmark tests demonstrate improved imaging of thin layers and localized subsurface features under the tested conditions. A two-cycle nearshore field experiment further demonstrates the operational feasibility of the workflow, in which QC evaluation of the initial acquisition guided the subsequent acquisition configuration and resulted in improved reflector continuity, penetration, and SNR. The results show that integrating AI-assisted numerical modeling and evaluation can guide acquisition-parameter refinement and reduce reliance on repeated offshore trial-and-error surveys. The proposed framework provides a proof-of-concept for integrating configurable high-resolution seismic hardware with AI-assisted, QC-guided acquisition optimization for marine shallow-subsurface imaging.

Seismic data reveal that the shelf edge of the Pearl River Mouth Basin in the northern South China Sea is characterized by slope channels that have consistently migrated in a north-easterly direction over millions of years. Previous research suggests that the channel migration is driven by the interplay between along-slope bottom currents and downslope turbidity currents. Here, we propose an alternative interpretation, suggesting the migrating channels are actually a series of channel–levee systems and the migration is driven by their own evolution of erosion–deposition under the influence of the Coriolis force. A detailed interpretation of high-resolution seismic data reveals seven types of architectural elements, characteristic of channel–levee systems, which are erosional bases, outer levees, inner levees, channel-axis fills, marginal slumps, drapes, and lobes. An analysis of the sequence stratigraphy and stacking pattern of channels suggests that channel migration from the middle Miocene to the present is discontinuous with at least three regional discontinuities within the channel migration sequence marked by regional drapes. Down-dipping reflections along the margin of channels, previously interpreted as bottom-currents deposits, are here reinterpreted as mass-transport processes along steep channel walls. The migration is most prominent in the middle reach, where erosion and deposition coexist and dominate alternately in two different phases. During the long-term canyon-filling turbidity currents prevailing phase, deposition dominates, leading to the development of a prominent asymmetric right-hand (west) inner levee due to the Coriolis force. In contrast, during the canyon-flushing turbidity currents prevailing phase, erosion dominates and the preferred right-hand (west) inner levee enforces the flow to erode eastward, then drives the channel migrating eastward. The alternating effects of erosion and deposition ultimately result in unidirectional channel migration.

The Yin’e Basin, lying in the west of the Inner Mongolia Autonomous Region of China, embodies a complex geologic setting involving diverse and deformed strata of sedimentary and igneous rocks. Two-dimensional seismic data, collected in the western portion of the basin for the frontier exploration of oil and gas, are highly complicated with low signal-to-noise ratio and interference of reflected and diffracted waves. Seismic imaging using conventional technologies was poor, and the subsurface structures and strata were unreliable for interpretation. We have applied an integrated processing-and-interpretation method based on seismic wavefield characteristics of common-scatterpoint (CSP) gathers to deal with the exploration challenges. First, we enhanced the amplitude of signals and distinguished weak signals from coherent noises by their kinematic characteristics. The signals, including reflected and diffracted waves, appear as scattered waves with highlighted hyperbolic curves, characterized by the vertex of the hyperbola coinciding with the subsurface scatterpoint and the curvature determined by the root-mean-square velocity of the overlying medium. In contrast, the coherent noises, such as surface and refracted waves, present nearly linear curves with remarkably intensified amplitude in CSP gathers. Then, we established an accurate velocity model and investigated the thickness of the strata. Finally, we ran a prestack time migration (PSTM) for reflected and diffracted waves. The images were improved by mapping signals to the correct position with accurate velocity; therefore, the target strata and the fault planes, with the higher lateral resolution, were more visible. With the aid of well control and the velocity model, main faults and target sequences over the surveyed area were interpreted. Our CSP-gather processing technique not only has enabled us to reveal subsurface complexities and reach the exploration target in the area but also has implications for seismic explorations in similar environments where structures and sequences vary dramatically.

Bottom-simulating reflections (BSRs) in seismic data have been widely accepted to indicate the base of the methane gas hydrate stability zone (MGHSZ), and free gas was thought to exist only below it. However, real geologic systems are far more complex. We have evaluated the results of 3D seismic, logging while drilling, in situ, and coring measurements at a venting gas hydrate system in the Shenhu area of the South China Sea. Our studies reveal that free gas has migrated upward through the thermogenic gas hydrate stability zone into the MGHSZ and become a part of the gas hydrate system. Seismic amplitude anomalies and core results suggest the presence of free gas above the base of MGHSZ at 165 mbsf and the presence of thermogenic gas hydrates below it in well SC-W01. Analyses of P-wave velocity, S-wave velocity, density, and porosity logs reveal that free gas occurs above and below the MGHSZ as well. Integrating log and core analysis with seismic interpretation suggests that the variation in seismic amplitude within the chaotic zone is associated with variable gas saturations, and a large amount of methane and thermogenic gases accumulate near the complex BSRs. We suggest that relative permeability likely plays a significant role in the free-gas distribution and the formation of gas hydrates within a venting gas hydrate system, whereas the effect of dissolved gas short migration is not ignored. Our results have important implications for understanding the accumulation and distribution of gas hydrates and free gas in the venting gas hydrate system and seeps at the seafloor.

The continuous bottom-simulating reflection (BSR) is commonly considered to mark the base of gas hydrate stability zone. Below this depth, gas hydrate gives away to free gas or water filling with pore spaces of sediments. We integrated and analyzed seismic data collected in 2008, and logging-while-drilling (LWD) data and coring results acquired by the Fugro Voyager in 2015 in the Shenhu area on the northern slope of the South China Sea. Based on seismic and well-log correlation, a BSR with typical characteristics of gas hydrates and free gas was identified at 237 m, below the mudline (BML). However, LWD data reveal a 63 m thick hydrate layer from 205 to 268 m BML. Increases in resistivity and velocity at 262 m BML indicate that gas hydrate is likely presented below the BSR. The observed pore-water freshening with depth and infrared image of core samples are consistent with geophysical interpretation. Seismic and well interpretations reveal continuous, discontinuous, and pluming BSRs in the Shenhu area. The continuous BSR indicates the base of the methane gas hydrate stability zone, and structure II gas hydrate is likely presented below the BSR. Deep thermogenic fluid locally entrapped within shallow-buried sediments may reinforce gas-hydrate accumulations near the discontinuous and pluming BSRs. We conclude that seismic responses of structure II gas hydrate can be distinct from structure I gas hydrate. Understanding the seismic characterizations of structures I and II will help in the evaluation of gas-hydrate reservoirs and inferring the presence of deep thermogenic reservoirs.