Search

Abstract The Engineering Ocean Seismic 3D system, which enables ultra-high-resolution seismic surveys on a smaller survey scale, was deployed in Yeongil Bay, Pohang, South Korea. The region is renowned for abundant shallow gas deposits and faults. Based on acoustic impedance contrast and abnormal behavior observations, two seismic anomalies termed channel and polka-dot anomalies have been identified in the seismic volume. Seismic attribute analysis based on the signal amplitude and structural characteristics reveals that these anomalies correspond to shallow biogenic gas deposits. Structural interpretation of the seismic volume revealed that the contractional deformation resulting from post-Miocene neotectonics has resulted in uplift and reverse faulting in the Pohang region, contributing to the formation of anomalies. The channel anomalies correspond to gas-saturated tidal channels that formed during eustatic sea-level changes in the post-Last Glacial Maximum period. The polka-dot anomalies are located in topographic lows and are overlain by neotectonic sediment. The different behaviors of these anomalies in a seismic volume can be attributed to the different thicknesses of overburden overlying each of the anomalies.

The ant-tracking algorithm is commonly used to extract faults in geological structures. However, obtaining 3D ant-tracking data requires the calculation of various volume attributes. To obtain satisfactory data, it is necessary to iterate through the parameters of these attributes to achieve reasonable continuity. Moreover, due to the numerous parameters involved, the algorithm can produce different outputs with each execution. In this study, we aimed to enhance the performance of the ant-tracking algorithm by combining it with a basic U-Net structure. The input and corresponding labels to the model are "cubic shaped" 3D data segmented from the original 3D seismic volume to facilitate cross-validation with distinct regions. We used the label data as the single ant-tracking result to minimize the operator’s bias by executing the ant-tracking with several different parameters and executors and then taking the average. An evaluative comparison of three different loss functions (MAE, RMSE, and MSE) was conducted to identify the optimal function for training the model. Across five out of the six metrics, MSE function demonstrated predominant performance, leading to its adoption. Apart from this, a significant number of misinterpreted faults led us to propose the post-processing algorithm named "Dual-Threshold Iteration." It was initially used to extract fine blood vessels branching out from large vessels in medical image segmentation and adapted in our work to ensure a high level of continuity while ignoring worthless noise. Comparison with the F1 score and the number of 3D-connected components confirmed that the proposed method could generate reduced bias and smoothly connected fault structures.