Accurate, rapid, and automatic seafloor sediment classification represents a crucial challenge in marine sediment research. To address this, our study proposes a seafloor sediment classification method integrating convolutional neural networks (CNNs) with small-sample multi-beam backscatter data. We implemented four CNN architectures for classification—LeNet, AlexNet, GoogLeNet, and VGG—all achieving an overall accuracy exceeding 92%. To overcome the scarcity of seafloor sediment acoustic image data, we applied a deep convolutional generative adversarial network (DCGAN) for data augmentation, incorporating a de-normalization and anti-normalization module into the original DCGAN framework. Through comparative analysis of the generated versus original datasets using visual inspection and grayscale co-occurrence matrix methods, we substantially enhanced the similarity between synthetic and authentic images. Subsequent model training using the augmented dataset demonstrated improved classification performance across all architectures: LeNet showed a 1.88% accuracy increase, AlexNet an increase of 1.06%, GoogLeNet an increase of 2.59%, and VGG16 achieved a 2.97% improvement.

Marine soft soils are characterized by high compressibility, low strength, and low permeability, which often result in excessive settlement and stability problems. Drainage consolidation methods are widely regarded as effective solutions for improving such soils. This review summarizes recent progress from four perspectives: optimization of traditional techniques, combined applications of multiple methods, development of emerging innovative approaches, and advances in drainage element materials and structures. Traditional methods such as surcharge and vacuum preloading have been refined through innovations in loading schemes, drainage improvements, and design approaches, while hybrid combinations with electroosmosis, thermal treatment, and dynamic loading have further enhanced their efficiency and applicability. In parallel, novel techniques such as siphon drainage, aerosol-assisted consolidation, and osmosis-based drainage show promise for sustainable applications. Furthermore, biodegradable and multifunctional drainage elements provide new directions for environmentally friendly and efficient soft soil improvement. Looking ahead, drainage consolidation technology is expected to move toward greener, low-carbon, and intelligent solutions. This review offers a comprehensive reference for engineering practice and a useful basis for guiding future research in marine soft soil improvement.

This study addresses the challenge of data scarcity in research on the migration patterns of shallow gas in submarine sediments. Taking the northern Hangzhou Bay area of the East China Sea as an example, we integrate borehole core geophysical surveys and geochemical data to elucidate the migration and fractionation mechanisms of shallow biogenic gas. A three-zone conceptual model—“disturbed zone–active zone–residual zone”—dominated by lithology-controlled migration is established, revealing the dominant roles of gas escape, mixing-homogenization, and adsorption fractionation in heterogeneous sedimentary systems. The results show that high-permeability sand layers can act as adsorption-fractionation windows, causing significant enrichment in δ¹³C-CH₄ (–57.4‰). We propose an analytical framework of “zonal verification–mechanism tracing”, which overcomes the limitations of traditional Rayleigh fractionation models and enables accurate interpretation of gas migration patterns in heterogeneous systems using limited data such as δ¹³C-CH₄ and CH₄ concentration. This provides a new paradigm for engineering surveys and risk assessment in low-data-density contexts. The findings indicate that the shallow seepage zone poses low engineering risks, while the residual zone serves as an indicator of depleted gas reservoirs. The proposed analytical approach can be extended to preliminary submarine engineering surveys and hazard assessments in other regions.