Advances in deep-sea hydrothermal circulation studies with machine learning

Deep-sea hydrothermal circulation systems play a crucial role in understanding the Earth’s material cycling and the evolution of extreme ecosystems. However, their unique environments have long constrained effective observation and in-depth study. In recent years, the application of machine learning techniques in geosciences has been expanded, providing new perspectives for exploring deep-sea hydrothermal systems. This paper systematically reviews recent advances in this area, focusing on the classification for hydrothermal vent detection and basement lithology identification, prediction for polymetallic sulfide distribution, analysis of hydrothermal biological habitat, and image recognition of subsea imagery, and ecosystem structure investigation. Additional to significant achievements made, challenges in data scarcity, spatiotemporal sampling bias, limited model generalization, and poor interpretability remain. To address these issues, future research should focus on improving deep-sea observation and data-sharing mechanisms, developing models specified for the complex deep-sea environment, and promoting interdisciplinary integration to construct a multi-source, data-driven research framework.