Abstract: The rapid development of spatiotemporal data, the Internet of Things, high-performance computing, and artificial intelligence has continuously accelerated the digital and intelligent transformation of maritime transportation. Maritime spatiotemporal data, including vessel trajectories, nautical charts, meteorological and hydrological observations, and shipping business-related data, have become an important information foundation for describing the operational status of waterborne traffic and supporting intelligent maritime governance. Through effective mining and prediction of such data, it becomes possible to improve operational efficiency, transport capacity, service quality, safety management, and environmental sustainability in maritime trans⁃ portation systems.Research published over the past decade has shown that maritime spatiotemporal data mining and prediction has gradually formed a relatively comprehensive methodological framework. Classical statistical analysis remains important for regression, clustering, classification, and time-series modeling, especially in tasks such as fuel-consumption estimation, trajectory pattern discovery, vessel classification, and trend analysis. Spatiotemporal statistical analysis further strengthens the characterization of spatial dependence and spatial heterogeneity by introducing spatial autocorrelation analysis, spatial pattern analysis, and spatial regression models, thereby providing effective tools for traffic-flow monitoring, hotspot identification, and regional risk assessment. With the rapid development of data-driven methods, machine learning and deep learning have been widely applied to trajectory prediction, traffic-state prediction, behavior recognition, and emission forecasting. In parallel, graph theory and complex network methods have provided effective support for representing in⁃ teractions among vessels, ports, and routes, and for analyzing network evolution, vulnerability, and resilience. Existing studies have also demonstrated broad application potential in vessel behavior modeling, waterborne traffic risk assessment, vessel path optimization, port scheduling and management, port and shipping enterprise operation evaluation, shipping network analysis, ship emission monitoring and prediction, vessel target recognition, and maritime knowledge graph construction and application. Overall, maritime spatiotemporal data are driving a transition from experience-based judgment toward data-driven analysis and intelligent decision support.Despite these advances, several challenges remain unresolved, including data heterogeneity, missing and noisy observations, cross-source fusion difficulty, uncertainty propagation, insufficient model interpretability, weak cross-scenario generalization, and the lack of unified evaluation benchmarks. Future research should place greater emphasis on high-quality multimodal data construction, real-time spatiotemporal computing, knowledgeenhanced modeling, graph intelligence, large-model-assisted maritime decision support, and integrated frameworks that combine perception, analysis, prediction, and dynamic control, so as to provide stronger support for intelligent shipping, smart ports, and modern waterborne traffic governance.