The maritime sector, responsible for around 2.9% of global greenhouse gas (GHG) emissions, is under increasing pressure to decarbonize towards the International Maritime Organization’s (IMO) 2050 targets. A number of exisiting reviews have summarized different aspects such as policy frameworks, alternative fuels, and ship technologies, yet few have examined the integrative role of data in advancing maritime decarbonization across these domains. This review establishes a four-pillar framework to capture how data-driven approaches are reshaping maritime decarbonization. It begins by analyzing emerging digital governance policies on the basis of conventional maritime policies. Second, it synthesizes the development of emission datasets, highlighting the indirect databases with direct real-time measurements such as unmanned aerial vehicle (UAV) and portable emission measurement systems (PEMS) monitoring. Third, it reviews methodological advancements in emission modeling, from empirical, economic, and hybrid models to machine learning-based approaches. Finally, it evaluates how these data and modeling advancements support coordinated decarbonization strategies across vessels, fuels, and ports, illustrating the growing role of data-informed decision-making in guiding systemic emission reductions throughout the maritime supply chain. This review links regulatory, data, modeling, and implementation perspectives through a data-centric perspective. By mapping research progress and identifying knowledge gaps across these four pillars, the review offers a structured foundation to support the acceleration of maritime decarbonization efforts.

Accurate ship trajectory prediction is crucial to ensure maritime safety. Most existing ship trajectory predictors face two issues: reliance on post-clustered trajectories and limited interpretability in decision processes. In this research, we address these challenges by proposing an explainable Ship Trajectory Predictor (STPredictor), which is facilitated by strong reasoning capabilities of large language models (LLMs). We reformulate the ship trajectory prediction as a language modeling problem, encoding heterogeneous maritime scenarios as naturallanguage prompts, and employing supervised fine-tuning to design LLMs specifically for the prediction task. Furthermore, we integrate the Chain-of-Thought (CoT) process into the inference pipeline to enhance the transparency and reliability of predictions, and include explanatory requirements in the inference stage to make the decision process align with human instructions. To comprehensively benchmark STPredictor against strong baselines, we construct two large-scale datasets from global Automatic Identification System (AIS) records, including a geospatial-domain dataset and a draught-domain dataset. Extensive experiments based on these datasets demonstrate the superior performance and interpretability of STPredictor in the trajectory prediction task. These findings indicate that LLMs can effectively encode rich interaction information for understanding complex maritime scenarios, thereby laying a solid foundation for reliable and interpretable decisionmaking in maritime safety.

Intelligent maritime transportation systems (IMTS) have become increasingly critical for enhancing navigational safety, improving operational efficiency, and supporting autonomous decision-making in maritime domains. With the growing volume and complexity of maritime operations, IMTS have evolved rapidly through the integration of emerging technologies such as the internet of things (IoT), satellite communication, and artificial intelligence (AI). Among these, deep learning (DL) has shown particular promise, offering powerful capabilities to extract complex patterns from large-scale maritime data and enabling advancements in applications such as ship detection, trajectory prediction, collision avoidance and traffic flow modeling. Despite these developments, a comprehensive review is lacking that critically assesses the strengths and weaknesses of these models, especially the DL-based models used in IMTS. As such, this study contributes to bridge this gap with a quantitative review of the technological evolution of IMTS, and a systematic analysis of DL-based research within IMTS, covering key domains such as risk assessment, autonomous navigation, situation awareness, and intelligent decision-making. Furthermore, this study highlights key challenges in recent researches and identifies future research directions. This study not only provides a holistic understanding of how DL has transformed maritime intelligence but also offers practical insights for developing safe and more efficient IMTS.