This session, organized by the UN Decade Program CoastPredict, aims to directly contribute to the UN Decade Challenge 6: Enhancing community resilience to ocean hazards. The focus is on addressing critical gaps in scientific knowledge, particularly in key areas such as coastal risk assessment, warning and mitigation strategies. Key topics include: (i) the collection and generation of observational and modeling datasets essential for risk assessment, including downscaled climate projections for coastal regions, all within a robust data-sharing frameworks; (ii) the promotion of interdisciplinary and international research and innovation to comprehensively address these challenges c, with a particular emphasis on approaches like Digital Twin technology; (iii) the enhanced Early Warning Systems for Ocean-related Hazards through Machine learning and Predictive Modeling, and (iv)the development of standards for risk communication at both national and international levels. The session will also explore multi-hazard early warning systems for events such as tsunamis, storm surges, marine heatwaves, and coastal biogeochemical hazards, including pollution and other extreme coastal events such as erratic extratropical cyclones Contributions on machine learning applications, compound event analysis, and disaster risk reduction strategies are strongly encouraged, as are science-based management practices for enhancing coastal resilience. By leveraging innovative tools like digital twins, this session highlights how predictive modeling can significantly improve risk assessment and response strategies. Its relevance extends to policymakers, scientists, and coastal communities, fostering collaboration to strengthen coastal resilience.