Development and Application of a Grid-Based Risk Assessment Framework for Reliable Maritime Traffic Safety Management

Purpose: This study develops a grid-based risk assessment framework for Korean maritime waters and establishes a pattern classification system to enhance the reliability of maritime traffic safety management.<br/>Methods: A total of 6,014 grids (3 ′×3 ′) were evaluated using eight indicators weighted through the Analytic Hierarchy Process (AHP). Among them, 984 complex high-risk grids were classified into three patterns based on hierarchical criteria: Pattern 1 (Traffic/Operation), Pattern 2 (Environment/Topography), and Pattern 3 (Complex/Non-standard). Statistical validation was conducted using analysis of variance (ANOVA), discriminant analysis, and effect size analysis.<br/>Results: The pattern distribution was 70.6% for Pattern 1, 24.9% for Pattern 2, and 4.5% for Pattern 3. All risk factors exhibited statistically significant differences among the patterns (p < 0.05). Linear discriminant analysis achieved a classification accuracy of 96.24%. Effect size analysis showed that pattern-related variables explained 16.2% of the total variance (η² = 0.162), with particularly strong effects observed for floating-debris entanglement risk (η² = 0.507), merchant-vessel traffic concentration (η² = 0.224), and annual mean significant wave height level (η² = 0.197).<br/>Conclusion: The statistically validated pattern classification supports differentiated safety strategies, including VTS enhancement for Pattern 1, weather-based control for Pattern 2, and floating object management for Pattern 3. The proposed framework provides a scientific basis for evidence-based maritime safety policy and can be extended to dynamic risk assessment and autonomous vessel development.