Data-driven Spatial Analysis and Temporal Prediction of Piracy and Armed Robbery Against Ships

Piracy and armed robbery against ships remain critical threats to global supply chains, evolving in response to economic pressures and enforcement measures. This study applies a multi-method quantitative design combining spatial density analysis and global Moran’s I, bias-corrected Cramér’s V, and ARIMA forecasting using International Maritime Bureau data. Official 2015–2025 summaries contained 1,795 reported incidents, while the annual forecasting series covered 1993–2025. Among 1,677 geocoded incidents with usable positions, 1,319 (78.7%) occurred within 20 km of the physical shoreline, and global Moran’s I indicated positive spatial autocorrelation (I = 0.1226; permutation p = 0.001). Boarded incidents were the dominant category, and tankers were the most frequently recorded vessel category in the descriptive tables. Hold-out validation selected ARIMA(0,2,3) (RMSE = 21.18), but its forecast of 103.5 incidents for 2025 underestimated the realised total of 137 by 24.5%. The 2026–2030 forecasts have wide intervals and are interpreted as an exploratory baseline. Overall, the findings show geographically concentrated reporting patterns and substantial uncertainty in medium-term forecasts.