MC-QRA: Quantitative Spatiotemporal Risk Assessment and Probabilistic Dependency Analysis of Maritime Cyber Attacks

Gong, J, Zhao, Y, Li, H orcid iconORCID: 0000-0002-4293-4763 and Yang, Z orcid iconORCID: 0000-0003-1385-493X (2026) MC-QRA: Quantitative Spatiotemporal Risk Assessment and Probabilistic Dependency Analysis of Maritime Cyber Attacks. Reliability Engineering & System Safety. ISSN 0951-8320

[thumbnail of MC-QRA_Quantitative Spatiotemporal Risk Assessment and Probabilistic Dependency Analysis of Maritime Cyber Attacks.pdf]
Preview
Text
MC-QRA_Quantitative Spatiotemporal Risk Assessment and Probabilistic Dependency Analysis of Maritime Cyber Attacks.pdf - Published Version
Available under License Creative Commons Attribution.

Download (7MB) | Preview

Abstract

The increasing digitalisation of maritime systems has raised concerns about cyber threats to shipping operations, port infrastructure and maritime safety. However, existing approaches often examine cyber-risk scenarios, geographical patterns and dependencies among risk factors separately. This study proposes Maritime Cyber-Quantitative Risk Assessment (MC-QRA), an integrated framework for scenario-based relative risk assessment, spatiotemporal analysis and probabilistic dependency modelling using reported maritime cyber incidents. The framework comprises four stages: latent class analysis (LCA) for scenario identification; frequency–severity scoring for relative risk assessment; kernel density estimation (KDE) and Getis–Ord Gi* analysis for spatiotemporal assessment; and mutual information (MI) and Tree-Augmented Naive Bayes (TAN) modelling for dependency analysis. The framework was applied to 444 incidents from the Maritime Cyber Attack Database (MCAD), identifying five distinct cyberattack scenarios. The results show an increase in reported incidents classified as medium or high risk after 2021, with risk-weighted spatial concentrations in Europe, North America and East Asia. Scenario-level risk decomposition indicates that ransomware-related incidents account for 52.2% of the summed relative risk scores, compared with 19.6% for spoofing-related incidents. Cyber threat type, exploited asset and target category provide the most information for distinguishing the assigned risk levels. The TAN model achieved a mean classification accuracy of 87.4% under stratified fivefold cross-validation, while identifying conditional probabilistic dependencies among incident characteristics. The resulting risk indices and spatial patterns describe relative differences within the reported MCAD sample rather than exposure-adjusted absolute risk. MC-QRA thus provides an integrated basis for comparing incident scenarios, examining spatiotemporal patterns and supporting preliminary maritime cyber-risk prioritisation.

Item Type: Article
Uncontrolled Keywords: 40 Engineering; 49 Mathematical Sciences; 35 Commerce, Management, Tourism and Services; 01 Mathematical Sciences; 09 Engineering; 15 Commerce, Management, Tourism and Services; Strategic, Defence & Security Studies; 35 Commerce, management, tourism and services; 40 Engineering; 49 Mathematical sciences
Subjects: Q Science > QA Mathematics > QA76 Computer software
T Technology > TA Engineering (General). Civil engineering (General)
Divisions: Engineering and Built Environment
Publisher: Elsevier
Date of acceptance: 5 November 2026
Date of first compliant Open Access: 20 November 2026
Date Deposited: 20 Nov 2026 14:09
Last Modified: 20 Nov 2026 14:09
DOI or ID number: 10.1016/j.ress.2026.113823
URI: https://researchonline.ljmu.ac.uk/id/eprint/29900
View Item View Item