생성형 AI 플랫폼을 활용한 난카이 지진 시나리오에서 남해안의 도달 파도 예측 연구 = A Study on Predicting Coastal Wave Arrival along the South Korea Southern Coast under Nankai Earthquake Scenarios Using a Generative AI Platform

다국어 초록 (Multilingual Abstract)

This study investigates tsunami propagation and wave arrival characteristics along the southern coast of Korea and Jeju Island, generated by a potential large-scale Nankai Trough earthquake (Mw 9.1), using a generative artificial intelligence (AI) platform. The AI framework, based on the linear shallow-water approximation and travel-time integration method, performed automated multi-scenario calculations, uncertainty quantification, and rapid prediction for 11 distributed fault scenarios. Three analytical approaches—Case 1 (single Okada model), Case 2 (multi-scenario ensemble), and Case 3 (Tohoku observation-based correction)—were compared. The results show tsunami arrival times of approximately 4–7 hours and maximum wave heights mostly below 1.0 m at major southern coastal ports. The corrected Case 3 results demonstrated strong agreement with existing observational and numerical model data. This study suggests that AI-based approximate modeling can serve as a practical tool for preliminary assessments and early-warning support.