<![CDATA[Tsunamis pose a significant threat to coastal areas, particularly in seismically active regions like Indonesia. The Tsunami Early Warning System (TEWS) plays a vital role in mitigating the impact of such disasters by providing timely and accurate warnings. This paper presents a comprehensive survey of TEWS developments, focusing on seismic parameter-based methods and machine learning approaches. The study reviews current methodologies, highlights their strengths and limitations, and identifies research opportunities for improving TEWS performance. The survey methodology involves collecting literature from leading databases such as IEEE Xplore, Springer, and Scopus, using keywords related to tsunami detection and warning systems. The selected articles were analyzed based on detection methods, seismic parameters, and system performance in terms of accuracy, computational efficiency, and handling imbalanced data. Key findings include the high accuracy of simulation-based systems like InaTEWS for predefined scenarios but their limitations in unexpected cases. Machine learning methods offer significant potential to address these gaps, but challenges remain regarding real-time processing and imbalanced data handling. Future research directions include developing time-efficient machine learning models for near-field tsunami detection, optimizing algorithms for imbalanced data, and integrating seismic with non-seismic data to enhance prediction accuracy. This paper provides insights for researchers and practitioners aiming to advance TEWS technology, contributing to more effective disaster mitigation strategies.]]>
Copyrights © 2024
