Evaluation, sensitivity and comparison of tsunami simulation programmes

Abstract

Although large tsunamis in the Mediterranean Sea are not frequent it does not exclude their occurrence. Furthermore, the Maltese islands do not go unaffected by these events, as seen clearly during the 1908 Messina event. The increase in coastal development and enhancement of Malta’s tourist areas has further increased the risk imposed by tsunamis therefore, tsunami early warning systems (EWSs) are of utmost importance for the Maltese islands, as these areas would endure the greatest impact. Hence, the importance of the evaluation of the performance and sensitivity of TOAST (Tsunami Observation and Simulation Terminal), a tsunami simulation software installed at the University of Malta as an EWS. However, for TOAST to produce rapid results it utilizes coarse bathymetry that only extends to the periphery of ports and harbours. Therefore, its results are compared with those produced by HySEA (Tsunami-Hyperbolic Systems and Efficient Algorithms), a high-performance software package, specifically designed to simulate earthquake generated tsunamis that makes use of nested bathymetry data. The 365 AD Crete event was taken as the reference simulated event. Its location of 35.25 °N and 23.53 °E was used, with a magnitude of 8.5 and fault parameters: dip – 35°, strike – 315°, rake – 90°, length – 160 km, width – 80 km and depth – 10 km. It was concluded that the simulated wave heights and inundations were larger from TOAST than from HySEA, with a maximum wave height of 5.39 m (TOAST) and 4.37 m (HySEA). This overestimation from TOAST is attributed to the use of coarser bathymetry data for quick simulations resulting in estimations of run-up and wave heights at the coast.