Geologically speaking, an earthquake event arises from the sudden release of energy that was accumulated along the fault segments due to tectonic plate movement. A big earthquake in deep sea would cause a significantly large ground uplift/subsidence at seabed, and lead to water waves traveling very fast (e.g. speed of jet aircrafts) and reaching the coast with huge wave height and large run-up to the land. This is called a tsunami. Figure 1 shows a mechanism of tsunami generation by a megathrust earthquake. Not all earthquake events in the sea, even with sufficiently large magnitude, will generate a tsunami. The generation and propagation of tsunami is very complex, and is related to not only the seabed condition, but also to the type of fault, fault rupture mechanism, etc.

Historical records of tsunamis in this region indicate quite clearly that it is not a question of whether a tsunami will occur again; rather, it is only a question of when and at which segment. However, the art and science of earthquake and tsunami prediction is still in the research stage.

At the Department of Civil Engineering, we are focusing on research for prediction of tsunami generation based on seismic signals and tsunami wave information. We are seeking to develop a more holistic approach for prediction of the tsunami-potential of earthquakes, using seismic signals and fault parameters from historical records, and also from real-time computations. There are two main advantages of this approach: (1) the prediction can be done in a very timely manner as the seismic signal travels to stations near Singapore from potential epicenters in Sumatra and Java in less than 15 minutes, while the potential tsunami will hit our region only 30 minutes to 2 hours later; (2) the method will be able to differentiate between earthquake events that are tsunamigenic and non-tsunamigenic. These research findings will be used for tsunami prediction and issuing timely warnings in the region.