Skip to main content
Blog Post

What do the Palu Bay earthquake, unexpected tsunami mean for risk models?

Forsikringsrådgivning og teknologi|Eiendom
Insurer Solutions

By Geoffrey Saville and Anawat Suppasri | July 15, 2019

An earthquake and tsunami that had not been predicted by available models hit Palu Bay, Indonesia hard enough to defeat the early warning system. Are those models still applicable and what data is needed?

An earthquake and an accompanying tsunami that had not been predicted by available models, hit Palu Bay, Indonesia with a ferocity that defeated the early warning system. Are those models still applicable and what data is needed?

The tsunami that hit Palu Bay on September 28, 2018, cost at least 2,100 lives. In conjunction with the earlier earthquake, it made 206,524 people refugees, caused a healthcare crisis and meant that 1,309 people were classified as missing. What made it more devastating was that the tsunami detectors had been damaged by the earthquake, and that it is possible warnings that were sent out were later retracted. As there was a festival in Palu and a paragliding festival, there were high numbers of visitors who became victims.

One of the most significant elements of the Palu Bay tsunami is that the warning systems that use databases of assumed earthquake information failed to predict it. Indeed, the warning was cancelled after a wave height of just six centimeters was reported at Mamuiu, 180 kilometers away from the epicenter, seeming to confirm the lack of urgency, although the gauge used to record wave amplitude in Palu Bay itself could not transmit because of an electricity blackout and should have meant that the warning remained in force.

And there is a further complication for tsunami models. The Palu Bay event was triggered by a strike-slip earthquake, which is generally not expected to generate a tsunami, as the movement is on a horizontal plane and lacks the vertical motion that generates waves.

Looking for reasons why

Looking more closely at horizontal movement, however, suggests a different picture. A tsunami can be generated by horizontal movement, provided that it happens near a slope, leading to the vertical movement of water increasing by as much as 50%. But in the Palu Bay instance, this would create waves of less than a meter It does not explain why there were five-meter flow depths along many parts of the coastline. What else might have contributed?

An aerial or submarine landslide would explain the extremely localized effect. Palu Bay is at the end of a channel that is 10 kilometers long, flanked by steep slopes and with river mouth inlets that have large amounts of soft sediments below sea level. Video footage from the incident, including one from a helicopter and another from a boat, suggests that there were indeed a number of wave-generating spots, while there is also evidence of aerial landslides in several locations that has been gathered by field surveys taken after the event. There is a hypothesis that these were induced by liquefaction.

In a preliminary study undertaken by Willis Research Network partners at Tohoku University, several areas which had a high possibility of slope failure were investigated. In addition, documented landslide areas from eyewitness accounts were also used in case studies. A simulation was produced using computer modeling of the physical environment around the bay to represent several aerial and submarine landslides and the waves they would generate. The landslide tsunami model was adjusted using the observed waveforms in Pantoloan Port inside Palu Bay.

Unanswered questions

There are still many unanswered questions. The timing of the landslides needs more eyewitness accounts to be reliable, and the exact locations of submarine landslides would need a detailed bathymetric survey and geomorphological modeling to establish. Large scale, detailed surveys would need to be conducted inland and underwater and a multi-hazard simulation would need to be used to produce event probabilities and uncertainties to gather accurate information.

Even in the presence of reliable data, such a complicated tsunami like Palu Bay is difficult to predict and will remain a challenge for modeling and simulations. Equally important is to address how resilient infrastructure in the face of such events. For example, maintaining the continuity of electricity supply during an emergency so that observing and warning systems remain operational. There are also issues around management. Limitations on residential development zones should be applied and then enforced in order to improve the vulnerability of certain areas.

This holds the possibility that while models are made more accurate and simulations given new and richer variables, future disasters can at least be partially mitigated. The impacts of those events that are predicted and even those that, like Palu Bay, happen after a string of untoward events and as a result of natural topography can be greatly lessened.


Senior Research Manager

Geoff joined Willis Towers Watson in 2013, and works with the Willis Research Network stakeholders and academic partners to match business needs to the latest in scientific research, and derive tangible outputs for Willis Towers Watson to help advise its clients to advance their understanding of risk from weather and climate related hazards.

His background is in meteorology and climate science, having worked in forecasting for over a decade for the UK Met Office and Bermuda Weather Service, in all aspects of delivering forecast services from media broadcasting to delivering warnings and actionable guidance on extreme weather phenomena such as tropical cyclones and heavy rainfall leading to flooding.

He holds a BSc in Environmental Science from the University of East Anglia, and a Masters (with distinction) in Climate Change from University College London. He is also an active Fellow of the Royal Meteorological Society.

Anawat Suppasri
Associate Professor, International Research Institute of Disaster Science, Tohoku University

Contact Us

Related solution