On September 28, 2018, following a magnitude 7.5 strike-slip fault earthquake, an unexpected tsunami inundated the coast of Palu bay, Sulawesi, Indonesia, causing many casualties and extensive property damage. However, the earthquake’s mechanism rarely generates a destructive tsunami. The tidal record at Pantoloan, located along the coast of Palu bay, indicates that the tsunami arrived 6 min after the earthquake and generated 2 m of receding water. It had a maximum wave height of 2 m and arrived approximately 2 min later. The tsunami had a relatively short period and caused devastation as far inland as 300 m. Additionally, 8 m high watermarks were observed near the coast; the flow depth decreased to 3.5 m inland (Fig. 1). Amateur videos and eyewitness accounts indicate that the tsunami did not enter the bay through its mouth but obliquely from an area inside the bay. Our hypothesis, therefore, is that the killer tsunami was most likely generated by an underwater landslide occurring inside Palu bay. While detailed bathymetric data are still needed to confirm this hypothesis, in this article we provide a preliminary analysis of the available data, supported by the results of a field survey, to strengthen this hypothesis and provide direction for further post-tsunami surveys and analysis.

The height of the tsunami caused by the Sulawesi earthquake in 2018 was higher than that estimated from the magnitude of the earthquake, and its cause is not well understood. Although it is conceivable that tsunamis originated from landslides in several cases, it is not known whether landslides were occurring in the coastal area at the time. This report describes the results of a field survey conducted to investigate the traces of subsidence and the characteristics of the tsunami incidents in the bay far from the epicenter. Subsidence was observed at a minimum of seven points along the coast; in accord with the reports of eyewitnesses, this suggests that subsidence might have generated tsunamis.