Submarine landslides and their tsunami hazard

Most tsunamis are generated by earthquakes, but in 1998, a seabed slump offshore of northern Papua New Guinea (PNG) generated a tsunami up to 15 m high that killed more than 2,200 people. The event changed our understanding of tsunami mechanisms and was the forerunner to two decades of major tsunamis that included those in Turkey, the Indian Ocean, Japan, and Sulawesi and Anak Krakatau in Indonesia. PNG provided a context to better understand these tsunamis as well as older submarine landslide events, such as Storegga (8150 BP); Alika 2 in Hawaii (120,000 BP), and Grand Banks, Canada (1929), together with those from dual earthquake/landslide mechanisms, such as Messina (1908), Puerto Rico (1928), and Japan (2011). PNG proved that submarine landslides generate devastating tsunamis from failure mechanisms that can be very different, whether singly or in combination with earthquakes. It demonstrated the critical importance of seabed mapping to identify these mechanisms as well as stimulated the development of new numerical tsunami modeling methodologies. In combination with other recent tsunamis, PNG demonstrated the critical importance of these events in advancing our understanding of tsunami hazard and risk. This review recounts how, since 1998, understanding of the tsunami hazard from submarine landslides has progressed far beyond anything considered possible at that time.