Failure mechanism and tsunamigenesis of the Anak Krakatau landslide on 22 December 2018

Tsunamis represent a major natural hazard and are capable of inundating low-lying coastlines, destroying infrastructure and homes, and causing injury and loss of life. Tsunamis are commonly associated with earthquakes beneath the seafloor; however, landslides into or beneath the sea can also generate catastrophic tsunamis. Volcanic islands are particularly susceptible to landslides due to their often steep and unstable submarine slopes. On December 22 2018, the SW slope of the Anak Krakatau volcanic island, in Indonesia, collapsed into the sea and created a tsunami that travelled across the Sunda Strait in less than 30 minutes. The tsunami destroyed over 2,750 buildings and 510 ships in southern Sumtra and northern Java, displaced over 43,000 people, injured 14,059 people, and caused 437 fatalities.

Although landslides from other volcanic islands have generated tsunamis, the case at Anak Krakatau is very unique as Anak Krakatau is a very young and active post-collapse cone. Anak Krakatau formed in the centre of the Krakatau caldera following the famous and catastrophic Krakatau caldera-collapse eruption in 1883. In 1926, renewed volcanic activity in the centre of the collapse-caldera created the volcanic island of Anak Krakatau. The landslide and tsunami at Anak Krakatau represents a new hazard scenario as collapses on such a young volcanic edifice have not been reported at this scale and not been reported to generated tsunamis that have resulted in such devastating destruction and loss of life. It is therefore urgent that we study this event to better understand why and how it happened.

We propose to survey the submarine slopes of SW Anak Krakatau and map the submarine landslide using multi-beam echosounders to analyse the seafloor surface and seismic reflection methods to investigate beneath the surface. This will provide invaluable insights into the size of the landslide, how the slope failed, and the dynamics of the landslide as it moved downslope. These properties are important towards understanding the tsunami that was generated, and will provide more accurate landslide measurements that can be used to model the landslide-tsunami process. This opportunity is very timely, as no submarine landslides have been studied so soon after the event. Volcanic islands are very dynamic environments, and photographic evidence shows that the embayment above sea level is already being infilled by lavas and volcanic sediments. Therefore, it is urgent to survey the submarine slopes of Anak Krakatau before they are modified by future landslides and sediment flows or are buried by volcanic sediments.

To better understand these natural hazards we have previously had to study volcanic island landslides in the geological record. However, study of these past examples often present large uncertainties in some of the properties we are trying to measure because the slopes and landslide deposits have been eroded, reworked and buried over time. In particular, landslide volume can have significant uncertainties, and yet it is arguably the largest contributing factor to generated tsunamis. Here, we have an opportunity to accurately study the properties of this particularly unique volcanic island collapse only months after the event.

We also propose to study the distal deposits of sedimentary flows from the Anak Krakatau landslide. These deposits can indicate whether the original landslide failed in a single or multiple stages. This is important to determine because if the landslide failed in a series of stages the volume was not released all at once, and so the tsunami potential would decrease. It is urgent to sample these deposits as the sedimentary record may become ambiguous should subsequent landslides and sediment flows occur and obscure the record of this event.

We propose to use the results of this study to better inform tsunami models used to determine mitigation strategies, and pass these results to stakeholders in Indonesia.

There are approximately 600 active volcanoes globally. Among these, 35 volcanoes represent direct analogues to the Krakatau collapse caldera in which the post-collapse cone Anak Krakatau is located. Understanding the relationships between volcanic island landslides and how they generate tsunamis has global implications, especially to countries in the Indo-Pacific and Caribbean regions where there is heightened volcanic active in areas adjacent to the sea with rapidly growing coastal communities.. In the last 100 years, in Indonesia alone volcanic landslide-tsunamis at Paluweh (1928), Illiwerung (1979, 1983) and Anak Krakatau (2018) have resulted in over 1,120 fatalities. This study will be particularly important to the Government, regional governments of Java and Sumatra, NGOs, and the public in Indonesia. The results of this study will be directly shared with the Volcanological Survey of Indonesia (PVMBQ) within the Indonesian Ministry of Energy and Mining, the Indonesian National Board for Disaster Management (BNPB), and the Meteorology, Climatology and Geophysics Agency (BMKG). This will be facilitated by collaborations with Partner Tappin and previous work conducted by Hunt in Indonesia and SE Asia.

This research project contributes towards the first priority of the Sendia Framework 2015: understanding disaster risk. This research project will benefit both scientists and stakeholders. For example, geoscientists, volcanologists, landslide and tsunami modellers, and natural hazard researchers will be able to use data and results from this study to better inform their models and theoretical understanding of this landslide tsunami. In governance, risk managers and NGOs will potentially have a greater knowledge base for hazard mitigation decisions using results from this study. In Indonesia, results from this study will likely enable improvement of hazard and risk assessments and design and implementation of mitigation strategies towards future volcanic, landslide and tsunami activity. The dissemination of the results, interactions with stakeholders and their utilisation of the results towards hazard mitigation would contribute towards the second priority of the Sendai Frameowrk 2015: strengthening disaster risk governance to manage disaster risk. To ensure dissemination of the project results to the stakeholders in Indonesia a series of online webinars will be held with stakeholders over the duration of the project. At the close of the project, meetings with stakeholders in Jakata, Indonesia will be held and a workshop for 20 in-country attendees from stakeholder NGOs will be organised.

Hunt and Clare are part of a science community studying submarine landslides, through which project results will be disseminated. Co-I Clare and Partner Tappin are also members of the UK Natural Hazards Partnership where natural hazards posed to UK and to UK interests abroad are discussed. Partner Tappin is also involved in studies of tsunamis in Indonesia and is PI of the NSFGEO-NERC project researching the 1883 Krakatau eruption and tsunami with stakeholders in Indonesia. The results of this study will be disseminated and promoted through science community and stakeholder contacts with Prof Tappin.

The timely dissemination of project results and engagement with the community is an important aspect of project impact, and will be facilitated using multi-media platforms. The project and its results will be made accessible through a project website hosted by the National Oceanography Centre (NOC). Partner Tappin has an extensive record with media and has been featured in scientifically oriented media, including the National Geographic, BBC, and Discovery channels.

The results of this study will also be published in high impact journals that have broad readership. PI Hunt, Co-I Clare and project partners have extensive track records of publishing in high impact journals, through which we intend to publish the results of this study