The very first PDM holder

In September 2022 the Post-graduate School of Societal Safety Sciences, awarded Karina Sujatmikoa a PDM degree. Faculty and students of PDM celebrated the birth of the very first PDM holder.

PDM (Ph.D. of disaster management) program, 3 years' doctoral course only in English was inaugurated by Post Graduate School of Societal Safety Sciences in 2019.

Karina Sujatmikoa, from Indonesia, finished redaction of her doctoral thesis and became the very first PDM holder. The title of the thesis is "Factors affecting the initiation of Tsunami evacuation". Her work was supervised by Professor Yoshihiro Okumura.


This study aims to understand better the natural phenomenon of landslide-induced tsunamis and the social phenomenon of a tsunami evacuation. The outcomes provide data to compare and contrast of the evacuation process during a typical seismic tsunami and a non-typical seismic tsunami. This research also compares the evacuation behavior in Indonesia and Japan. With this information, researchers could identify issues and make evacuation advice for future non-typical seismic and non-seismic tsunami events.

Chapter 1 explained the background and the context of this study by detailing the aims and objectives of this thesis. This chapter introduces the 2018 Palu earthquake and tsunami events and provides information on the location, disaster timeline, and casualties. The development of tsunami risk reduction in Indonesia also present to give the illustration of the current situation.

Chapter 2 present our finding from the analysis of landslide-induced tsunami that submarine landslide is enough to generate a destructive tsunami. The scientific communities and Government of Indonesia focused the research and countermeasure for tsunami earthquake disaster. Development of tsunami warning technology also addressed for earthquake tsunami. This tendency is understandable since more than 75% tsunami generated by earthquake and survey for landslide-induced tsunami is hard to conduct. This study proposed a new approach to address the difficulty of determining the parameter of submarine landslide by using the soil investigation inland to be used as input for the landslide-induced tsunami simulation. Hopefully, in the future tsunami hazard analysis, the landslide-induced tsunami is not neglected. It is possible that only a landslide, without a strong earthquake, is enough to generate a destructive tsunami. It means ground motion will not be available as a critical evacuation trigger. Therefore, the evacuation plan will consider the risk of landslide-induced tsunami disaster.

Chapter 3 explains the definitions and concepts about tsunami evacuation used in this research. The framework of this study is the intuitive tsunami evacuation, specifically the behavior changes from the response stage into the evacuation initiation stage. Two crucial aspects affecting their decision are reality-of-evacuation-start (RES) and awareness-level-of-danger (ALD). The environmental, social, and warning information cues caused the "must-escape" situation, or in this research called as RES sources are the interest of this study. The method was developed to analyze RES sources quantitively and qualitatively from the post-disaster survey, video analysis and numerical simulation.

Chapter 4 shows the standard ranking of evacuation trigger and revealed six triggers worked for Indonesian people. Which is receiving a message from the authorities, seeing evacuees, feeling ground motion, seeing unusual sea surface or tsunami, hearing loud sounds from the sea, and hearing evacuees calling for evacuation. The observation of 53 individuals from 6 static cameras installed in the hotel located at the east coast of Palu Bay provides information about the impact and exposure that the field survey does not capture. Feeling ground motion is the first RES source exposed to people. People did not evacuate immediately after feeling ground motion. Early evacuees are 38 individuals (68%) who start evacuating before tsunami coming. Late evacuee are 17 individuals (32%), who started to evacuate after seeing the tsunami come into their location. The observation of 200 individuals captures on the 4 phone cameras in the commercial area (Palu Grand Mall), located on the west coast of Palu bay found that hearing other people calling for an evacuation is likely more impactful than revealed in the post-tsunami interview surveys. At least in this case study, RES sources generated by social cues, both seeing and/or hearing other people, are not constantly influential and gradually increase as the distance between the source and people decreases.

Chapter 5 shows the implementation of tsunami evacuation simulation in the residential area located on the east coast of Palu bay. This model analyzes the reality-of-evacuation-start in a spatiotemporal manner. The influenced weight of RES sources generated by social cues, both seeing and/or hearing other people, are not constant. Instead, the impact should gradually increase as the distance between the source and people decreases. Based on the scenario, it is found that to achieve 75% of residents decide to do early evacuation before tsunami came, then at least 47% of resident should evacuate immediately after feeling ground motion. This group of people should not be waiting for other cues, such as receiving message to evacuate from the authorities, observing other people behavior or waiting until they see or hear the tsunami arrival. The comparison between tsunami evacuation in 2018 Palu Tsunami, Indonesia and 2011 Tohoku Tsunami in Ishonomaki, Japan revealed that in Indonesia, seeing other people had a significantly higher influence on people to initiate evacuation. In Japan, hearing had a greater impact than seeing. The low percentage of hearing was probably because of small exposure area of RES source and due to unclear instruction from shouting people. Case study comparison has the advantage to show the difference in detail, but a general comparison is needed to provide a broader picture. General comparison of evacuation standard ranking in Indonesia and Japan shows that receiving messages from authorities is the number one rank of evacuation trigger in both countries. The feeling of ground motion become the number two rank in Japan and number three in Indonesia. Interestingly, there is a wide gap between the rankings of "seeing other evacuees", it is on rank two for Indonesian people, but it is the lowest rank for Japanese people. In the video and the questionnaire, Indonesian people tend to quit the building after feeling ground motion and staying outside the building. In Japan, after the earthquake, people stayed inside the building and sought disaster information through media. Therefore, they could not see what other people were doing outside the building.

Further research on the difference in tsunami evacuation initiation is important because this indicates that the disaster risk reduction strategy should not be generic among countries. Educational material for tsunami training and drills must consider the local characteristics. Further research should examine RES sources in order to determine appropriate evacuation strategies for non-seismic tsunamis. To mitigate future similar disasters in the region, the high-frequency ocean surface radars is a potential tsunami observing systems. This warning system could be very impactful for Indonesian people, especially if we consider that receiving message from authorities rank as the number one RES sources.