The 22 December 2018 Mount Anak Krakatau volcanogenic tsunami on Sunda Strait coasts, Indonesia: tsunami and damage characteristics

On 22 December 2018, a tsunami was generated from the Mount Anak Krakatau area that was caused by volcanic flank failures. The tsunami had severe impacts on the western coast of Banten and the southern coasts of Lampung in Indonesia. A series of surveys to measure the impacts of the tsunami was started 3 d after the tsunami and lasted for 10 d. Immediate investigations allowed the collection of relatively authentic images of the tsunami impacts before the clearing process started. This article investigates the impacts of the 2018 Sunda Strait tsunami on the affected areas and presents an analysis of the impacts of pure hydrodynamic tsunami forces on buildings. Impacts of the tsunami were expected to exhibit different characteristics than those found following the 2004 Indian Ocean tsunami in Aceh. Data were collected from 117 flow depths along the Banten and Lampung coasts. Furthermore, 98 buildings or houses were assessed for damage. Results of this study revealed that the flow depths were higher in Banten than in Lampung. Directions of the tsunami arrays created by the complex bathymetry around the strait caused these differences. Tsunami-induced damage to buildings was mostly the result of impact forces and drag forces. Damping forces could not be associated with the damage. The tsunami warning system in Indonesia should be extended to anticipate non-seismic tsunamis, such as landslides and volcanic processes driven by tsunamis. The lack of a tsunami warning during the first few minutes after the generation of the first wave led to a significant number of human casualties in both of the affected areas.

. Local area communities received no warning of the 22 December 2018 tsunami that was generated by the Anak Krakatau. This was one reason for the large number of human casualties on both sides of the affected area. What made this tsunami of particular interest was the rare process that led to its generation, and the fact that it occurred at night, hindering any direct visual anticipation and interaction by area communities.

5
In some cases, people living around the volcano are not well prepared to face a potential tsunami that can be generated by eruptions. This was the case was for those communities around the Thyrrenian coast of Italy (Teresita et al., 2019). Sustained public engagement is necessary, as in the Mount Zao case in Japan (Donovan et al., 2018). As the record of volcanogenic tsunamis is long, proper mitigation of nontectonic tsunamis could be difficult in practice.  (Bani et al., 2015;Zen, 1970).' Mount Anak Krakatau is considered among the most active 20 volcanos in the world. Before the 2018 eruption, it erupted 40 times over the past 85 years (GVP, 2019).
Nonetheless, only the 2018 eruption caused a tsunami wave that affected the southern coast of Sumatra (Lampung Province) and the western coast of Java Island (Banten Province). In 2012, Giachetti et al. identified an active zone at the southwestern flank of Mount Anak Krakatau that expanded from time to time. They also revealed that the southwestern flank failure could generate a-45 m wave toward the small 25 islands surrounding the volcano complex that could reach the Banten area with 1.5 m wave heights within 35-45 minutes after the collapse (Giachetti et al., 2012). The growth of Mount Anak Krakatau was observed toward the southwest based on a survey in 1994 (Deblus et al., 1995). This made the slope of compiled in a database that is stored at the Mendeley database (Syamsidik et al., 2019b). Some locations with measured flow depths, damaged buildings and tsunami boulders can be referenced by the database.
In this article, reported impacts of the tsunami on buildings provide novel findings where field tsunami impact data has mostly been based on tectonic tsunamis. Here we present results of an analysis of the damage to buildings due purely to tsunami hydrodynamic forces. Notwithstanding the presented findings, 5 we acknowledge some limitations of the study as elucidated in Section 6, Discussions and Limitations.
This article is expected to contribute to tsunami engineering studies, and to a better understanding of tsunami mitigation efforts, especially on analysis of damage to buildings due to tsunamis.

10
The tsunami-affected area was largely reported to be from the southern coasts of Lampung and western coasts of Banten. In this study, we investigated five areas as shown in Fig. 1. Two areas were Pandeglang district and Serang District of the western coasts of Banten. The other three areas were South Lampung and Tanggamus districts of Lampung Province. Serang District has a total population of about 1.56 15 million as of 2017, while Pandeglang has a total population about 1.21 million as of 2017. Serang is famous for its industrial areas, where a large steel company is located at its coast. Pandeglang is known for tourism, where a number of hotels and resorts, located along its coast, are heavily occupied during the long holiday season. Some areas in Pandeglang were difficult to assess using land transportation due to road damage, and some routes are still not constructed. This made it difficult to investigate the most 20 southern part of Pandeglang. Along Banten, we investigated about 112 km of its coastlines.
In Lampung area, about 57 km of coastline were investigated, covering the two districts at the southern part of the province. South Lampung district has a population of about 980 thousand people, most of whom reside at the coastal area. Kalianda is the capital city of the district, which was also affected by the 2018 Sunda Strait Tsunami. Another area in the district that was affected by the tsunami is Rajabasa.

25
People in the two areas are mainly farmers and fishermen. A large ferry port accommodating a number of ferry lanes servicing Banten and Lampung is located in this district. Fortunately, there was no major impact on the port. The other district in Lampung investigated in this study is Tanggamus. Here, only one victim reportedly died from the tsunami. We investigated a small bay where human casualties were reported in this district. triangles. Bathymetry data was adopted from Topex (2019).

Survey method
A series of surveys was conducted at the five affected areas as explained in the previous section from 24 December 2018 until 3 January 2019. The ten-day survey was initiated to measure the impacts of the 5 tsunami on Banten area. In this area, the team spent six days measuring tsunami flow depths, tsunami boulders, and damage to housing or buildings in the affected area. Another four days of the survey were spent in Lampung area measuring the same data. Measurement of flow depths was done using a levelling staff and water pass to measure water marks, broken twigs, or stranded tsunami debris. A handheld GPS was used to locate the coordinates of the measured flow depths. The GPS was also used to measure limit 10 of tsunami inundation. A similar method was used to measure impacts of the 29 September 2009 American Samoa tsunami and the 1946 Aleutian tsunami, the 2004 Indian Ocean tsunami in Banda Aceh (Borrero et al., 2006), and the 2018 Palu tsunami (Syamsidik et al., 2019a). A drone was also utilised in this survey to capture images from the tsunami-affected area. In total, we managed to measure 117 flow depths from both sides of the affected area. All data was stored in the Mendeley dataset (Syamsidik et   15 al., 2019b). The survey was not performed at offshore islands around Mount Anak Krakatau, as the area was restricted by the Government for any activity due to the threat of volcanic activity and the tsunamis.
The tsunami arrival at the coastal area was analysed based on water elevations measured at four tidalgauge stations. The locations of the tidal-gauge stations can be seen in Fig. 1. To separate the tsunami 20 wave data from the long-frequency data influenced by astronomical components, a low pass filter was applied. Threshold frequency for the filtering was 0.0805 cycle per day (cpd) as suggested by Emery and Thomson (2001). Arrival times were interpreted based on the first peak of the wave recorded at the tidalgauge stations. To confirm the conditions around the arrival times of the tsunami, we also performed a number of interviews in the local community. Three main questions were asked, i.e. indications or some 25 sign before the tsunami arrived, the number of waves, and the evacuation process. Ten persons were asked at the Banten coasts and five in Lampung. Since the number of interviewees was limited, results of the interviews were meant to confirm the conditions before the tsunami arrival qualitatively.

30
Velocity data was inferred from tsunami boulder transportation. Types of boulders were classified based on size and material composition. During the survey, they were dominated by two kinds of boulder materials, namely coral and ruble mound material from revetment structures. A description of the analytical solutions for inferring tsunami boulder transportation was given by Noormets et al. (2004).
Tsunami boulders were measured for their dimensions, original locations (based on interviews and 35 materials), and distance of transport. Tsunami boulder transportation velocities were inferred using eq (1), as suggested by Paris et al. (2010): where umin is minimum estimated velocity (m/s), μ is friction coefficient, which is 0.7 as suggested by Noormets et al. (2004), g is gravitational acceleration (m/s 2 ), Cd is drag coefficient, which was considered as 1.95, An is the areal of the boulder perpendicular to the tsunami flow direction, and ρw is water density (1027 kg/m 3 ).
Distance travelled by the tsunami boulder depended on tsunami velocity, size of the boulder, and boulder 5 material. The range of seawall material transported as a tsunami boulder was 3-4 m/s for sliding or rolling movement, and 11-12 m/s if the boulder moved as a saltation mode (Nandasena et al., 2011;Paris et al., 2009).

20
The types of the damage were classified into five damage states (DS) as suggested by Suppasri et al. (2014) and Macabuag et al. (2016). The list of DS classifications can be seen in Table 1. To exclude houses altered due to clearing processes, the assessment was limited to the confined masonry-brick infill buildings (CM) and wooden houses. Types of houses in the Banten and Lampung affected areas were relatively similar. Most of the buildings located near the coastal areas functioned as residences or 25 villas/cottages.
Data collected was analysed using fragility functions to produce cumulative probabilities of damage caused by the tsunami. Equation (2) was used to calculate the fragility functions:

30
where P is the cumulative probability of damage, ɸ is the standardised normal distribution function, x is the hydrodynamic feature analysed for damage (in this case, flow depth), and and are mean and standard deviations of x, respectively.
The fragility function was developed to estimate the future impacts of a tsunami. In previous cases, the 35 function was developed based on tectonic tsunamis (Koshimura et al., 2009;Suppasri et al., 2015). In this study, as there was no earthquake preceding the tsunami, analysis of the damage was based purely on tsunami wave propagation.

25
Comparing the two main affected areas, a significant flow depth decay on both sides of the study areas was found. Highest flow depth was found at Cipenyu of Bantern, where the tsunami wave attacked directly from the area around Krakatau Complex. In the southern and the northern parts of this location, flow depth decreased. This was also found to be true in Lampung area. Detailed findings from both of 30 these areas are elucidated as follows.

a. Serang District
Impacts of the tsunami in Banten were found to be more severe than in Lampung. Along the western 35 coast of Banten, impacts were visible but not seen to be continuous.

20
At Cinangka sub-district, maximum tsunami flow depth was measured at 3.75 m (F25). This was identified from the peeled skin of a tree in Bulakan village. At Bulakan Pondok coast, in the same subdistrict, flow depth was measured at 2.08 m (F10), identified by a thick tsunami deposit at the stairway 25 of a villa located about 12 m from coastline (see Fig. 4).

b. Pandeglang
Unlike Serang, where most of the coastal area is dominated by villas for tourism, Pandeglang's coastal 30 area is mostly residential. Suka Ramai village, located at Sambolo Bay, was severely affected by the tsunami as can be seen from a cross profile of tsunami height and aerial images in Figs. 5 and 6. Tsunami wave direction is marked with a white arrow in the figure. Direction was identified from the direction of the fallen trees, swept away by the tsunami wave. At this location, tsunami flow depth was at 4.10 m (F16) based on a broken tree branch. A higher flow depth was at 4.85 m at point F20 (Appendix).

35
At this location, a transect was performed to measure tsunami height.     identified from a broken branch of a Pandanus Odorifer sp. tree. According to an interview with the 5 resort staff, seven people died. They were guests and staff of the resort. Severe erosion was also seen along the coastline near the resort. The deadly force of the tsunami wave was also revealed by a large tsunami boulder (Bo-5), which was found around 87 m from its initial source (see Fig. 8). The boulder's origin was based on eyewitness interviews stating that the boulder was at the beach area close to the resort before the tsunami. Another tsunami flow depth's mark close to the boulder was measured at 5.10 Sumur sub-district was the most affected sub-district at the southern part of Banten coast. This sub-20 district was isolated for about three days after the tsunami due to massive damage to roads and bridges connecting the sub-district to other areas. According to interviews with eyewitnesses, they experienced two tsunami waves where the second wave was the largest and the most destructive one. Before the first wave, residents heard a roaring sound from the sea that motivated most of the them to evacuate to higher ground. The residential area in this sub-district is situated very close to the sea. Some of the houses were 25 located immediately behind a seawall. Tsunami flow depth was measured at 4.75 m from ground level (Point F60 in the Appendix). This point is located about 40 m from the seawall. A higher tsunami flow depth, measured at 5.25 m (F61), was also found in this sub-district. Most houses in this area were semipermanent-type houses, where impacts of the tsunami waves on the houses were severe. Tsunami inundation limit was measured at 155 m from the coastline.

30
After Sumur, a survey was performed at Kertajaya sub-district, which was the last survey area in Banten.
Tsunami impacts were investigated carefully at Cinibung resort. Here, flow depths were between 2.25 m-3.15 m. The survey at Banten area was completed on 30 December 2018, eight days after the tsunami.  Distributions of tsunami flow depths in Kalianda can be seen in Fig. 9. In general, the majority of the Way Kiayi coastal residential community was deserted. Although tsunami flow depth was as high as 10 2.00 m and could still be found at an area about 100 m from coastline, not much major damage to buildings was identified. Nonetheless, some houses that were located around the coastal area were demolished by the tsunami wave, leaving floors as the only visible elements of the houses remaining. In Way Urang village, some houses were completely destroyed by the tsunami. A tsunami flow depth of 3.90 m was measured by one broken tree branch in the coastal area (Point F78 in the Appendix).

15
Maja Village's coastal area is a residential fishing village. A seawall was constructed to protect the houses from high waves generated from the Sunda Strait. Highest tsunami flow depth was 2.0 m from ground level (F70 and F71), measured at a house close to the seawall.
Impacts of the tsunami on Rajabasa area were worse than in Kalianda. We surveyed three villages in this sub-district, namely Way Muli, Kunjir and Batu Balak villages. Measured tsunami flow depths in this 20 area were between 2.0 m-4.5 m as can be seen in Fig. 10. This is a fishing community, and the majority of the coastal area is a residential area for the fishermen. Other related buildings, such as a shrimp hatchery, were also damaged by the tsunami.

20
Most buildings in the surveyed areas can be classified into two types, namely, confined masonry and wood. Wall components were brick. Roofs were wooden-framed with either tiled or zinc roofing.
Detailed investigation of building damage was performed at 98 sites including the remaining houses and one school. Seventy-three of them were non-engineered, lightly reinforced concrete houses or confined masonry (CM), and 25 were wooden timber. There were 16 CM-type houses as classified by DS0.

25
Impacts of tsunami debris were minor at houses near the coastline. Some tsunami debris was found to have contributed to house damage although most of the CM-type houses were mostly unaffected. Large debris was found at hotel complexes in Banten, where cars were carried by the tsunami flow and stopped by vegetation around the complexes. Damage to CM walls was mostly due to impulse or punching force produced by the tsunami flow. Unlike cases with tectonic events, shear-cracks on walls due to lateral 30 forces were not found. Damage to columns was observed as collateral failure of walls combined with impacts of hydrodynamic forces. Fig. 13 shows completely destroyed houses (DS4), where their types and material composition were known; these were also directly within tsunami flow depths as measured around the houses. Due to the limited number of surveyed houses, the fragility function was analysed using CM-type houses. A summary of the fragility functions analysis is given in  There were 25 wooden houses surveyed in the affected area. Due to the limited number of surveyed wooden houses, analysis of the damage was performed by classifying the damage as seen in Table 1.
Only three classifications of damage could be identified, namely, DS2, DS3 and DS4. Two houses could 10 be classified as DS2, seven houses were DS3, and 19 houses were DS4. Examples of the damage found at wooden houses due to tsunami impacts can be seen in Fig. 13. Based on the number of surveyed wooden houses, it was found that a tsunami flow depth higher than 2.0 m could completely wash away the houses, provided there was no debris in the flow. If the tsunami flow contained debris, the limit of the tsunami flow depth that could wash away a house was lower. In cases where the flow depths were

Characteristics of Building Damage
Damage to buildings and houses was more severe in Banten than in Lampung. Most of the damages were found in wall components. As a majority of buildings in Banten and Lampung were constructed with   southern Sumatra coasts, respectively. Starting the 10-day survey just two days after the tsunami offered the team a better opportunity to collect undisturbed tsunami evidence. Study conclusions are as follows: a. Tsunami flow depths were found to be relatively higher at Banten area than at Lampung. Arrival times were slightly faster in Banten than on the Lampung side although it is important to note that