How to identify tsunami prevention zones: a case study of Putuoshan Island in China

Tsunamis can cause massive damage and loss of life in coastal areas. This study uses Putuoshan Island as a case study to discuss ways to identify tsunami prevention zones based on the analysis of physical factors such as inundation, maximum wave amplitude, and tsunami flow. In the Ryukyu Trench Mw 9.0 tsunami scenario, the tsunami prevention zones of Putuoshan Island consist of land areas and marine areas. Our analysis identified two inundation areas in the western and southern Putuoshan Island as tsunami prevention zones. With the exception of the western marine area, the coastal waters of Putuoshan Island are also identified as tsunami prevention zones. The tsunami flow impact is greater in the waters of north and south Putuoshan Island. The results of this study could be used by emergency departments to take protective measures and thus to prevent the disastrous effects of tsunamis.


Introduction
Tsunamis are a relatively low-probability hazard, but they often cause substantial damage. The 2004 Indian Ocean Tsunami ( Wang and Liu 2006) and the 2011 Japan Tsunami (Wei et al. 2013;Wang et al. 2016) caused hundreds of thousands of deaths and massive devastation of infrastructure in coastal areas. When a tsunami leaves the deep water of the open ocean and enters shallow water near the coast, it slows down and the height of the wave increases. As a result, a tsunami that is small in deep water could grow to a height of tens of meters after reaching shallow water. Tsunamis are devastating because they carry a huge amount of energy (Zhao et al. 2012;Wang et al. 2021). A United Nations report estimated that about 60,000 people and US$ 4 billion in property are exposed to tsunamis each year globally (UNISDR 2015). It is expected that the devastation caused by tsunamis will increase over time due to population growth, population migration to coastal areas, and global climate change (Li et al. 2018).
A tsunami can flatten an entire town and wash away buildings, bridges, and cars. It surges in one direction and falls in the opposite direction, which is different from the unidirectional effects of flooding events. The inundation associated with a tsunami usually lasts tens of minutes (Exton et al. 2019). In the 2011 Japan tsunami, the inundation in Kesennuma lasted for about 25 min. The Sendai Plain was inundated for more than 5 km (Mori et al. 2011). Most tsunamis do not cause significant inundation, but coastal facilities can be severely damaged by the tsunami's strong currents.
A tsunami with a small amplitude may cause significant flow when it enters narrow inlets and ports (Allan et al. 2012;Gao et al. 2021). High-speed tsunami flow can destroy buildings and wash away ships. The strong flows of the 2011 Japan Tsunami damaged more than 200 small vessels and some dock facilities in Hawaii, with an estimated total damage of US$ 30 million (Dunbar et al. 2011;Fiedler et al. 2014). Observational data on tsunami flow has not been available until the last two decades. Some measurements of flow velocity have been found in the literature (Fritz et al. 2012;Wilson et al. 2012). Recently, research on tsunami forces and engineering design standards has increased (Yim et al. 2009). Nistor et al. (2011) used the smooth particle hydrodynamics (SPH) model to investigate the hydrodynamic force caused by a tsunami on circular cylinder and square cylinder structures. Yim et al. (2015) calculated the Tsunami Design of Coastal Infrastructure using the COMCOT model. Mutsuda et al. (2015) calculated the tsunami fluid force and tsunami impact pressure using numerical method. Yeh et al. (2014) analyzed the time-varying combined force caused by tsunami waves. The first peak force on the coastal structure is estimated to be 1.5 times the subsequent hydrodynamic force (Attary et al. 2017). Tsunami forces are affected by many factors, including inundation depth, slope of the bottom, bottom roughness, and distance from the coastline. The US Federal Emergency Management Agency divides the forces caused by tsunami waves into hydrostatic force, buoyant force, hydrodynamic force, impulsive force, and debris impact force (FEMA 2019).
The goal of this paper is to identify tsunami prevention zones by modeling potential tsunami events. Physical parameters such as tsunami inundation and tsunami flow are analyzed, and the method of using the results to identify tsunami prevention zones is discussed. Tsunami prevention zones can be used to prepare tsunami mitigation measures in advance.

Methodology and tsunami model
This paper uses modeling to analyze the tsunami hazard and determine where future tsunami disasters may occur. Two kinds of disastrous tsunami effects are considered-tsunami inundation on land and tsunami flow impact in coastal waters. The study area includes marine areas 3 km from the coastline, which is the main area for activities such as marine aquaculture and marine tourism. Figure 1 is the analysis framework for the tsunami prevention zones. There are two main considerations for research on prevention zones: (1) On the land areas, the tsunami inundation areas are classified as tsunami prevention zones. Objects on the ground within inundation areas including commercial facilities and housing, nuclear power plants, and industrial complexes may be affected by a tsunami to a certain extent. (2) In marine areas, the impact 1 3 of tsunami currents on coastal facilities is the main consideration. Tsunami forces were used to calculate the impact of tsunami flow. Facilities that could be affected by tsunami flows include ports, tourist places, and marine aquaculture operations.
The areas affected by tsunamis differ depending on the source of the tsunami. Therefore, an analysis of prevention zones should be based on a particular source and tsunami magnitude. For the purpose of this paper, four earthquakes were assumed to be the tsunami sources. The COMCOT model was used to calculate the physical parameters of the tsunamis. This model, developed by Cornell University, uses a leapfrog scheme on a staggered and nested grid system to solve both linear and nonlinear shallow water equations. It has been successfully used for the simulation of several historical tsunami events (Heidarzadeh and Satake 2014;Hou et al. 2016).
A three-level nested grid was used in the tsunami model to analyze tsunami prevention zones for Putuoshan Island. The first level of the model covers the Northwest Pacific Ocean and Putuoshan Island, while the third level covers only Putuoshan Island and its coastal waters. According to the distance between the tsunami sources and the Putuoshan Island, the total calculation time in the model was 15 h. In deep water, the tsunami wave is affected mainly by the Coriolis force and the pressure gradient, so the governing equation of the first level uses linear shallow water equations in spherical coordinates. When the tsunami reaches shallower water, nonlinear interaction increases. Therefore, nonlinear shallow water equations in Cartesian coordinates are used in the second and third levels. where η is the water fluctuation; (P, Q) stand for the volume flux in the X and Y direction; (φ, ψ) are the latitude and longitude; f is the Coriolis force; H = η + h represents the total water depth; g is gravity acceleration; R is the radius of the earth; (Fx, Fy) denote the bottom friction in the X and Y direction.
where n is the Manning's roughness coefficient. The grids settings of numerical model are shown in Table 1.
ETOPO2v2 2-arcminute data (NGDC 2006) was used for simulation at the first level, which covers an area from 0°N to 50°N and from 100°E to 160°E. The second level, which covers the entire archipelago where Putuoshan Island is located, uses SRTM 15 + 15-arcsecond bathymetry data (Tozer et al. 2019). The bathymetry data for the third level were derived from the fusion of measured bathymetric data with a resolution of 1/32 arcminutes.

Overview of Putuoshan Island
Putuoshan Island is located in the East China Sea, facing the Pacific Ocean (Fig. 2). Several academic studies have analyzed the potential tsunami hazard in the East China Sea (Wen et al. 2008;Hou et al. 2020;Ren et al. 2021Ren et al. , 2022. There are tourist areas, ports, and docks in the waters around Putuoshan Island. If a disastrous tsunami occurs, some important facilities may be damaged. Therefore, this paper uses Putuoshan Island as a case study to analyze tsunami prevention zones. Putuoshan Island is 8.6 km long from north to south and 3.5 km wide from east to west, with a coastline of 30 km. The terrain is steep in the northwest and gentle in the

Numerical calculation
In this paper, four tsunami scenarios that could affect Putuoshan Island were considered, including the Nankai Trough scenario, Japan Trench scenario, Ryukyu Trench scenario, and the Philippine Trench scenario. These four tsunami scenarios were simulated using COMCOT numerical model. The focal depth, strike angles, dip angles, and slip angles ( Table 2) are derived from historical data statistics and the Slab 1.0 (Hayes et al. 2012). A 9.0-magnitude earthquake in the Ryukyu Trench was assumed in the "Exercise Pacific Wave 11" tsunami exercise launched by the Intergovernmental Oceanographic Commission of UNESCO in November 2011 (UNESCO/IOC 2011). The magnitudes of the four tsunami scenarios were assumed to be 9.0. Four output points around Putuoshan Island were included in the numerical model to characterize the tsunami wave (Fig. 3). It can be seen from the numerical results (Fig. 4) that Ryukyu Trench scenario is the scenario that has the greatest impact on Putuoshan Island. Therefore, the Ryukyu Trench scenario was chosen to analyze the tsunami prevention zones in this study.
The maximum amplitude of Ryukyu Trench scenario can be seen from Fig. 4. Figure 5 shows the tsunami travel times of Ryukyu Trench scenario. As can be seen from the figures, the tsunami wave comes from east of Putuoshan Island. In the Ryukyu Trench 9.0 earthquake tsunami scenario, most of the tsunami amplitude near Putuoshan Island exceeds 3 m, with a maximum tsunami amplitude of 6.3 m.
The tsunami waves reach Putuoshan Island in 5.6 h and spread to all of the coastal areas of Putuoshan Island in 6 h. Figure 6 is the time series at four model output points. It shows that the maximum tsunami amplitude at each point is between 2.3 and 5.1 m and the first wave reaching the coast is not always the highest. The tsunami amplitudes at points b and d are relatively larger. The point b is facing the tsunami wave, and tsunami wave gathers in the bay. When the tsunami waves pass through the island, they diffract and split. Tsunami wave do not gather at point a and point c. After passing point a and point c, the tsunami waves gather at point d again. So the tsunami amplitude at point d is not small. 1 3

Identification of tsunami prevention zones
The result of the model was used to characterize the tsunami hazard and identify tsunami prevention zones in land and sea areas of Putuoshan Island. After a tsunami inundates land, both inundation and flow impact can cause disasters. Therefore, the inundation areas should be considered tsunami prevention zones. In the sea area, the tsunami flow hazard was established by calculating the velocity of tsunami flow and the tsunami hydrodynamic force.

Tsunami hazard analysis for land areas
Tsunami waves inundate two areas in western and southern Putuoshan Island. There is no inundation in the eastern part of the island due to the protection of seawalls. The inundation areas cover total area of 1.2 km 2 (Fig. 7). There are scenic spots, residential areas, hotels, and other important facilities in the inundation area. The inundation depth is 1.1-3.8 m. Tsunamis, with longer wavelengths than normal waves, carry more power for inundation, so they can travel further inland than normal waves. The inundation distance for Putuoshan Island is as far as 0.76 km from the coast. Since inundation can cause damage to objects on land, this paper classifies the inundation areas as tsunami prevention zones.

Tsunami hazard analysis for marine areas
As a tsunami approaches the shallow coastal area of Putuoshan Island from open sea, the tsunami flow hazard increases. The flow direction of the tsunami indicates the path the disaster will take, which is important information for tsunami prevention and mitigation. The tsunami flow is instantaneous and varies with time, so this paper studied the tsunami flow at specific moments. Figure 8 shows the tsunami flow direction off the coast of Putuoshan Island, where the tsunami wave flows from east to west. Due to the irregular shape of Putuoshan Island, the tsunami flows become very complex when the tsunami waves reach the shore. The tsunami waves spread around the north and south sides of Putuoshan Island to the western marine area of the island. Most of the tsunami flow damage is caused by the rapid movement of water. Rapid tsunami flows in coastal waters can cause damage to ships, docks, and other facilities. The velocity of the tsunami flow increases after tsunami waves propagate. The maximum tsunami flow velocity reaches a maximum value of 8.3 m/s (Fig. 9). The tsunami velocity in the eastern, southern and northern waters of Putuoshan Island is relatively higher than the velocity in the west.
In addition to velocity, water density and water depth also determine the tsunami flow hazard. Tsunami forces were calculated to assess the destructiveness of tsunami flow. Tsunami forces consist of hydrodynamic force, hydrostatic force, and buoyant force. The hydrodynamic force of the tsunami flow was calculated to determine the impact of tsunami flow off the coast of Putuoshan Island. Hydrodynamic force is imposed on buildings by water flowing against and around them. The numerical results from COMCOT model were used to investigate the hydrodynamic force. We assumed that the hydrodynamic force was acting on an object with a width of 1 m off the coast of Putuoshan Island. The equation for tsunami hydrodynamic force (FEMA 2012) is: where ρ s is the fluid density, including sediments. C d stands for drag coefficient, conservatively taken as C d = 2.0. B is the width of the structure in the plane normal to the direction of flow. h denotes the water depth of the tsunami flow. u is the flow velocity at the location of the structure.
The maximum value of tsunami hydrodynamic force at each moment was calculated and is shown in Fig. 10. The highest value is 443 kN. It should be noted that this result is used only to show the variation in tsunami hydrodynamic force at different locations off the coast of Putuoshan Island.
The maximum amplitude, flow velocity, and hydrodynamic force at the four output points were calculated and are shown in Table 3. As can be seen from this table, the flow velocity and hydrodynamic force off the eastern, southern, and northern coast of Putuoshan Island are higher than off the western coast.

Tsunami prevention zones
The tsunami modeling results were used to determine the locations of potential tsunami prevention zones for Putuoshan Island. Our analysis modified the modeling results in three ways.
(1) The lakes and rivers were removed in the inundation areas to determine the boundary of the prevention zones.
(2) The entire building group (such as scenic spots, residential areas, and so on) was considered as the tsunami prevention zone, if The tsunami prevention zones proposed in this paper are shown in Fig. 11. They can be considered priority areas for tsunami disaster prevention and mitigation measures. For land areas these measures include: avoiding new developments; protecting existing facilities through renovations; and developing tsunami evacuation plans. Tsunami prevention zones in marine areas include areas where the tsunami hydrodynamic force Fig. 8 Tsunami flow direction (a is the moment 5.45 h after the earthquake occurs, b is the moment 5.65 h, c is the moment 5.85 h, and d is the moment 6.05 h) exceeded 40 kN, which is an unbearable force for small fishing boats. Our modeling results show that all coastal waters of Putuoshan Island except the western marine area are in the tsunami prevention zone. Measures that can be taken in these areas include careful planning of maritime tourism projects, adjustment of marine aquaculture areas, and strengthening of ports.

Conclusions
This study uses Putuoshan Island as a case study to discuss a method of identifying tsunami prevention zones by assuming a tsunami worst-case scenario of a magnitude 9.0 earthquake in the Ryukyu trench. The results of the modeling identified tsunami prevention zones on the land and in the marine areas of Putuoshan Island. On the land, the inundation areas were identified as prevention zones. The total inundation area is 1.2 km 2 , with an inundation depth reaching 3.8 m. The tsunami inundation distance is up to 0.76 km Fig. 11 Tsunami prevention zones from the coast. In the marine areas, places where hydrodynamic force was greater than 40 kN were identified as prevention zones. The tsunami hydrodynamic force is higher on the north and south sides of the island, with a maximum value of 443 kN. The tsunami prevention zones indicate disaster-prone areas for tsunami disaster prevention and mitigation. Tsunamis are very dangerous with huge amounts of energy, so tsunami prevention and mitigation should be taken seriously. The suggested measures for tsunami disaster prevention include: (1) Building seawalls and other tsunami protection facilities; (2) Avoiding development or construction of important facilities in the prevention zones; (3) Increasing tsunami education and raising public awareness of tsunami disaster prevention.
Tsunami disaster prevention is difficult due to the sudden occurrence of tsunamis. Therefore, tsunami prevention measures should be made in advance. Tsunamis are not very dangerous in the deep sea, but are likely to cause disaster after reaching coastal areas. Tsunami prevention zones can be used to reduce the tsunami disaster through developing appropriate measures.
Each different tsunami source causes a different kind of tsunami disaster and requires distinct tsunami prevention zones. The location, magnitude, and focal mechanism of tsunami sources can influence the results of tsunami prevention zone analysis. The tsunami with highest probability of occurrence should be analyzed first for tsunami prevention zones.
Our analysis identified waters with a hydrodynamic force greater than 40 kN as tsunami prevention zones. These areas correspond roughly to areas where the velocity is greater than 3 m/s. However, the scientific classification criteria of hydrodynamic forces need to be discussed further. In addition to the hydrodynamic force considered in this study, tsunami load effects include several forces. The other tsunami forces need to be analyzed in the future.