geosciences Article
Systematic Evaluation of Different Infrastructure Systems for Tsunami Defense in Sendai City Kwanchai Pakoksung *, Anawat Suppasri and Fumihiko Imamura International Research Institute of Disaster Science, Tohoku University, 468-1 Aramaki-Aza, Aoba-ku, Sendai 980-0845, Japan;
[email protected] (A.S.);
[email protected] (F.I.) * Correspondence:
[email protected]; Tel.: +81-22-752-2090
Received: 5 February 2018; Accepted: 8 May 2018; Published: 10 May 2018
Abstract: The aim of this study is to assess the performances of different infrastructures as structural tsunami countermeasures in Sendai City, based on the lessons from the 11 March 2011, Great East Japan Tsunami, which is an example of a worst-case scenario. The tsunami source model Ver. 1.2 proposed by Tohoku University uses 10 subfaults, determined based on the tsunami height and the run-up heights measured for all tsunami affected areas. The TUNAMI-N2 model is used to simulate 24 cases of tsunami defense in Sendai City based on a combination of 5 scenarios of structural measures, namely, a seawall (existing and new seawall), a greenbelt, an elevated road and a highway. The results of a 2D tsunami numerical analysis show a significant difference in the tsunami inundations in the areas protected by several combinations of structures. The elevated road provides the highest performance of the single schemes, whereas the highest performance of the 2-layer schemes is the combination of an existing seawall and an elevated road. For the 3-layer scenarios, the highest performance is achieved by the grouping of an existing seawall, a new seawall, and an elevated road. The combination of an existing seawall, a new seawall, a greenbelt and an elevated road is the highest performing 4-layer scenario. The Sendai City plan, with a 5-layer scenario, reduces the tsunami inundation area by 20 sq. km with existing structural conditions. We found that the combination of an existing seawall, a greenbelt, an elevated road and a highway (a 4-layer scheme) is the optimum case to protect the city against a tsunami similar to the 2011 Great East Japan Tsunami. The proposed approach can be a guideline for future tsunami protection and the evaluation of countermeasure schemes. Keywords: Great East Japan Tsunami 2011; numerical modeling; tsunami simulation
1. Introduction A magnitude 9.0 earthquake generated a huge tsunami in the Tohoku region of Japan on 11 March 2011, affecting human lives and causing economic losses. The affected area was recorded as approximately 561 sq. km and covered 3 prefectures (Iwate, Miyagi, and Fukushima) along the Pacific coast of Japan [1,2]. The maximum recorded tsunami run-up height was approximately 40 m located in Iwate Prefecture [3]. The fatality number reported by the National Police Agency was approximately 15,866 people, with approximately 2946 people missing [4]. The number of buildings and houses that collapsed or were washed away was approximately 130,411 [5,6]. The economic losses associated with the disaster were approximately 25 trillion yen, which was 30% of the national income of Japan in 2011 [7]. After the event occurred, approximately 82,000 people lost their houses [8]. Then, the recovery efforts of the central and local government began, including the conceptual planning of the reconstruction of infrastructures including dike renovation, transportation, land-use zoning and urban relocation.
Geosciences 2018, 8, 173; doi:10.3390/geosciences8050173
www.mdpi.com/journal/geosciences
Geosciences 2018, 8, 173
2 of 25
Tsunami modeling is an important aspect prior to protection planning and includes using models such as the Cornell Multi-grid Coupled Tsunami Model (COMCOT), the Method of Splitting Tsunami (MOST) and Tohoku University’s Numerical Analysis Model for Investigation of Near-field Tsunami, No.2 (TUNAMI-N2), which are based on the non-linear shallow water equation. The COMCOT, developed by Liu et al. (1998) uses the leap-frog finite difference method on a staggered mesh scheme [9,10]. Titov and Gonzalez (1997) developed the MOST model, which splits a 2D system into 1D systems for the early prediction of a tsunami [11]. The TUNAMI-N2 model, developed by Imamura (1995) [12] to model the tsunami process, has been widely utilized in a number of research and assessments though the TIME (Tsunami Inundation Modeling Exchange) project. In previous research, several studies on tsunami modeling were performed. For example, the Arabian Sea tsunami, generated by the 1945 Makran earthquake, has been simulated by the TUNAMI-N2 model [13], based on several historical earthquake hypotheses [14]. The vulnerability of the Car Nicobar coast to tsunami hazards has also driven the use of TUNAMI-N2. A study was performed on the 2004 Indian Ocean tsunami to understand the effects of wave run-up in the Koodankulam region of the Tamil Nadu Coast using TUNAMI-N2. The 2004 tsunami was simulated using 28 scenarios with a run-up wave range between 1.30–3.54 m [15]. In addition, the fragility curve of the 2004 tsunami in Thailand was estimated by using numerical modeling and satellite remote sensing. The tsunami modeling of the fragility curve provided by the TUNAMI-N2 model demonstrated reasonable accuracy [16]. The TUNAMI-N2 model has been applied to the 2011 Tohoku-oki tsunami to understand inundation characteristics and sedimentation processes [17,18]. The results show that the flood depth ranges between 2.4–6.0 m and the flow velocity ranges between 3.4–6.2 m/s; furthermore, the sand layer is distributed on land approximately 2.8 km from the coastline. For the early warning system, the TUNAMI-N2 was modified using parallel programming to reduce the computation time such that the computation time for this model in this study is 75 times faster than real time [19]. The perfect early warning system with a non-structural scheme is difficult to obtain and to use as a tsunami protection method. Combinations of hard and soft structures perform better than single schemes [20]. The concept of multi-layered safety has been proposed in the distributed reconstruction in the Tohoku area after the 11 March 2011 tsunami [21,22]. It was realized that there was a need for multiple layers of safety to ensure the protection of coastal residents. Research in the field of disaster risk management usually focuses on the perceived shared benefits of coastal structures that can increase the safety of people against coastal hazards [21,23]. Coastal structures are categorized as offshore and onshore structures. Offshore structures include breakwaters, jetties and groins, whereas onshore structures include seawalls, dikes and roads. However, there have been few studies that define and measure the performance of coastal structures in tsunami protection. The performance of a breakwater in reducing tsunami inundation was measured in [24]. The study suggests that a breakwater cannot reduce the flood inundation; however, the arrival time of a tsunami wave is delayed by a breakwater, based on the 2011 tsunami event in Japan. For inland structures, the protection efficiency of a multi-layered structure was evaluated for the 2004 Indian tsunami in Sri Lanka [25]. The multi-layered structure (revetment seawall and railway embankment) demonstrated a potential to reduce damage by approximately 30% compared with existing conditions. Sendai City was one of the impact areas of the 2011 Great Tsunami, and the prefecture and local governments have a plan to develop multilayered infrastructure systems to protect the city from future tsunamis under the limitations and uncertain conditions of funding [26,27]. This idea was developed after the existing highway located approximately 4–5 km from the sea helped to protect the city from the 2011 tsunami by chance. The conceptual plan is a combination of existing and new infrastructure systems such as seawall reconstruction, land use zoning, and construction of new roads. As shown in Figure 1a, the plan is to reduce the tsunami flood depth to less than 2 m in the newly planned residential area [28]. The reconstruction plan of Sendai City, as presented in Figure 1b, uses a combination of several features such as seawalls, forests, parks, new elevated roads and existing highways to reduce the disaster losses. The planned view of the multi-structural system of Sendai
Geosciences 2018, 8, 173
3 of 25
City is based on a 7.2 m seawall reconstruction (originally approximately 6.0 m) and a 6.0 m elevated 2018, 8, x FOR PEER REVIEW 3 of 25 road.Geosciences As of 2017, the seawall reconstruction has already been completed, while the elevated road is still under The protection efficiency of theofelevated roadroad andand seawall ininthis road is construction. still under construction. The protection efficiency the elevated seawall thisplan plan was evaluated for the 2011 tsunami scenario [2]. The study reveals that the 2011 tsunami would flow was evaluated for the 2011 tsunami scenario [2]. The study reveals that the 2011 tsunami would flowover the top ofthe a 7.2 and 6.0and m elevated road. road. over topm ofseawall a 7.2 m seawall 6.0 m elevated
From Koshimura et al., 2014 [2], and Sendai City, 2011 [28] (a )
(b ) Figure 1. Actual plans to develop Sendai City tsunami disaster protection in the future: (a) Conceptual
Figure 1. Actual plans to develop Sendai City tsunami disaster protection in the future: (a) Conceptual plan to develop Sendai City for protection against tsunamis, presented as a profile; (b) Location of plan to develop Sendai City for protection against tsunamis, presented as a profile; (b) Location of Sendai City and the line of tsunami prevention facilities. Sendai City and the line of tsunami prevention facilities.
For Sendai City, the reconstructed and new structures have already been evaluated for the 2011 tsunami scenario previously mentioned, thestructures study considers the reconstructed seawall For Sendai City,asthe reconstructed andbut new have only already been evaluated for the and the new elevated road, without the greenbelt and economic approach. Currently, few studies 2011 tsunami scenario as previously mentioned, but the study considers only the reconstructed seawall havenew beenelevated performed on investigating determining explicit approach. benefits of multi-layered coastal and the road, without theand greenbelt and the economic Currently, few studies structures, which would involve a comprehensive economic estimate of their disaster prevention have been performed on investigating and determining the explicit benefits of multi-layered coastal value. To address this gap, this study aims to assess the tsunami prevention of a multi-infrastructure structures, which would involve a comprehensive economic estimate of their disaster prevention value. system, with an economic approach to selecting the optimum combination of structures. To address this gap, this study aims to assess the tsunami prevention of a multi-infrastructure system, The objective of this study is to understand the tsunami behavior on different terrains to obtain with new an economic to selecting the optimum combination of structures. insights onapproach the damage for the previously mentioned multi-infrastructure system. The tsunami The objective of this study is to understand the tsunami behavior on different terrainsmodel to obtain modeling of the Sendai Plain was reestablished and studied using a numerical mathematical new called insights on the damage for theapplies previously mentioned system. The tsunami TUNAMI-N2. The model the shallow watermulti-infrastructure theory using an initial water level and seafloor estimated using fault parameters. Theusing numerical model results were then modeling of deformation the Sendai Plain was reestablished and studied a numerical mathematical model
Geosciences 2018, 8, 173
4 of 25
called TUNAMI-N2. The model applies the shallow water theory using an initial water level and seafloor deformation estimated using fault parameters. The numerical model results were then calibrated with the field measurements. The tsunami hazard maps for each component of the infrastructure system for protecting against tsunamis were analyzed with respect to the tsunami impacts on the Sendai City coastal area of the Sendai Plain generated by the 2011 Tohoku earthquake. The findings of this paper present the significance of infrastructure system management for protecting against future tsunami disasters. The evaluation of these plans is generally academic research to raise awareness of loss reduction for future disasters and has application in other areas. The paper is organized as follows. Section 2 presents the detailed study method, including tsunami modeling. The results and discussion are given in Section 3, with consideration to the tsunami inundation map and evaluation of the optimum scenario. Section 4 presents conclusions. 2. Methods This section presents a tsunami model that is used to evaluate countermeasure tsunami defense scenarios composed of multiple levels of infrastructure. It consists of four subsections: (i) the numerical tsunami simulation, (ii) the tsunami source model, (iii) the calibration and validation of the tsunami model, and (iv) a tsunami simulation of the multi-infrastructure scenario. 2.1. Numerical Tsunami Simulation To obtain the tsunami inundations for different infrastructure-system conditions, a numerical tsunami simulation was conducted with the TUNAMI-N2 model [12]. The simulation was run on a computational domain using a nesting grid system, from large areas to small areas, along the coastline of Sendai City. The TUNAMI-N2 model was first developed at Tohoku University to model tsunami propagation and inundation on land and operates using the nonlinear theory of the shallow water equation, which is solved using a leap-frog staggered-grid finite difference scheme. The nonlinear shallow water equation is shown by Equations (1)–(3), wherein the finite difference method is applied to run the nonlinear equation with the bottom friction represented by Manning’s roughness coefficient. ∂η ∂M ∂N + + = 0, ∂t ∂x ∂y MN ∂η gn2 p + gD + 7 M M2 + N 2 = 0, D ∂x D3 2 ∂N ∂ NM ∂ N ∂η gn2 p + + + gD + 7 N M2 + N 2 = 0, ∂t ∂x D ∂y D ∂y D3
∂ ∂M + ∂t ∂x
M2 D
+
∂ ∂y
(1)
(2) (3)
where η is the water level, M and N are the fluxes of water in the x and y directions, D is the total depth, g is gravitational acceleration, and n is Manning’s roughness coefficient. The preparation process was performed for the bathymetry data for the tsunami propagation and inundation simulations. The tsunami model was run on five computational domains as shown the nested grid system in Figure 2a; the data were provided by Geospatial Information Authority of Japan (GSI). The projection of collected data is based on the Cartesian coordinates from the Japanese Geodetic Datum 2000 and the Tokyo Peil (T.P.) datum for the horizontal and vertical references, respectively. The fixed ratio of the two domains is one-third (see Figure 2b). In the smallest domain, the inundations were calculated for the proposed protective structure, which was inserted into the bathymetry by using GIS techniques, as presented in Figure 2c. The GIS task was done by the Quantum GIS (QGIS) software on Conversion tool and Raster calculator tool. The Conversion tool used to convert the vector format (hypothesis structure line) to the rater format as same as the format of bathymetry data. The combination between hypothesis structure in rater format and existing bathymetry is done by the Rater calculator tool. Note that the simulation without a greenbelt area is run with a constant Manning coefficient of 0.025.
Geosciences 2018, 8, 173 Geosciences 2018, 8, x FOR PEER REVIEW
5 of 25 5 of 25
(a)
(b) Figure 2. Cont.
Geosciences 2018, 8, 173
6 of 25
Geosciences 2018, 8, x FOR PEER REVIEW
6 of 25
(c) Figure 2. 2. Bathymetry and inundation. inundation. (a) (a) The Figure Bathymetry data data to to compute compute tsunami tsunami propagation propagation and The largest largest modeling modeling domain with a 1215 m grid size. The black boxes represent the grid for smaller domains; (b) The domain with a 1215 m grid size. The black boxes represent the grid for smaller domains; (b)ratios The of different grid sizes; (c) The method of inserting the hypothesized infrastructure into the existing ratios of different grid sizes; (c) The method of inserting the hypothesized infrastructure into the mesh. mesh. existing
2.2. Tsunami Source Model 2.2. Tsunami Source Model An important variable for simulating tsunami propagation and inundation is the initial seafloor An important variable for simulating tsunami propagation and inundation is the initial seafloor deformation that is computed from the fault modeling of the tsunami source model. Because of the deformation that is computed from the fault modeling of the tsunami source model. Because of the complexity and uncertainty of the fault rupture processes, this study estimated the initial water level complexity and uncertainty of the fault rupture processes, this study estimated the initial water level based on a rectangular fault model and assumed that the change in the sea surface is the same as the based on a rectangular fault model and assumed that the change in the sea surface is the same as the seafloor deformation. The rectangular faults are assumed to have a fast movement toward the sea seafloor deformation. The rectangular faults are assumed to have a fast movement toward the sea surface that is only a vertical displacement [29]. The fault parameter of the 2011 Great Pacific Coast surface that is only a vertical displacement [29]. The fault parameter of the 2011 Great Pacific Coast Japanese earthquake is based on the Tohoku University model [30]. The Tohoku fault model (V. 1.2) Japanese earthquake is based on the Tohoku University model [30]. The Tohoku fault model (V. 1.2) establishes 10 subfaults. Figure 3 presents the initial water level distribution generated by the Tohoku establishes 10 subfaults. Figure 3 presents the initial water level distribution generated by the Tohoku University model based on the Okada formula [29] where the maximum water level or uplift is about University model based on the Okada formula [29] where the maximum water level or uplift is about +14.3 m T.P., while −4.7 m T.P. is the minimum water level or subsidence. The T.P. datum is the Tokyo +14.3 m T.P., while −4.7 m T.P. is the minimum water level or subsidence. The T.P. datum is the Tokyo Peil datum referenced by the sea water level in Tokyo bay [1]. Peil datum referenced by the sea water level in Tokyo bay [1].
Geosciences 8, x FOR PEER REVIEW Geosciences 2018,2018, 8, 173
7 of 257 of 25
Figure 3. Initial sea floor deformation from the tsunami source based on the Tohoku University Model V. 1.2. Figure 3. Initial sea floor deformation from the tsunami source based on the Tohoku University Model V. 1.2.
2.3. Calibration and Validation of Tsunami Model
2.3. Validationand of Tsunami Model collected from the survey data taken from the GNSS TheCalibration tsunami and inundation flow depth
(Global Navigation System) measurements afterfrom the 2011 Greatdata Pacific Coast Tsunami The tsunamiSatellite inundation and flow depth collected the survey taken fromJapan the GNSS Navigation System) measurements the 2011 Great PacificGroup Coast Japan were(Global collected by TTJS Satellite (The 2011 Tohoku Earthquakeafter Tsunami Joint Survey) [31]. Tsunami The tsunami were collected bydetermined TTJS (The 2011 Tohoku Earthquake Tsunami Joint Survey) Group The tsunamidata run-up was further using a numerical model since insufficient field[31]. measurement was The further determined using amodel numerical sinceusing insufficient field measurement datasuch wererun-up collected. numerical tsunami was model calibrated a performance parameter, were collected. The numerical tsunami model was calibrated using a performance parameter, such as the geometric mean K or geometric standard deviation k proposed by Aida [32], to reproduce as the geometric mean 𝐾 or geometric standard deviation 𝑘 proposed by Aida [32], to reproduce the tsunami inundation and flow depth data. The K and k values were chosen for the calibration the tsunami inundation and flow depth data. The 𝐾 and 𝑘 values were chosen for the calibration of of tsunami run-up heights. The K value was derived from the mean of the K and k values, which tsunami run-up heights. The 𝐾 value was derived from the mean of the 𝐾 and 𝑘 values, which refersrefers to the or or variances ofobserved observedand and computed data. Japan to deviations the deviations variancesfrom fromthe theproportion proportion of computed data. The The Japan Society of Civil Engineering (JSCE) recommends values of 0.95 < K < 1.05 and k < 1.45 for the model Society of Civil Engineering (JSCE) recommends values of 0.95 < 𝐾 < 1.05 and 𝑘 < 1.45 for the results to achieve agreement” when evaluating the tsunami and propagation models [33]. model results “good to achieve “good agreement” when evaluating the source tsunami source and propagation According the According K and k values above,mentioned the simulation requires further calibration methods modelsto[33]. to thementioned 𝐾 and 𝑘 values above, the simulation requires further calibration methods to find a “good agreement” model because the proposed model was based on data to find a “good agreement” model because the proposed model was based on field measurement from all along the Tohoku coast while this study focuses only on Sendai City. The parameters K and k are determined as shown below: 1 n log K = log Ki (4) n i∑ =1
Geosciences 2018, 8, x FOR PEER REVIEW
8 of 25
field measurement data from all along the Tohoku coast while this study focuses only on Sendai City. The parameters 𝐾 and 𝑘 are determined as shown below: Geosciences 2018, 8, 173
log 𝐾 =
1 𝑛
log 𝐾
8 of 25
(4)
s
1 n 2 − (log K )2 (5) log k = 1 ∑ (𝐾log) K−i )(log (5) n i(log log 𝑘 = 𝐾) = 1 𝑛 xi Ki = (6) 𝑥 yi 𝐾 = (6) 𝑦 where xi and yi are the observed and simulated data, respectively, at point i. where 𝑥 and 𝑦 are the observed and simulated data, respectively, at point 𝑖. The calibration method, as recommended in this study, is meant to improve the initial water level The calibration method, as recommended in this study, is meant to improve the initial water using the mean ratio bias correction method. The corrected data were first adjusted by finding the level using the mean ratio bias correction method. The corrected data were first adjusted by finding mean ratio between the recorded and computed data [34–36], as shown in Equation (7). the mean ratio between the recorded and computed data [34–36], as shown in Equation (7). R (7) (7) Ri 𝑅== ̅ ·∙S𝑆i , , S where 𝑅 is the mean of the observed data, 𝑆̅ is the mean of the simulated data, 𝑅 is the corrected where R is𝑆the of the observed is theismean of theusing simulated data, Ri is the method corrected data, data, and is mean the simulated data. Thedata, ratioSvalue calculated the cross validation [37], as and Si is simulated The ratio value is calculated thevalue crossisvalidation as shown inthe Figure 4. The data. advantage of this technique is that using the bias removed method from the[37], mean, shown in Figure 4. The advantage of this technique is that the bias value is removed from the mean, whereas its disadvantage is failing to correct the time series data. whereas its disadvantage is failing to correct the time series data.
Figure 4. Method for calibrating the initial water level based on the mean ratio algorithm. Figure 4. Method for calibrating the initial water level based on the mean ratio algorithm.
2.4. Tsunami Simulation 2.4. Tsunami Simulation of of the the Multi-Infrastructure Multi-Infrastructure Scenario Scenario Determining the the impact impact of of multiple disasters is is the Determining multiple structures structures on on preventing preventing tsunami-related tsunami-related disasters the target of this study. The structural system shown in Figure 1 as part of the Sendai City plan has five target of this study. The structural system shown in Figure 1 as part of the Sendai City plan has five components: an existing existing seawall, seawall, aa reconstructed reconstructed seawall, seawall, aa greenbelt greenbelt area area (park (park and and coastal coastal forest), forest), components: an an elevated road, road, and andan anexisting existinghighway. highway.The Thegreenbelt greenbeltarea areawas was assumed include grass areas an elevated assumed toto include grass areas (n (n = 0.030) in parks and pine trees (n = 0.160) in coastal forests [38]. Based on the components of = 0.030) in parks and pine trees (n = 0.160) in coastal forests [38]. Based on the components of the the existing and reconstructed structures, the scenario of the simulation is divided into six schemes: existing and reconstructed structures, the scenario of the simulation is divided into six schemes: (1) (1) no-structure, (2) single structure, 2 components, 3 components, 4 components and5 no-structure, (2) single structure, (3) 2(3) components, (4) 3(4) components, (5) 4 (5) components and (6) (6) 5 components. The no-structure condition is used to model the actual terrain without any tsunami components. The no-structure condition is used to model the actual terrain without any tsunami protection For the the single-structure single-structure condition, condition, one one preventative preventative structure structure is is selected protection structures. structures. For selected and and input into the numerical tsunami model using the structure-insertion method previously input into the numerical tsunami model using the structure-insertion method previously described described in Using the the condition condition of of an an inundation inundation depth depth of of less less than than two two meters meters behind behind the the road, road, an in Figure Figure 2c. 2c. Using an elevated road was elevated road was suggested suggested for for the the protection protection of of Sendai Sendai City City against against tsunamis tsunamis [2]. [2]. This This condition condition can be used to reveal the performance of each preventative structure. The scenario of 2 components of preventative structures is used to reveal the performances of a combination of structures, such as the case of a seawall and a highway, as mentioned by the existing case in the calibration dataset.
Geosciences 2018, 8, 173
9 of 25
In the time line of development, the new and renovated preventative structures are contracted as 3-, 4-, and 5-component scenarios. The scenario of 5 components of preventative structures is the fullest extent of Sendai City’s ability to protect against tsunamis in the future. This scheme has 2 options in which the tsunami wave either can or cannot overflow the elevated road. These schemes are used to evaluate transportation safety during tsunami occurrences in the future. To understand the effects of the different protective structures during the progress of a tsunami wave, 24 sets of bathymetry (as shown in Table 1) were established and applied in the inundation simulation. The results of the 24 cases are presented according to the inundation depth to reveal their performance in protecting against tsunamis. Table 1. Scenarios for the simulation of the tsunami impact and the corresponding multi-condition protective structures. Scenario
Case
Description
0 layer
01
Non-structure
1 layer
02 03 04 05
Existing seawall Greenbelt Elevated road Existing highway
2 layers
06 07 08 09 10 11 12
Existing seawall + New seawall Existing seawall + Greenbelt Existing seawall + Elevated road Existing seawall + Existing highway Greenbelt + Elevated road Greenbelt + Existing highway elevation Elevated road + Existing highway
3 layers
13 14 15 16 17 18
Existing seawall + New seawall + Greenbelt Existing seawall + New seawall + Elevated road Existing seawall + New seawall + Existing highway Existing seawall + Greenbelt + Elevated road Existing seawall + Elevated road + Existing highway Greenbelt + Elevated road + Existing highway
4 layers
19 20 21 22
Existing seawall + New seawall + Greenbelt + Elevated road Existing seawall + New seawall + Greenbelt + Existing highway Existing seawall + New seawall + Elevated road + Existing highway Existing seawall + Greenbelt + Elevated road + Existing highway
5 layers
23 24
Existing seawall + New seawall + Greenbelt + Elevated road (overflow) + Existing highway Existing seawall + New seawall + Greenbelt + Elevated road (no-overflow) + Existing highway
3. Results and Discussion The performances of multilayered structures for tsunami mitigation in Sendai City were investigated by comparing the results of different geometries, with and without preventative structures. The TUNAMI-N2, a numerical tsunami model, uses a tsunami source model based on Tohoku University’s fault model (V. 1.2) of the 2011 Great Tsunami. The numerical model is calibrated with bias correction of the mean ratio algorithm. The simulation results with the corrected initial water level is evaluated by the K (kappa) coefficient. In addition to the actual structures, their heights are represented in the terrain data, and the hypothetical structures were recommended by the plan to protect Sendai City against tsunamis. The existing and planned structures with the highest crests were evaluated in order to investigate the effects of the overtopping of the structures on the behaviors of the tsunami flows. The last case is an enclosed structure that is hypothesized to be fully protected by multiple structures to examine the effects of implementing multilayer infrastructures to protect against large tsunamis. Finally, a demand/supply curve is used to evaluate the optimum combination of structures by using the intersected region between the demand line and the supply line.
Geosciences 2018, 8, x FOR PEER REVIEW
10 of 25
Geosciences 2018, 8, 173
10 of 25
3.1. Calibration and Validation Results
The tsunami run-up height from the numerical simulation based on the Tohoku University fault 3.1. Calibration and Validation Results model and bathymetry data with existing structures present during 2011 (Case 09) was calibrated Themeasurement tsunami run-up height from theinnumerical on the University using field data, as shown Figure 5.simulation Figure 5abased presents a Tohoku comparison of thefault model and bathymetry with existing present during 2011Information (Case 09) wasAuthority calibratedofusing inundation extent with thedata inundation line structures provided by the Geospatial field measurement as shown in Figure 5. Figure 5a presents a comparison thewherein inundation extent Japan [39]. Scatter plotsdata, of the observed and simulated data are shown in Figureof5b, 𝐾 is inundation line provided by the Geospatial Information Authority of Japan [39]. Scatter plots 0.874with andthe 𝑘 is 1.359. of the observed and simulated data are shown in Figure 5b, wherein K is 0.874 and k is 1.359.
(a ) Figure 5. Cont.
Geosciences 2018, 8, 173 Geosciences 2018, 8, x FOR PEER REVIEW
11 of 25 11 of 25
(b ) Figure Figure5.5.Validation Validationofofthe thewater waterlevel levelresulting resultingfrom fromthe theexisting existinginitial initialwater waterlevel levelfor forthe thetsunami tsunami source achieved using using the thetsunami tsunaminumerical numericalmodel modelbased basedon sourcescenarios: scenarios:(a) (a)Tsunami Tsunamiinundation inundation extent extent achieved onthe the Tohoku University Model 1.2 (Existing (b) Scatter plot between the observed and Tohoku University Model V. 1.2V.(Existing case); case); (b) Scatter plot between the observed and simulated simulated (existing case) data at the maximum water level. The modeled results reveal that K is (existing case) data at the maximum water level. The modeled results reveal that K is approximately approximately 0.874 and k is approximately 1.359. 0.874 and k is approximately 1.359.
In this study, the observed water level data have a mean value of approximately 4.314 m, In this study, the observed water level data have a mean value of approximately 4.314 m, whereas whereas the existing runs have a mean value of approximately 4.749, indicating that the simulation the existing runs have a mean value of approximately 4.749, indicating that the simulation results results overestimated the observed data. The bias correction parameter is approximately 0.908 and overestimated the observed data. The bias correction parameter is approximately 0.908 and was used was used to improve the existing initial water level. The bias parameter was multiplied with the to improve the existing initial water level. The bias parameter was multiplied with the spatial existing spatial existing initial water level. The improved initial water levels have a maximum value of initial water level. The improved initial water levels have a maximum value of approximately +12.6 m approximately +12.6 m and a minimum of approximately −4.1 m. Comparing the corrected data with and a minimum of approximately −4.1 m. Comparing the corrected data with the existing data, the existing data, the maximum increases by approximately 0.6 m, while the minimum decreases by the maximum increases by approximately 0.6 m, while the minimum decreases by approximately approximately 1.7 m. The modified initial water level is input into the TUNAMI-N2 model again for 1.7 m. The modified initial water level is input into the TUNAMI-N2 model again for the inundation the inundation simulation, and the simulated results were used for the validation, as shown in Figure 6. simulation, and the simulated results were used for the validation, as shown in Figure 6. Figure 6a Figure 6a shows the validation of the inundation extents with the inundation line, while the scatter shows the validation of the inundation extents with the inundation line, while the scatter plot of the plot of the observations and simulations is shown in Figure 6b using a 𝐾 of 0.957 and a 𝑘 of 1.339, observations and simulations is shown in Figure 6b using a K of 0.957 and a k of 1.339, which agree well which agree well with the mentioned form of the JSCE. The evaluation of the simulation results in a with the mentioned form of the JSCE. The evaluation of the simulation results in a spatial scheme. For a spatial scheme. For a detailed evaluation, Figure 7 shows five cross sections of the simulated terrain detailed evaluation, Figure 7 shows five cross sections of the simulated terrain inundation. Sections A inundation. Sections A and E represent the outside of the protected area to the north and south of and E represent the outside of the protected area to the north and south of Sendai City, respectively. Sendai City, respectively. The other sections, Sections A, B, and C, are in the reconstruction area. The The other sections, Sections A, B, and C, are in the reconstruction area. The evaluations of the sections evaluations of the sections reveal that the corrected model is closer to the field measurement data reveal that the corrected model is closer to the field measurement data than the existing model. than the existing model.
Geosciences 2018,2018, 8, x FOR Geosciences 8, 173PEER REVIEW
12 of 12 25of 25
(a) Figure 6. Cont.
Geosciences 2018, 8, 173 Geosciences 2018, 8, x FOR PEER REVIEW
13 of 25 13 of 25
(b) Figure 6. 6. Validation Validation of of the the water water level level resulting resulting from from the the calibrated calibrated initial initial water water level level of of the the tsunami tsunami Figure source scenario. scenario. (a)(a)Tsunami Tsunamiinundation inundation extent achieved with calibrated initial water level; source extent achieved with thethe calibrated initial water level; (b) (b) Scatter between observed and simulated data (calibratedinitial initialwater waterlevel) level)at atthe themaximum maximum Scatter plotplot between thethe observed and simulated data (calibrated K is approximately 0.957 and kk is approximately 1.339. water level. The modeled results reveal that K 1.339.
3.2. Simulation for for Each Each Scenario Scenario 3.2. Inundation Inundation Results Results of of the the Tsunami Tsunami Simulation The forfor tsunami protection were evaluated by the The performances performancesofofthe themultilayered multilayeredstructures structures tsunami protection were evaluated by inundation area,area, flow depth, velocity and arrival The simulated results based on the 2011on Great the inundation flow depth, velocity and time. arrival time. The simulated results based the Tsunami derived 24 cases of cases different geometric scenarios of tsunami protection. The 2011 Greatwere Tsunami werefrom derived from 24 of different geometric scenarios of tsunami protection. inundation results for some cases areare shown in in Figure 8 as a aspatial The inundation results for some cases shown Figure 8 as spatialmap. map.The Thenumber numberof ofinundation inundation areas areas at at each each depth depth and and the the ranges ranges of of the the areas areas are are shown shownin inTable Table 22 and and are are separated separatedinto into55classes. classes. These classes are arebased basedononthethe prevention represented in Figure and the inundation area These classes prevention areaarea represented in Figure 1, and1,the inundation area within within the prevention boundary is shown in 9Figure 9 with each combination scenario. the prevention boundary is shown in Figure with each combination scenario. The firstscenario scenariois is a scheme of 0 layers, presenting a no-structure tsunami The first a scheme of 0 layers, presenting a no-structure approachapproach to tsunamitoprotection, protection, wherein the single case is termed case 01. Case 01 is the base case used to evaluate the wherein the single case is termed case 01. Case 01 is the base case used to evaluate the protective protective structure The of isthis caseon is the based on theterrain existing terrain model, which structure cases. Thecases. terrain of terrain this case based existing model, which originally originally includedstructures existing such structures such as andbut highways, all ofstructures the existing included existing as seawalls andseawalls highways, all of thebut existing are structures are removed using GIS techniques before modeling to reveal the actual tsunami removed using GIS techniques before modeling to reveal the actual tsunami behavior on abehavior natural on a natural terrain, is athe same as a no-protection The results of case 01 show a total terrain, which is the which same as no-protection scheme. Thescheme. results of case 01 show a total inundation inundation area of approximately 28.56 sq. km in the prevention area, of which the inundation depth area of approximately 28.56 sq. km in the prevention area, of which the inundation depth is mostly in is mostly range of approximately the rangein ofthe approximately 2.0–4.0 m. 2.0–4.0 m. The singletsunami tsunamiprotection protection scheme the second scenario. It is on a component structural The single scheme is theissecond scenario. It is based on based a structural component of Sendai City, either planned or existing, and has a hypothetical structure of of 4 of Sendai City, either planned or existing, and has a hypothetical structure of one of 4 cases: aone seawall, cases: a seawall, a greenbelt, road, a highway. The objective this scheme is to reveal a greenbelt, an elevated road,an or elevated a highway. Theor objective of this scheme is toof reveal the performance of the of the each component of structures the countermeasure structures inand mitigating tsunamis, and eachperformance component of countermeasure in mitigating tsunamis, their performances are their performances arenonstructural evaluated against the nonstructural scheme. In summary, area evaluated against the scheme. In summary, the inundation area ofthe thisinundation scenario varies of this20.0–27.0 scenario sq. varies 20.0–27.0 with an average of approximately 24.0 sq. km. from km from with an averagesq. ofkm approximately 24.0 sq. km.
Geosciences 2018, 8, 173 Geosciences 2018, 8, x FOR PEER REVIEW
14 of 25 14 of 25
(a) Figure 7. Cont.
Geosciences 2018, 8, 173 Geosciences 2018, 8, x FOR PEER REVIEW
15 of 25 15 of 25
(b) Figure 7. 7. Validation of the the simulation simulation results, results, existing existing and and calibrated, calibrated, in in aa section. section. (a) (a) Location Location of of Figure Validation of Sections A–E, presenting a comparison of the observed and simulated (existing and calibrated) Sections A–E, presenting a comparison of the observed and simulated (existing and calibrated) results; results; (b) Comparison of theofresults of the observations and simulations in the cross section. (b) Comparison of the results the observations and simulations in the cross section.
Geosciences 2018, 8, 8, 173 x FOR PEER REVIEW Geosciences 2018,
16 16of of 25 25
(a)
(b)
(c)
(d)
Figure 8. Some Figure 8. Some examples examples of of the the inundated inundated regions regions for for different different sets sets of of multiple multiple prevention prevention structure structure conditions: (d) case case 23. 23. conditions: (a) (a) case case 01; 01; (b) (b) case case 09; 09; (c) (c) case case 22; 22; and and (d)
Geosciences 2018, 8, 173 Geosciences 2018, 8, x FOR PEER REVIEW
17 of 25 17 of 25
Table2.2. Inundation Inundation area area inside inside the the prevention prevention area area for for each each depth depth range range resulting resulting from from different different Table multi-structural conditions. multi-structural conditions. Scenario Case (Layer) Scenario (Layer) 0 01 02 0 03 1 04 105 06 07 08 2 09 210 11 12 13 14 15 3 316 17 18 19 20 4 421 22 23 5 524
Inundation Area in Each Flow Depth Range, sq. km 0.01–2 m Inundation 2–4 Area m in Each Flow 4–6 mDepth Range, 6–8sq. m km Case 0.323 0.01–2 m 8.492 2–4 m 4–613.272 m 6–8 m 6.455 8–10 m 0.271 11.562 11.962 2.272 01 0.323 8.492 13.272 6.455 0.018 0.325 13.928 8.420 0.753 02 0.271 11.562 11.962 2.272 0.000 0.358 12.456 7.497 0.351 03 0.325 13.928 8.420 0.753 0.000 7.826 9.242 040.350 0.358 12.456 7.497 0.351 8.6700.000 13.379 9.454 050.290 0.350 7.826 9.242 8.670 0.4240.045 0.240 12.907 10.221 1.142 06 0.290 13.379 9.454 0.424 0.000 0.479 10.085 2.125 07 0.240 12.907 10.221 1.142 0.0000.000 9.646 11.985 080.316 0.479 10.085 2.125 0.000 2.9440.000 090.608 0.316 9.646 11.985 2.944 0.0190.000 11.151 3.927 100.218 0.608 11.151 3.927 0.019 2.7360.000 9.727 10.910 110.336 0.218 9.727 10.910 2.736 0.3510.000 11.491 8.033 12 0.336 11.491 8.033 0.351 0.000 0.353 14.705 7.198 0.375 130.433 0.353 14.705 7.198 0.375 0.0000.000 6.855 0.691 140.251 0.433 6.855 0.691 0.000 0.4240.000 11.925 10.051 15 0.251 11.925 10.051 0.424 0.000 0.419 7.380 0.605 0.000 16 0.419 7.380 0.605 0.000 0.000 9.817 2.125 170.431 0.431 9.817 2.125 0.000 0.0000.000 9.683 1.743 180.350 0.350 9.683 1.743 0.000 0.0000.000 0.218 4.024 0.229 0.000 19 0.218 4.024 0.229 0.000 0.000 0.309 13.715 7.596 20 0.309 13.715 7.596 0.203 0.2030.000 6.776 0.691 210.395 0.395 6.776 0.691 0.000 0.0000.000 0.373 7.080 0.547 22 0.373 7.080 0.547 0.000 0.0000.000 4.083 0.235 230.221 0.221 4.083 0.235 0.000 0.0000.000 0.000 0.000 240.000 0.000 0.000 0.000 0.000 0.0000.000
8–10 m 0.018 0.000 0.000 0.000 0.045 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
Figure Figure 9. 9. Inundation Inundation area area of of each each scenario scenario and and case case for for tsunami tsunamiprotection protectionin inSendai SendaiCity. City.
The simulatedcases casesofof2-layer 2-layer schemes resulted in several sets of that dataare that are explained The 77 simulated schemes resulted in several sets of data explained below. below. The inundation area of the 2-layer schemes varies from 12.0–25.0 sq. km with a range The inundation area of the 2-layer schemes varies from 12.0–25.0 sq. km with a range of of approximately inundation area of this scenario is approximately 20.7420.74 sq. km. approximately12.5 12.5sq. sq.km. km.The Theaverage average inundation area of this scenario is approximately sq. km. The maximum value is for case 09, while case 08 has the minimum value. The 3-layer schemes to
Geosciences 2018, 8, 173
Geosciences 2018, 8, x FOR PEER REVIEW
18 of 25
18 of 25
The maximum value is for case 09, while case 08 has the minimum value. The 3-layer schemes to protectagainst againsttsunamis tsunamisinclude includecombining combiningthe theproposed proposedstructures structuresinto into66cases. cases. To Tosummarize summarizethis this protect scheme, the the inundation inundation area area varies varies from from 7.5–23.0 7.5–23.0 sq. sq. km km with with aa mean mean of of approximately approximately 14.3 14.3 sq. sq. km. km. scheme, The 4-component structural combination schemes are divided into 4 cases for their evaluations as The 4-component structural combination schemes are divided into 4 cases for their evaluations as tsunami protection. To summarize this scheme, the inundation area varies from 5.0–22.0 sq.km with tsunami protection. To summarize this scheme, the inundation area varies from 5.0–22.0 sq.km with a a mean approximately 10.54 km. mean of of approximately 10.54 sq.sq. km. The two conditions of the 5-component structural combinations combinations are are categorized categorized as as with with and and The two conditions of the 5-component structural without overflow of the elevated road and represent the full development of Sendai City’s protection without overflow of the elevated road and represent the full development of Sendai City’s protection against future future large large tsunamis. tsunamis. The The overflow overflow case case shows shows aa 24.02 24.02 sq. sq. km km reduction reduction in in the the tsunami tsunami against inundation area. The tsunami does not overflow the elevated road; however, the tsunami wave inundation area. The tsunami does not overflow the elevated road; however, the tsunami wave overflowsthe theelevated elevatedroad roadby byan anaverage averagedepth depthof ofapproximately approximately0.30–1.0 0.30–1.0m. m.The Theoutput outputshows showsthe the overflows inundation resulting from the no-overflow case, which reduces the inundation area by 28.56 sq. km inundation resulting from the no-overflow case, which reduces the inundation area by 28.56 sq. km relativeto tothat thatof ofthe thenonstructural nonstructuralscheme. scheme. relative 3.3.Performances PerformancesofofSeveral SeveralCombinations CombinationsofofMultilayered MultilayeredStructures Structures 3.3. Other results results show show aasignificant significantdifference differencein inthe theinundation inundationof ofthe theprotected protectedareas areasfor forseveral several Other combinations of structures. The reduced inundation areas were calculated by computing the combinations of structures. The reduced inundation areas were calculated by computing the difference difference between case 01 and other cases. The performance of each case shown in Figure 10 are between case 01 and other cases. The performance of each case shown in Figure 10 are based on the based on the reduction of the inundation area compared with the nonstructural (case 01). Case reduction of the inundation area compared with the nonstructural case (case 01).case Case 04 (with the 04 (with the elevated road) is the highest performing single-structure scheme, while the highest elevated road) is the highest performing single-structure scheme, while the highest performing 2-layer performing 2-layer scheme is caseand 08 elevated (existingroad). seawall road). For 3-layer scenarios, scheme is case 08 (existing seawall Forand theelevated 3-layer scenarios, thethe highest performing the highest performing case is case 14, which combined theseawall, existingand seawall, a newroad. seawall, and case is case 14, which combined the existing seawall, a new an elevated Case 19 an is elevated road. Case 19 is the combination of the existing seawall, a new seawall, a greenbelt, and an the combination of the existing seawall, a new seawall, a greenbelt, and an elevated road and has the elevated road andofhas best performance of the 4-layer scenarios. best performance thethe 4-layer scenarios.
Figure of of each casecase for the against tsunamis in Sendai City. Figure10. 10.Evaluating Evaluatingthe theperformances performances each for protection the protection against tsunamis in Sendai City.
3.4. Performances of the Existing and Reconstructed Structural Measures against the 2011 Tsunami 3.4. Performances of the Existing and Reconstructed Structural Measures against the 2011 Tsunami The existing structural system in Sendai City includes 2 structures, the seawall and the highway, as in case 09 existing of this study. However, after the 2011 Great Tsunami occurred, the Sendai Cityand identified a plan The structural system in Sendai City includes 2 structures, seawall the highway, to large tsunamis in theafter future, whichGreat is presented Figure 1 Sendai as case City 23 inidentified this study.a asprotect in case against 09 of this study. However, the 2011 Tsunamiinoccurred, To show the performance of thetsunamis Sendai City severalwhich cases were compared, case23 23 in result plan to protect against large in plan, the future, is presented in showing Figure 1 that as case this study. To show the performance of the Sendai City plan, several cases were compared, showing that case 23 result in a reduction of the tsunami inundation area over the existing structural conditions
Geosciences 2018, 8, 173
19 of 25
in a reduction of the tsunami inundation area over the existing structural conditions (case 09), which is Geosciences 2018, 8, x FOR PEER REVIEW 19 of 25 equivalent to approximately 20.35 sq. km. In the Sendai City plan, hills in the park in the greenbelt region are used as safety areas when facing tsunamis; thus, the result of case 23 can be used to recommend the (case 09), which is equivalent to approximately 20.35 sq. km. In the Sendai City plan, hills in the park hill height. The hill’sregion heightare is recommended to bewhen higherfacing than tsunamis; 6.0 m in the northern area the23greenbelt, in the greenbelt used as safety areas thus, the result ofof case can whilebeinused the southern area, the height is recommended to be higher than 8.0 m. to recommend the hill height. The hill’s height is recommended to be higher than 6.0 m in the northern area of the greenbelt, while in the southern area, the height is recommended to be higher
3.5. Construction than 8.0 m. Cost Perspectives of the Multilayered Countermeasures
The construction costs for the 3 structures (the seawall, greenbelt, and road) are the costs for reconstructing and introducing new components to protect against future tsunami disasters in Sendai construction costs the 3onstructures (the seawall, greenbelt, and are City the costs for City. The The construction costs arefor based the established materials needed forroad) Sendai to construct reconstructing introducing new components to protect disasters in cost Sendai the elevated road,and which assumes either a unit length oragainst a unit future area. tsunami The construction of the City. The construction costs are based on the established materials needed for Sendai City to construct elevated road is a base cost required to estimate the construction costs of the other structures, such the elevated road, which assumes either a unit length or a unit area. The construction cost of the as the highway and seawall. The elevated road cost is approximately 1.92 billion yen for a length of elevated road is a base cost required to estimate the construction costs of the other structures, such 10.2 km [28]. The construction material is assumed to have 2 components: the cost of building up as the highway and seawall. The elevated road cost is approximately 1.92 billion yen for a length of the earth and road the manpower which to is have assumed to account 75% of 10.2 km [28]. Theand construction materialcost, is assumed 2 components: thefor costapproximately of building up the the material cost. The assumptions used to estimate the construction cost above is used for the earth and road and the manpower cost, which is assumed to account for approximately 75% of theother structures, theassumptions seawall andused the highway. 3 shows the cost of the structures in which materialsuch cost.asThe to estimateTable the construction cost above is 3 used for the other the unit cost is such based material cost of the elevated road.the The cost structure is used to structures, as on thethe seawall and the highway. Table 3 shows cost of of theeach 3 structures in which the unit is based oncost the for material cost of the elevated road.11 The cost ofthe each structure is cost usedin toeach calculate thecost construction each protection case. Figure shows construction calculate the construction cost for each protection case. Figure 11 shows the construction cost in each case within the scenario groupings. case the scenario groupings. The within economic value of the Tohoku region can be obtained from the Research and Statistics The economic value of the Ministry Tohoku region can be Trade obtained the Research Statistics Department Minister’s Secretariat of Economy, andfrom Industry (METI),and Japan. The total Department Minister’s Secretariat Ministry of Economy, Trade and Industry (METI), Japan. The total economic cost is approximately 57,267 billion yen for the total area of approximately 66,890 sq. km for economic cost is approximately 57,267 billion yen for the total area of approximately 66,890 sq. km the Tohoku area [40]. Given that, the economic unit area cost is approximately 0.856 billion yen per for the Tohoku area [40]. Given that, the economic unit area cost is approximately 0.856 billion yen sq. km, value can be used benefit protection. Figure 12Figure presents the benefit/cost perasq. km,that a value that canto beestimate used to the estimate theofbenefit of protection. 12 presents the analysis (BCA) for each method of protecting Sendai City. The construction cost is estimated using benefit/cost analysis (BCA) for each method of protecting Sendai City. The construction cost is the unit estimated cost and the quantity of units. The benefit can be estimated by the reduction of the inundation using the unit cost and the quantity of units. The benefit can be estimated by the reduction area of (protected area) plus economic unit area The BCA can be estimated by can the be cost of the inundation areathe (protected area) plus thecost. economic unitindex area cost. The BCA index estimatedarea by the cost ofby thethe protected area divided by trend the construction Thefor trend of theprotection BCA the protected divided construction cost. The of the BCAcost. index tsunami index forwhen tsunami is protected increased when number of protectedThus, lines (scenarios) is increased theprotection number of lines the (scenarios) increases. the layers increases. of protection Thus, the layers of protection have an effect on the tsunami behavior that is reflected in the BCA. have an effect on the tsunami behavior that is reflected in the BCA. 3.5. Construction Cost Perspectives of the Multilayered Countermeasures
Figure 11.11. Construction oftsunami tsunamiprotection protection Sendai City. Figure Constructioncosts costsfor foreach each case case of in in Sendai City.
Geosciences 2018, 8, 173
Geosciences 2018, 8, x FOR PEER REVIEW
20 of 25
20 of 25
Figure 12.12.Benefit/cost analysis(BCA) (BCA)index index each of tsunami protection in Sendai Figure Benefit/cost analysis forfor each casecase of tsunami protection in Sendai City. City. Table Theconstruction construction cost to protect Sendai City City fromfrom a tsunami. Table 3.3The costofofthree threestructures structures to protect Sendai a tsunami.
Structure Unit Cost Quantity Cost (Billion yen) Structure Unit Cost Quantity Cost (Billion yen) Elevated road 0.19 billion yen/km 10.20 km 1.92 Elevated road 0.140.19 billion yen/km 10.20 1.92 New seawall billion yen/km 1.65 km km 1.65 New seawall 1.000.14 millionbillion yen/km 1.65 hectare km 1.65 Greenbelt yen/hectare 486.00 0.49 Greenbelt 1.00 million yen/hectare 486.00 hectare 0.49 3.6. Performance of Multilayered Structures from an Economic Construction Cost Perspective
3.6. Performance of Multilayered Structures an Economic Construction Cost Perspective Several cases used to protect againstfrom tsunamis have provided 2 different results: those of the tsunami areato(damaged area) ontsunamis the demand and those of the protected Severalinundation cases used protect against haveside provided 2 different results:area those of (reduction area from the nonstructural case) on the supply side. Figure 13 shows the relationship the tsunami inundation area (damaged area) on the demand side and those of the protected area between the tsunami inundation area and the construction costs. The relationship trend decreases (reduction area from the nonstructural case) on the supply side. Figure 13 shows the relationship when the damage area increases, leading to a smaller construction cost. The relationship between the between the tsunami inundation area and the construction costs. The relationship trend decreases protected area and the construction cost is shown in Figure 14. The trend increases when the when the damage area leading to a smaller cost. The relationshipbalance between the protected area and theincreases, construction costs increase. The construction optimum point is the recommended protected area and the construction cost is shown in Figure 14. The trend increases when the between the damaged and benefited areas for the optimization of the protections against tsunamisprotected in area and the construction costs increase. The optimum point is the recommended balance between the Sendai City. Based the results mentioned the performance of large numbers components, such City. damaged andon benefited areas for the above, optimization of the protections againstoftsunamis in Sendai as Based in the 5-layer have higherabove, efficiencies than those of other numbers of layers, on average. such on thescenarios, results mentioned the performance of large numbers of components, However, certain cases, such as the 5-layer case 23 andthan the 4-layer case 19, have the same performance as in the 5-layer scenarios, have higher efficiencies those of other numbers of layers, on average. in protecting against tsunamis, and several cases have the same results. The optimum to However, certain cases, such as the 5-layer case 23 and the 4-layer case 19, have the samepoint performance consider is an important key to identifying the best scenario for tsunami prevention. To determine in protecting against tsunamis, and several cases have the same results. The optimum point to consider the best multilayer scenario, finding the equilibrium point between demand and supply in an is an important key to identifying the best scenario for tsunami prevention. To determine the best economic sense is required. The equilibrium point can be determined by using the cost and quantity multilayer scenario, finding the equilibrium point between demand and supply in an economic sense domain, and the point can be presented as the intersection point between the demand and supply is required. The equilibrium point be determined using the curve, cost and quantity domain, and the curves [41–43]. For this study, thecan demand side is the by damage-cost while the supply is the point can be presented the intersection point between the demand and supply curves [41–43]. For this protection-cost curve.as Figure 15 shows the interaction between the damage costs and the protection study, theshowing demand side is the damage-cost curve, while the supply is the protection-cost curve. Figure 15 costs, the optimization of the investment with respect to potential damage. The equilibrium value is interaction approximately 5–6 billion yen forcosts construction Of thecosts, 24 cases considered, 2 cases are of the shows the between the damage and the costs. protection showing the optimization in the equilibrium zone: 4-layerdamage. scenariosThe of case 21 (existing seawall + new seawall + elevated investment with respect to the potential equilibrium value is approximately 5–6 billion yen for road + highway) case2422cases (existing seawall +2greenbelt elevated road + highway). A comparison construction costs. and Of the considered, cases are+in the equilibrium zone: the 4-layer scenarios of both based on the +performance of + the protected area+ in Figure 10and andcase the BCA presentedseawall in of case 21 cases (existing seawall new seawall elevated road highway) 22 (existing +
greenbelt + elevated road + highway). A comparison of both cases based on the performance of the protected area in Figure 10 and the BCA presented in Figure 12 indicates that the optimum case is case 22. Thus, the results of case 22 of the 4-layer scenario reveal that the inclusion of the greenbelt
Geosciences 2018, 8, x FOR PEER REVIEW Geosciences 2018, 8, x FOR PEER REVIEW Geosciences 2018, 8, 173
21 of 25 21 of 25 21 of 25
Figure 12 indicates that the optimum case is case 22. Thus, the results of case 22 of the 4-layer scenario Figure 12 indicates that the optimum case is case 22. Thus, the results of case 22 of the 4-layer scenario reveal that the inclusion of the greenbelt has a better performance in protecting against tsunamis in reveal that the inclusion ofin theprotecting greenbelt against has a better performance in City, protecting against in has a better performance tsunamis in Sendai reflecting thistsunamis structure’s Sendai City, reflecting this structure’s performance in the single-layer scenarios mentioned above. Sendai City, reflecting this structure’s performance the single-layer scenarios mentioned above. performance in the single-layer scenarios mentionedinabove.
Figure Figure 13. 13. Relationship Relationship between between the the tsunami tsunami inundation inundation area area and and construction construction costs. costs. Figure 13. Relationship between the tsunami inundation area and construction costs.
Figure 14. Relationship between the protected area (benefit area) and construction costs. Figure 14. Relationship between the protected area (benefit area) and construction costs. Figure 14. Relationship between the protected area (benefit area) and construction costs.
Geosciences 2018, 8, 173 Geosciences 2018, 8, x FOR PEER REVIEW
22 of 25 22 of 25
Figure 15. Relationship between the protected area (benefit area) and the damaged area, used to identify Figure 15. Relationship between the protected area (benefit area) and the damaged area, used to the optimum point of the multilayer countermeasures for preventing tsunami damage in Sendai City. identify the optimum point of the multilayer countermeasures for preventing tsunami damage in Sendai City.
4. Conclusions
4. Conclusions The present study investigated the performance of structural countermeasures of Sendai City
in the face of thestudy tsunami on 11 March and assessed the performance of multilayer structural The present investigated the 2011, performance of structural countermeasures of Sendai City in countermeasures for future tsunami mitigation. The tsunami source model was calculated using the face of the tsunami on 11 March 2011, and assessed the performance of multilayer structural the 10 subfaults based on the Tohoku University fault parameters 1.2).was Thecalculated 10-subfault model countermeasures for future tsunami mitigation. The tsunami source (V. model using the was evaluated with the recorded water levels collected by the TTJS group, using the mean ratio 10 subfaults based on the Tohoku University fault parameters (V. 1.2). The 10-subfault model was bias correction to calibrate initial water level. by The resulted several evaluated with the recordedthe water levels collected theTUNAMI-N2 TTJS group, model using the mean in ratio bias inundationtodatasets 24 cases based 5 scenarios of structural combination: structure, correction calibratefrom the initial water level.on The TUNAMI-N2 model resulted in severalnoinundation a seawall,from a greenbelt, elevated or a highway. datasets 24 casesanbased on 5road scenarios of structural combination: no structure, a seawall, a The results of the road tsunami greenbelt, an elevated or amodeling highway. demonstrate significant differences in tsunami inundations whenThe evaluating combinations structures. The elevateddifferences road (case in 04)tsunami providesinundations the highest results ofseveral the tsunami modelingofdemonstrate significant performance of the single schemes, whereas the highest performance of the 2-layer schemes is the when evaluating several combinations of structures. The elevated road (case 04) provides the highest combination of and an elevated road (case 09). For the scenario, the highest performance of an theexisting single seawall schemes, whereas the highest performance of3-layer the 2-layer schemes is the performance is case 14, which is the grouping of an existing seawall, a new seawall, and an elevated road. combination of an existing seawall and an elevated road (case 09). For the 3-layer scenario, the highest The combination of an existing seawall, a new seawall, a greenbelt and an elevated road (case 19) is the performance is case 14, which is the grouping of an existing seawall, a new seawall, and an elevated highest scenario. The 5-layer scenarioa reduces inundation area road. Theperforming combination4-layer of an existing seawall, a new seawall, greenbeltthe andtsunami an elevated road (case 19)by is 20 sq. km via existing structural conditions. Finally, a demand/supply curve was applied to determine the highest performing 4-layer scenario. The 5-layer scenario reduces the tsunami inundation area by thesq. optimum combination of structures that Finally, can protect Sendai City against The to best solution 20 km via existing structural conditions. a demand/supply curvetsunamis. was applied determine indicates that the combination of the existing seawall, a greenbelt, an elevated road and a highway the optimum combination of structures that can protect Sendai City against tsunamis. The best (case 22) indicates in the 4-layer would provide the city from tsunamis similar solution that scenario the combination of the optimal existing protection seawall, a to greenbelt, an elevated road and to a the 2011 Great Tsunami. highway (case 22) in the 4-layer scenario would provide optimal protection to the city from tsunamis The results this study are limited by the inundation values estimated from the similar to the 2011 of Great Tsunami. TUNAMI-N2 model. studyare thelimited tsunami, are severalvalues required variables, such the traveling The results of thisTostudy bythere the inundation estimated from theasTUNAMI-N2 time of To thestudy tsunami its duration peak, and thevariables, ranges ofsuch the inundation depths. For model. the wave, tsunami, there are and several required as the traveling time ofcost the estimations, the construction cost was estimated by the unit cost to demonstrate the performance of the tsunami wave, its duration and peak, and the ranges of the inundation depths. For cost estimations, prevention structures, butestimated these structures several other costs in real applications. In future work, the construction cost was by the have unit cost to demonstrate the performance of the prevention the arrival time, duration and peak times can be used to identify the performance of the prevention structures, but these structures have several other costs in real applications. In future work, the arrival structures in the face of tsunami Finally, thethe results of this study can be used tostructures evaluate time, duration and peak times candisasters. be used to identify performance of the prevention tsunami protection anddisasters. mitigationFinally, schemes a countermeasure methodology. in the face of tsunami thewith results of this study can be used to evaluate tsunami protection and mitigation schemes with a countermeasure methodology.
Geosciences 2018, 8, 173
23 of 25
Author Contributions: K.P., A.S. and F.I. contributed to the implementation of the research, analysis of the results and writing of the manuscript. Acknowledgments: The observational data used to calibrate and verify the tsunami models of the 2011 Great Tsunami in Japan were provided by TTJS. This research was funded by the Willis Research Network (WRN) under the Pan-Asian/Oceanian tsunami risk modeling project through the International Research Institute of Disaster Science (IRIDeS) at Tohoku University. Conflicts of Interest: The authors declare no conflict of interest.
References 1. 2. 3. 4.
5.
6.
7. 8. 9.
10. 11. 12. 13.
14. 15. 16.
17.
18.
Geospatial Information Authority of Japan (GSI). Available online: http://www.gsi.go.jp/ (accessed on 14 March 2011). Koshimura, S.; Hayashi, S.; Gokon, H. The impact of the 2011 Tohoku earthquake tsunami disaster and implications to the reconstruction. Soils Found. 2014, 54, 560–572. [CrossRef] Mori, N.; Takahashi, T. The 2011 Tohoku Earthquake Tsunami Joint Survey Group. Nationwide post event survey and analysis of the 2011 Tohoku earthquake tsunami. Coast. Eng. J. 2012, 54, 125001. [CrossRef] Lacharote, P.; Leelawat, N.; Suppasri, A.; Thamarax, P.; Imamura, F. Estimation of fatality ratios and investigation of influential factors in the 2011 Great East Japan Tsunami. Int. J. Disaster Risk Reduct. 2017, 29, 37–54. [CrossRef] Suppasri, A.; Koshimura, S.; Imai, K.; Mas, E.; Gokon, H.; Muhari, A.; Imamura, F. Damage characteristic and field survey of the 2011 Great East Japan tsunami in Miyagi prefecture. Coast. Eng. J. 2012, 54, 1250005. [CrossRef] National Police Agency. Report of the Damage Caused by the 2011 Tohoku Earthquake and Tsunami. Available online: http://www.npa.go.jp/archive/keibi/biki/higaijokyo.pdf (accessed on 9 December 2016). (In Japanese) Ministry of Finance Japan, 2011. Japan’s Fiscal Condition. Available online: www.mof.go.jp/english/ budget/budget/ (accessed on 20 December 2011). Ministry of Land, Infrastructure, Transport and Tourism (MLIT). 2011. Available online: http://www.mlit. go.jp/common/000138154.pdf (accessed on 3 March 2014). Liu, P.L.F.; Cho, Y.S.; Yoon, S.B.; Seo, S.N. Numerical Simulation of the 1960 Chilean Tsunami Propagation and Inundation as Hilo, Hawaii; Recent Development in Tsunami Research; Kluwer Academic Publishers: Dordrecht, The Netherlands, 1994; pp. 99–115. Liu, P.L.F.; Woo, S.B.; Cho, Y.S. Computer Programe for Tsunami Propagation and Inundation; Technical Report; Cornell University: Ithaca, NY, USA, 1998. Titov, V.; Synolakis, C.E. Numerical Modeling of Tidal Wave Run-up. J. Waterw. Port Coast. Ocean Eng. 1998, 124, 157–171. [CrossRef] Imamura, F. Review of tsunami simulation with a finite difference method. In Long-Wave Runup Models; World Scientific Pub Co Inc.: Singapore, 1995; pp. 25–42. Jais wal, R.K.; Singh, A.P.; Rastogi, B.K. Simulation of the Arabian Sea tsunami propagation generated due to 1945 Makran earthquake and its effect on the western parts of Gujarat, India. Nat. Hazards 2008, 48, 245–248. [CrossRef] Usha, T.; Ramana Murthy, M.V.; Murthy, N.T.; Murty, T.S. Vulnerability Assessment of Car Nicobar to Tsunami Hazard using Numerical Model. Sci. Tsunami Hazards 2009, 28, 15–34. Chenthamil Selvan, S.; Kankara, R.S. Tsunami model simulation for 26 December 2004 and its effect on Koodankulam region of Tamil Nadu Coast. Int. J. Ocean Clim. Syst. 2016, 7, 62–69. [CrossRef] Koshimura, S.; Oie, T.; Yanagisa, H.; Imamura, F. Developing fragility curves for tsunami damage estimation using numerical and post-tsunami data from Banda Aceh, Indonesia. Coast. Eng. J. 2009, 51, 234–273. [CrossRef] Yamashita, K.; Sugawara, D.; Takahashi, T.; Imamura, F.; Saito, Y.; Imato, Y.; Kai, T.; Uehara, H. Numerical simulation of large-scale sediment transport caused by the 2011 Tohoku Earthquake Tsunami in Hirota Bay, Southern Sanriku Coast. Coast. Eng. J. 2016, 58, 1640015. [CrossRef] Sugawara, D.; Goto, K. Numerical modeling of the 2011 Tohoku-oki tsunami in the offshore and onshore of Sendai Plain, Japan. Sediment. Geol. 2012, 282, 110–123. [CrossRef]
Geosciences 2018, 8, 173
19.
20. 21. 22.
23.
24. 25.
26.
27. 28.
29. 30.
31. 32. 33.
34. 35. 36. 37.
38. 39.
24 of 25
Oishi, Y.; Imamura, F.; Sugawara, D. Near-field tsunami inundation forecast using the parallel TUNAMI-N2 model: Application to the 2011 Tohoku-Oki earthquake combined with source inversions. Geophys. Res. Lett. 2015, 42, 1083–1091. [CrossRef] Sato, S. Characteristics of the 2011 Tohoku Tsunami and introduction of two level tsunamis for tsunami disaster mitigation. Proc. Jpn. Acad. Ser. B Phys. Biol. Sci. 2015, 91, 262–272. [CrossRef] [PubMed] Vana Tsimopoulou, J.K.V.; Sebastiaan, M.K.; Jonkman, N. Cost-Efficient Design of Multilayer Safety Systems against Large-Scale Coastal Disasters, 1st ed.; Elsevier Inc.: Oxford, UK, 2015. Tsimopoulou, V.; Jonkman, S.N.; Kolen, B.; Maaskant, B.; Mori, N.; Yasuda, T. A multi-layered safety perspective on the tsunami disaster in Tohoku, Japan. In Proceedings of the Flood Risk 2012 Conference, Rotterdam, The Netherlands, 19–23 November 2012; pp. 1–10. Khew, Y.T.J.; Jarzebski, M.P.; Dyah, F.; San Carlos, R.; Gu, J.; Esteban, M.; Aránguiz, R.; Akiyama, T. Assessment of social perception on the contribution of hard-infrastructure for tsunami mitigation to coastal community resilience after the 2010 tsunami: Greater Concepcion area, Chile. Int. J. Disaster Risk Reduct. 2015, 13, 324–333. [CrossRef] Takagi, H.; Bricker, J.D. Assessment of the effectiveness of general breakwaters in reducing tsunami inundation in Ishinimaki. Coast. Eng. J. 2014, 56, 1–21. [CrossRef] Samawasekara, R.S.M.; Sasaki, J.; Esteban, M. Assessment of the co-benefits of structures in coastal areas for tsunami mitigation and improving community resilience in Sri Lanka. Int. J. Disaster Risk Reduct. 2015, 13, 324–333. [CrossRef] Suppasri, A.; Latcharote, P.; Bricker, J.D.; Leelawat, N.; Hayashi, A.; Yamashita, K.; Makinoshima, F.; Roeber, V.; Imamura, F. Improvement of tsunami countermeasures based on lessons from the 2011 great east japan earthquake and tsunami -Situation after five years. Coast. Eng. J. 2016, 58, 1640011. [CrossRef] ¯ Strusinska-Correia, ´ A. Tsunami mitigation in Japan after the 2011 Tohoku Tsunami. Int. J. Disaster Risk Reduct. 2017, 22, 397–411. [CrossRef] Sendai City, Sendai City Earthquake Disaster Reconstruction Plan, Digest Version. Available online: https://www.city.sendai.jp/shinsaifukko/shise/daishinsai/fukko/kanren/kekaku/documents/ english14511_cover-p2.pdf (accessed on 5 February 2018). Okada, Y. Surface deformation due to shear and tensile faults in a half-space. Bull. Seismol. Soc. Am. 1985, 75, 1135–1154. Imamura, F.; Koshimura, S.; Murashima, Y.; Akita, Y.; Shintani, Y. The Tsunami Source Model of the 2011 Tohoku Earthquake (ver.1.2). Available online: http://www.tsunami.civil.tohoku.ac.jp/hokusai3/J/events/ tohoku_2011/model/dcrc_ver1.2.pdf (accessed on 25 April 2012). The 2011 Tohoku Earthquake Tsunami Joint Survey (TTJS) Group. Nationwide field survey of the 2011 off the Pacific coast of Tohoku earthquake tsunami. J. Jpn. Soc. Civ. Eng. 2011, 67, 63–66. Aida, I. Reliability of a tsunami source model derived from fault parameters. J. Phys. Earth 1978, 26, 57–73. [CrossRef] Suppasri, A.; Koshimura, S.; Imamura, F. Developing tsunami fragility curves based on the satellite remote sensing and the numerical modeling of the 2004 Indian Ocean tsunami in Thailand. Nat. Hazards Earth Syst. Sci. 2011, 11, 173–189. [CrossRef] Ines, A.V.M.; Hansen, J.W. Bias correction of daily GCM rainfall for crop simulation studies. J. Agric. For. Meteorol. 2006, 138, 44–53. [CrossRef] Acharya, N.; Chattopadhyay, S.; Mohanty, U.C.; Dasha, S.K.; Sahooc, L.N. On the bias correction of general circulation model output for Indian Summer Monsoon. J. Meteorol. Appl. 2013, 20, 349–356. [CrossRef] Asadullah, A.; Mcintyre, N.E.I.L.; Kigobe, M.A.X. Evaluation of five satellite products for estimation of rainfall over Uganda. Hydrol. Sci. J. 2010, 53, 1137–1150. [CrossRef] Pakoksung, K. Runoff Analysis Using Satellite Data for Regional Flood Assessment; Spatial and Time Series Bias Correction of Satellite Data. Ph.D. Thesis, Kochi University of Technology, Kami, Japan, 23 September 2016; p. 132. Hayashi, S.; Koshimura, S. The 2011 Tohoku tsunami flow velocity estimation by the aerial video analysis and numerical modeling. J. Disaster Res. 2013, 8, 561–572. [CrossRef] Geospatial Information Authority of Japan (GSI). On Estimation of Flooded Area by the 2011 Tsunami. Available online: www.gsi.go.jp/common/000059939.pdf (accessed on 5 May 2013). (In Japanese)
Geosciences 2018, 8, 173
40.
41. 42.
43.
25 of 25
Research and Statistics Department Minister’s Secretariat, Ministry of Economic, Trade and Industry (METI), Japan. Available online: http://www.meti.go.jp/english/statistics/tyo/tiikiio/index.html (accessed on 5 February 2018). Leng, L. Health Economic Evaluation: A Primer for Healthcare Professionals. Prim. Health Care 2016, 6, 223. [CrossRef] Locke, J. Some Considerations of the Consequences of Lowering the Interest and Raising the Value of Money. Available online: http://la.utexas.edu/users/hcleaver/368/368LockeSomeConsiderationsAlltable.pdf (accessed on 5 February 2018). Denham-Steuart, J. Inquiry into the Principles of Political Economy. Published in The Strand. 1767. Available online: https://www.marxists.org/reference/subject/economics/steuart/index.htm (accessed on 27 October 2014). © 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
