Tsunami Fragility A New Measure to Identify Tsunami Damage

Tsunami Paper: Tsunami A New Measure to Identify Tsunami Damage Shunichi Koshimura, Yuichi Namegaya, and Hideaki Yanagisawa Graduate School of Engineering, Tohoku University Aoba --11-11, Aramaki, Aoba-ku, Sendai 9-579, Japan E-mail: koshimura@tsunami.civil.tohoku.ac.jp Geological Survey of Japan, National Institute of Advanced Industrial Science and Technology C7, 1-1-1 Higashi, Tsukuba 35-57, Japan E-mail: yuichi.namegaya@aist.go.jp Tokyo Electric Power Services Co., Ltd. 3-3, Higashiueno 3-Chome, Taito-ku, Tokyo 11-15, Japan E-mail: h-yanagi@tepsco.co.jp [Received June 9, 9; accepted November 5, 9] Abstract. Tsunami fragility (fragility curve, or fragility function) is a new measure, we propose, for estimating structural damage and fatalities due to tsunami attack, by integrating satellite remote sensing, field survey, numerical modeling, and historical data analysis with geographic information system (GIS). Tsunami fragility is expressed as the structural damage probability or fatality ratio related to hydrodynamic features of tsunami inundation flow, such as inundation depth, current velocity and hydrodynamic force. It expands the capability of estimating potential tsunami damage in a quantitative manner. Keywords: fragility curve, tsunami damage estimation, remote sensing, numerical modeling, historical tsunamis 1. Introduction In tsunami damage estimation efforts, several empirical relationships between tsunami hazard and vulnerability have been used. Shuto (1993) proposed the tsunami intensity scale to discuss the structural damage based on the empirical data from historical tsunamis in Japan, in terms of the damage and local tsunami height [1], and this has been widely used in tsunami disaster assessment by the Japanese government as a measure of tsunami damage. When the local tsunami inundation depth exceeds m, for example, Shuto s tsunami intensity scale suggests, complete destruction of wooden houses as shown in Fig. 1. Izuka and Matsutomi () suggested a structural destruction threshold related to the hydrodynamic features of tsunami inundation flow throughout the field surveys and laboratory experiments []. And some engineering studies have proposed tsunami design forces on structures based on the laboratory experiments [3, ], but have not suggested the procedures for estimating structural damage. Since the Sumatra-Andaman earthquake tsunami disaster, numerous efforts have been made to identify the tsunami damage mechanisms by widely deployed posttsunami survey teams reporting tsunami height, inundation zone extent and damage [5 9]. These efforts have led to new understandings of local aspects of tsunami inundation flow and damage mechanisms. However, their findings based on the inspection of local aspects of tsunami damage make it difficult to identify the vulnerability in a quantitative manner. The nature of vulnerability is associated with multitude of uncertain sources, such as hydrodynamic features of tsunami inundation flow, structural characteristics, population, land use, and any other site conditions. To view tsunami vulnerability comprehensively requires humongous amounts of damage data, whereas post-tsunami survey rarely provides sufficient data because of limited survey time and human resources. As Shuto (1993) concluded, degree of damage may change with these uncertainties and a statistical approach to these uncertainties should be taken. Herein, we propose Tsunami fragility (fragility curve or fragility function) as a new measure for estimating tsunami damage. Tsunami fragility is defined as the structural damage probability or fatality ratio with particular regard to the hydrodynamic features of tsunami inundation flow, such as inundation depth, current velocity and hydrodynamic force. In principle, the development of tsunami fragility requires that tsunami hazard information and damage data should be used synergistically. We thus incorporate several approaches to constructing the tsunami fragility. In order to obtain tsunami hazard information such as inundation depth and current velocity, we performed a numerical modeling of tsunami inundation with some model validations, especially focusing on the Sumatra- Andaman earthquake tsunami. In terms of the damage data including structural damage and fatalities, we use the recent advances of remote sensing technologies expanding capabilities of detecting spatial extent of tsunami af- Journal of Disaster Research Vol. No., 9 79

Koshimura, S., Namegaya, Y., and Yanagisawa, H. Tsunami intensity 1 3 5 Tsunami height (m) 1 1 3 Damage : wooden house partly damaged completely destroyed Damage : masonry house withstand no data completely destroyed Damage : reinforced concrete building Damage : fishing boats withstand damage 5 damage no data completely destroyed completely damaged Fig. 1. Tsunami intensity scale and damage [1], modified from the original figure. fected areas and damage on structures. In addition, some field data and historical documents are used to obtain the tsunami fragility from historical events. Throughout the data integration and statistical analysis, we propose a concept and framework developing tsunami fragility, and apply it in several approaches according to available data types from recent and historical events.. Developing Tsunami Methods curves (or fragility functions) have conventionally been developed in performing seismic risk analysis of structural systems to identify structural vulnerability against strong ground motion using damage data associated with historical earthquakes and spatial distribution of observed or simulated seismic responses [1]. And they have been implemented in estimating structural damage against potential seismic risks in which various uncertain sources such as seismic hazard, structural charachteristics, soil-structure interaction are involved [11, 1]. In earthquake engineering studies, the fragility curves are defined by the following Eq. (1) or () ; [ ] lnx λ (x) = Φ = x ξ [ ] x µ (x) = Φ σ = x ( 1 exp πξt 1 πσ exp ( ) (lnt λ ) ξ dt (1) ) (t µ) σ dt. () where (x) is the damage probability of structures, as a function of x or lnx such as the maximum ground acceleration, velocity, and seismic intensity. Here, (x) is expressed by the standard lognormal or normal cumulative distribution function Φ[(lnx λ )/ξ ] or Φ[(x µ)/σ], with two statistical parameters (λ,ξ ) or (µ,σ), asthe mean of lnx (or x), and the standard deviation respectively. To develop tsunami fragility, we take a statistical approach synergistically using remote sensing, numerical model results, field surveys and historical documents, in five steps. 1 Damage data acquisition : obtaining damage data from satellite images, field surveys or historical documents, e.g. numbers of destroyed or survived structures with its spatial information. Tsunami hazard estimation : estimating the hydrodynamic features of tsunami by numerical modeling, field surveys and from historical documents. 3 Data assimilation between the damage data and tsunami hazard information : correlating the damage data and the hydrodynamic features of tsunami through the GIS analysis. Calculating damage probability : determining the damage probabilities by counting the number of damaged or survived structures, for each range of hydrodynamic features above. 5 Regression analysis : developing the fragility curves by regression analysis of discrete sets of damage probability and hydrodynamic features of tsunami. In the sections that follow, we apply the above procedure in several approaches to construct tsunami fragility, according to the data types such as satellite data, numerical models, field surveys and historical documents. 3. Tsunami Determined from Satellite Remote Sensing and Numerical Modeling Taking an advantage in satellite remote sensing, we identify the spatial distribution of structural damage by tsunami. The highest spatial resolution of commercial optical satellite imaging is up to -7 centimeters (Quick- Bird owned by DigitalGlobe) or 1 meter (IKONOS operated by GeoEye). Fig. shows the result of visual interpretation [13] on structural damage by using a set of pre and post-tsunami (the tsunami) satellite images (IKONOS) from Banda Aceh, Indonesia. Through inspecting a set of pre and post-tsunami satellite images visually or manually, presence of building roofs can be interpreted. The advantage of using high-resolution optical satellite images for damage interpretation is the capability of understanding structural damage visually and enables us to comprehend its spatial extent in regional scale where post-tsunami survey hardly get through because of Journal of Disaster Research Vol. No., 9

Tsunami Fig.. (a) Spatial distribution of structural damage interpreted from the pre and post-tsunami satellite images (IKONOS) [13]. Black dots indicate the interpreted structures as destroyed, and the gray dots as survived. The arrow points the expanded region shown in the right panels (b) pre-tsunami and (c) post-tsunami satellite images and (d) interpreted damage. (e) The maximum tsunami inundation depth obtained from the numerical model [1, 15]. the limitation of survey time and resources. However, note that no structural types were identified by the interpretation of satellite images. Also, the damage feature which can be identified from satellite images is only structural destruction or major structural failure which reveals change of roof s shape, namely collapsed and major or severe damage. Accordingly, the interpretation Destroyed in Fig. means collapsed or major or severe damage, and Survived is either of moderate, minor, slight and no damage. To obtain the tsunami hazard information such as inundation depth and current velocity, we performed a numerical modeling of the Sumatra-Andaman earthquake tsunami [1, 15], using high-resolution bathymetry and topography data. The model results were validated by the field measurements of post-tsunami survey teams [, 7], using Aida s formula [1] in terms of the reliability of tsunami numerical model (see ref.[15] for detail). Damage interpretation shown in Fig. (a) is combined with the numerical model results, e.g. Fig. (e), to obtain the tsunami damage statistics as shown in Fig. 3. Using GIS, we sampled all of the structures in the tsunami inundation zone and made a table (spread sheet) of structure ID, damage interpretation (Destroyed or Survived) and tsunami hydrodynamic features (inundation depth, current velocity and hydrodynamic force) spatially equiva- Number of 1 1.... 1. 1. 1.3 1.5 1.7 1.9.1.....9 3.1 3.3 3.5 3.7 3.9...3.5.7. 5. 5.1 5.3 5..7. Inundation depth (< x m) Destroyed Survived Fig. 3. Histogram of the numbers of destroyed and survived structures in terms of inundation depth range. Each inundation depth range is determined by exploring a range which includes approximately 1, structures. lent to the position of each structure. After sorting the table by each level of hydrodynamic features, we determined the groups of structures to calculate damage probability so that roughly 1 structures are involved in each group. Then we determined the damage probability in each group according to the range of inundation depth, current velocity and hydrodynamic force obtained by the Journal of Disaster Research Vol. No., 9 1

Koshimura, S., Namegaya, Y., and Yanagisawa, H. 1.. Table 1. Statistical parameters of tsunami fragility curves (Fig. ) for structural damage. R is the coefficient of determination obtained through the least-squares fitting... curve µ σ λ ξ R (a) d max (m).99 1.1 N/A N/A.99 (b) v max (m/s) N/A N/A...97 (c) F (kn/m) N/A N/A 1.7.75.99. 1. (a) 1. (b) 1.. d max Fig.. A discrete set of damage probabilities and the median values of inundation depths that were compiled from sample data......... Least-squares fit d max (m) v max (m/s) 5 1. (c). d max (m) 3.. 1.. -3 - -1 1 3 F -1 Fig. 5. An example of the plot on normal probability paper. numerical model. And as a result of counting the number of destroyed and survived structures within each inundation depth range (group), we obtain a relationship between the damage probability and inundation depth, as a discrete set of structural damage probabilities and tsunami inundation depths shown in Fig.. Then, we explore this relationship with the form of fragility curve by performing the regression analysis. Taking an analogy of earthquake engineering studies [1, 11, 17], we assume that the cumulative probability of damage occurrence is given as either Eq. (1) or () with two statistical parameters, (λ,ξ )or(µ,σ). Here, the statistical parameters λ (or µ) andξ (or σ) are obtained by plotting x (or lnx) and the inverse of Φ (Φ 1 ) on normal or lognormal probability paper, and conducting the least-squares fitting of this plot, as shown in Fig. 5. Hence, these parameters are obtained by taking the intercept (= λ or µ) and the angular coefficient (= ξ or σ) in Eq. (1) or (). Through regression analysis, the parameters are determined as shown in Table 1, to obtain the best fit of fragility curve with respect to the maximum inundation depth (measured above the local ground level) d max (m), the maximum current velocity v max (m/s) and the maxi- 1 F (kn/m) 3 Fig.. Tsunami fragility curves for structural destruction, in terms of (a) the maximum inundation depth, (b) the maximum current velocity and (c) the maximum hydrodynamic force obtained from the numerical model. The solid lines are the best-fitted curves of the plot ( : the distribution of damage probabilities) with the parameters in Table 1. mum hydrodynamic force F acting on a structure per unit width (kn/m). Here F is defined as the maximum drag force per unit width of structures; F = 1 C Dρ max{v d} 1 3....... (3) where C D is the drag coefficient (C D = 1. for simplicity), ρ water density (= 1, kg/m 3 ), v current velocity (m/s) and d inundation depth (m), and both of v and d are obtained at each time step of the tsunami inundation modeling. Note that the tsunami fragility with respect to the inundation depth is given by the standardized normal distribution function with µ andσ, while those to the current velocity and hydrodynamic force are by the standardized lognormal distribution functions with λ and ξ. The selection of which curve is applied should be made by checking its fit to the datasets. curves shown in Fig. indicate the damage probabilities of structural destruction equivalent to the hydrodynamic features of tsunami inundation flow. Houses in Banda Aceh, for example, were especially vulnera- Journal of Disaster Research Vol. No., 9

Tsunami ble when the local inundation depth exceeded or 3 m, the current velocity exceeded.5 m/s or hydrodynamic load exceeded 5 kn/m. Note that the observed structural damage at a site might consist of both damage by tsunami and strong ground motion. Major structure types in the tsunami-affected area were low-rise wooden house, timber construction, and non-engineered RC construction lightly reinforced, and it was reported that the large number of the wooden houses survived the earthquake with minor damage and non-engineered RC structures were degraded by strong ground motion, then they were destroyed by the tsunami [1]. We supposed that the structural destruction was likely to be induced by the tsunami inundation, but many of the structures were degraded by strong ground motion before the tsunami attack. In this sense, the proposed tsunami fragility may involve the structures with minor damage or degraded seismic performance. Note also that the tsunami damage on structures were caused by both hydrodynamic force/impact and the impact of floating debris, i.e. these facts are reflected on the damage probabilities but not on the numerical model results (the estimated hydrodynamic features). Thus, the present tsunami fragility may indicate overestimation in damage probabilities to the hydrodynamic features of tsunami inundation flow. Further to be mentioned is that tsunami fragility proposed herein is for assessing the number of damaged (destroyed) structures by applying tsunami fragility curves to the number of exposed structures against a given hydrodynamic condition of tsunami. This is not for a prediction whether a structure is destroyed or survives under a given probability of occurrence. The proposed tsunami fragility curves here was based on the regression analysis of the relations between the modeled tsunami hazards and the damage probabilities that were sampled in each group of approximately 1 data. How many data should be included in each group (so called data bin) still needs some discussions in statistical point of view. Different selection of bin size (the range to determine the damage probability) may cause different result of regression. In addition, we assumed that the numerical model results represent the features of tsunami inundation flow in the study area without errors. This is the issue that should be discussed as statistical analysis of tsunami fragility curves considering the proper selection of bin size and uncertainties in the numerical model.. Tsunami for Fatality Estimation Tsunami fragility for fatality is determined by using the post-tsunami data in terms of the number of dead, missing and survivors. Fig. 7 shows the spatial distribution of the ratio of dead, missing and survivors in each desa (village) in Banda Aceh city (as of 1 April 5), normalized by the pre-tsunami desa population [13]. GIS analysis of the fatality information and the numerical model results in Fig. (e) yields a fragility curve for tsunami fatalities as the relationship between the fatality ratio (both dead Death ratio Fig. 7. Proportion of fatalities in Banda Aceh city [13], calculated by using reported number of fatality and pre-tsunami population. 1...... (a) d max (m) 1 Death ratio 1...... (b) h max (m) Fig.. Tsunami fragility curves for fatality in terms of (a) the inundation depth and (b) inundation height. The solid line is the best-fitted curve of the plot ( : the distribution of fatality ratio) with Eq. (). and missing) and the hydrodynamic features of tsunami. Based on tsunami fatality data in Fig. 7, the representative value of local hydrodynamic feature of tsunami inundation or the inundation height is calculated by taking the median value of modeled inundation depths and inundation heights within each desa. Figure shows the tsunami fragility expressed as the fatality ratio with regard to the representative values of the maximum inundation depth d max (measured above the local ground level) and inundation height h max (measured above the pre-tsunami tide level) calculated by taking the median value of the numerical model results within each desa as shown in Fig. 7. The fragility curve is determined by assuming the standardized normal distribution function of Eq. () with the parameters of Table obtained through the least-squares fitting, where x is the median value of the inundation depth or inundation height (m) in each desa, calculated by using of the numerical model results. Note that the fatality ratio distribution is the result of the post-tsunami investigation based on the pre-tsunami registration data [13]. It is highly unknown where the residents were affected by the tsunami inundation flow, because it is easily guessed that the residents who were 1 Journal of Disaster Research Vol. No., 9 3

Koshimura, S., Namegaya, Y., and Yanagisawa, H. Fig. 9. The result of visual interpretation of structural damage in Banda Aceh city and the points of field measurements. See their original paper [19] for all the points and areas investigated. Journal of Disaster Research Vol. No., 9

Tsunami Table. Statistical parameters of tsunami fragility curves for fatality (Fig. ). curve µ σ R (a) d max (m) 3.9 1.15. (b) h max (m) 5.37.7.7 aware of tsunami arrival have evacuated and tried to survive. In other words, the fragility curve of Fig. does not indicate the human s survival possibility according to the local hydrodynamic features of tsunami inundation flow. Also, taking median to obtain the representative values of tsunami inundation depth according to each desa reflects higher variance of the plot compared with that of Fig.. For the above reasons, this fragility function should be interpreted as a macroscopic measure of tsunami impact, i.e. the occurrence of tsunami fatality significantly increase when the local inundation depth exceeds approximately m and the inundation height.5 (m), and almost impossible to survive when the local inundation depth exceeds m. Table 3. The damage probabilities and measured tsunami heights obtained by the visual inspection of satellite images and field measurements. The area and points A to O are equivalent to Fig.. Area h max d max Destroyed Pre-tsunami Damage (m) (m) structures structures probability A 7. 5. 1.9 B 7.1 5. 51 57.9 C 7..7 3 5.71 D 7.1 5. 7 3.7 E..9 57.1 F. 3.7 51.5 G..5 31. H.7.1 1 3. I 1. 7 75.9 J 9.7 7.9 1.9 K.1 3.3 51 5.91 L. 3. 3 5.7 M..5 3 5.7 N.5. 1 3.7 O. 31.19 1. (a) 1. (b).. 5. Tsunami from Satellite Remote Sensing and Field Survey.... Developing tsunami fragility may also be viewed using satellite images and post-tsunami surveys. Namegaya and Tsuji () investigated the structural damage in Banda Aceh city by the tsunami, using the visual inspection of QuickBird pre and post-tsunami satellite images acquired on June 3, and December,, in four areas of Banda Aceh city together with the measurements of tsunami inundation depth and height [19]. Fig. is showing their result of visual interpretation of structural damage in Banda Aceh city and the points of tsunami measurements. The markers and in the figure denote survived and destroyed (washed-away) structures interpreted from the satellite images focusing on the presence of their roofs. Using field survey results presents difficulties in correlating the tsunami heights at all the points where the structural damage was inspected. In this case, the damage probability is calculated by counting the number of survived and destroyed structures within the dashed-line circles of 1 m diameter (see Fig. ) and the tsunami inundation depth or height is represented by the measured value at the center of each solid-line circle. Consequently, the relationships between the damage probabilities and tsunami heights are obtained at 13 to 15 points in Banda Aceh city (see Namegaya and Tsuji () for details). Table 3 is the result obtained at each area equivalent to Fig.. The tsunami fragility curves are determined as Fig. 1, with statistical parameters of Table. Note that two fragility curves in Fig. 1(a), although very similar, show differences in approaches of damage data compilation (including number of samples and the inspected area) and methods to obtain tsunami hazard information... d max (m) (Fig.a).. h max (m) Survey Fig. 1. Tsunami fragility curves for structural destruction, in terms of (a) the maximum inundation depth and (b) the maximum inundation height, by using the visual inspection of satellite images and field measurements. The dashed line in (a) indicates the fragility curve from Fig. (a).. Tsunami from Historical Data In constructing the tsunami fragility from historical events, we incorporated the historical tsunami data on local tsunami damage and height. In Japan, the posttsunami surveys would be conducted by many different organizations and individuals. After the 19 Meiji Sanriku earthquake tsunami, which caused approximately, casualties, the damage survey efforts were conducted by the central government [] and an engineer Soshin Yamana delegated by Iwate prefectural government (his report was published by Yamashita (19) with comments and interpretations) [1]. Seismological researchers and Japan Meteorological Agency also conducted the survey after the 1933 Showa Sanriku earthquake tsunami which caused approximately 3, casualties [, 3]. Hatori (19) compiled the house damage data from the historical documents of the 19 Meiji Sanriku, 1933 Showa Sanriku and the 19 Chile tsunami events as listed in Table 5. He defined the structural damage prob- 1 1 Journal of Disaster Research Vol. No., 9 5

Koshimura, S., Namegaya, Y., and Yanagisawa, H. Table. Statistical parameters for fragility curves (Fig. 1). 1. (a) 1. (b) x µ σ R d max (m) 3.33 1.5. h max (m).7 1.3.53...... Table 5. Historical tsunami data in Japan used for developing tsunami fragility. Data compilation Events Original data [References] [References] Hatori (19) [5] 19 Meiji-Sanriku [1] 1933 Showa-Sanriku [] 19 Chile [3] Shuto (197a, 1993) 19 Meiji-Sanriku [, 1,, 7 3] [1, ].. 1. (c)... 19 Meiji (Hatori) Historical data 5 1 15 Tsunami height (m).. 1.... (d) 1933 Showa (Hatori) Historical data 5 1 15 Tsunami height (m) ability as Eq. () by counting the number of houses in three damage categories ; destroyed/washed-away, moderate and only flooded, in each reported area or settlement with tsunami height ; = a + b/ a + b + c............. () where a, b and c is the number of the houses categorized as destroyed/washed-away, moderate damage and only flooded, respectively. Shuto (197a, 1993) also compiled the documents and reports from the 19 Meiij Sanriku tsunami (Table 5), and determined the damage probability with four damage categories ; = a + b + c/ a + b + c + d........... (5) where a, b, c and d is the number of the houses categorized as washed-away, completely destroyed, moderate damage and only flooded, respectively. To increase the reliability of data, he conducted the numerical modeling and the additional field survey to determine the reliability of the documents. Figure 11 plots historical data (damage probability versus tsunami height and inundation depth) compiled by Hatori (19) and Shuto (197a, 1993), and the fragility curves (solid lines) obtained by the least-squares fitting of Eq. (). Since the historical data is highly dispersed, the dashed lines are also added to indicate the maximum and minimum limits by the authors interpretation (probably with less statistical meaning). The statistical parameters of fragility curves with the solid and dashed lines are summarized in Table. The high dispersion is probably caused by numerous uncertain factors in terms of the reliability of historical data. It is quite unknown, for example, how the tsunami heights were measured (datum) and represented in their reports, e.g. single or multiple measurements. Accordingly, these fragility curves should be interpreted as a coarse measure with uncertainty, e.g. m tsunami is equivalent to cause -3 % of probability that a house would be destroyed (Fig. 11(d)). In another way, by using fragility curves, the magnitude of tsunami.. 1...... (e) 19 Chile (Hatori) Historical data 5 1 15 Tsunami height (m) 19 Meiji (Shuto) Historical data 5 1 15 Tsunami height (m).. 1...... (f) All (Hatori) Historical data 5 1 15 Tsunami height (m) 19 Meiji (Shuto) Historical data Inundation depth (m) Fig. 11. Historical tsunami data of Hatori (19) and Shuto (197a, 1993), and tsunami fragility curves. (a) 19 Meiji Sanriku tsunami by Hatori, (b) 1933 Showa Sanriku tsunami by Hatori, (c) 19 Chile tsunami by Hatori, (d) Three events by Hatori, (e) and (f) 19 Meiji Sanriku tsunami by Shuto. Table. Statistical parameters of tsunami fragility curves (Fig. 11). µ,σ for the regression of the historical data (solid line) and µ,σ, µ,σ for upper and lower limits. curve µ σ R µ σ µ σ (a) 5. 3..3. 1.35 9. 5. (b) 5.9.5.7.9 1.35 7..5 (c). 1.9. N/A N/A N/A N/A (d) 5.97..55. 1.3 1. 3. (e).5.9...9 1.5 5. (f) 5.9 1..3 N/A N/A N/A N/A can be speculated from the documented damage, e.g. 3 % of structural damage would be potentially caused by the tsunami of.1-7. m height, as shown in Fig. 11(d), as an empirical relationships between tsunami hazards and local vulnerability learned from the historical Sanriku tsunami disasters. Historical tsunami fragility aims to identify the relationships among the tsunami hazards, the damage and uncertain historical documents. For instance, large numbers of descriptions can be found in historical documents, saying An abnormal tide reached to the entrance of a shrine 1 Journal of Disaster Research Vol. No., 9

Tsunami which was X m above the sea level or An abnormal tide penetrated in the village to cause Y houses washedaway. To identify the origin or cause of the descriptions above, the former requires the interpretation of potential damage equivalent to the reported tsunami height of X m. Also the latter requires the potential tsunami height equivalent to the reported amount of damage Y. 7. Concluding Remarks We proposed a new measure called Tsunami fragility throughout the statistical analysis of tsunami damage data interpreted from the high-resolution satellite images or field survey, numerical modeling and historical documents, to identify the relationship between tsunami hazard and vulnerability. Tsunami fragility is expressed by the structural damage probability or fatality ratio as the functions of hydrodynamic features of tsunami, such as inundation depth, current velocity and hydrodynamic force. Especially, the integration of satellite remote sensing and numerical modeling leads to a significant knowledge on structural vulnerability against the Sumatra- Andaman earthquake tsunami, in Banda Aceh, Indonesia. We suggest that tsunami fragility is implemented for an assessment of structural damage and fatalities within the exposed area against potential tsunami hazard scenarios. Multiplying the number of exposed structures and populations by the damage probability from the fragility curves equivalent to the estimated tsunami hazards provides the quantitative estimation of tsunami damage. It is still highly speculative, however, to say that the proposed tsunami fragility can become an universal measure of tsunami impact or damage. As stated in the introduction, the tsunami damage should be characterized by numerous uncertain factors. In this sense, tsunami fragility proposed here includes some of uncertainties, but not all. In other words, they may not be applicable in considering tsunami vulnerability when changing the areas of interest or considering other tsunami scenarios. Thus, we also suggest that careful use and interpretations are required in using proposed tsunami fragility when applying. We believe that further more precise investigations from the and other historical events can expand the applicability of tsunami fragility. Acknowledgements This research was financially supported, in part, by the Industrial Technology Research Grant Program in (Project ID : E51a) from New Energy and Industrial Technology Development Organization (NEDO), and the Grant-in-Aid for Scientific Research (Project Number : 19119) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan. References: [1] N. Shuto, Tsunami intensity and disasters, Tsunamis in the World, Kluwer Academic Publishers, pp. 197-1, 1993. [] H. Izuka and H. Matsutomi, Damage estimation due to tsunami inundation flow, Proceedings of Coastal Engineering, JSCE, Vol.7, pp. 31-35, (in Japanese). [3] R. Asakura, K. Iwase, T. Ikeya, M. Takao, N. Fujii, and M. Omori, An experimental study on wave force acting on on-shore structures due to over flowing tsunamis, Proceedings of Coastal Engineering, JSCE, Vol.7, pp. 911-915, (in Japanese). [] H. Yeh, Design Tsunami Forces for Onshore Structures, Journal of Disaster Research, Vol., No., pp. 531-53, 7. [5] J. Borrero, Field survey of northern Sumatra and Banda Aceh, Indonesia after the tsunami and earthquake of December, Seismological Research Letters, Vol.7, No.3, pp. 39-31, 5. [] H. Matsutomi, T. Sakakiyama, S. Nugroho, and M. Matsuyama, Aspects of inundated flow due to the Indian Ocean tsunami, Coastal Engineering Journal, Vol., No., pp. 17-195,. [7] Y. Tsuji, Y. Tanioka, H. Matsutomi, Y. Nishimura, T. Kamataki, Y. Murakami, T. Sakakiyama, A. Moore, G. Gelfenbaum, S. Nuguroho, B. Waluyo, I. Sukanta, R. Triyono, and Y. Namegaya, Damage and height distribution of Sumatra earthquake Tsunami of December,, in Banda Aceh city and its environs, Journal of Disaster Research, Vol.1, No.1, pp. 13-115,. [] K. Fujima, Y. Shigihara, T. Tomita, K. Honda, H. Nobuoka, M. Hanzawa, H. Fujii, H. Otani, S. Orishimo, M. Tatsumi, and S. Koshimura, Survey results of the Indian Ocean tsunami in the Maldives, Coastal Engineering Journal, Vol., No., pp. 91-97,. [9] K. Satake, T. T. Aung, Y. Sawai, Y. Okamura, K. S. Win, W. Swe, C. Swe, T. L. Swe, S. T. Tun, M. M. Soe, T. Z. Oo, and S. H. Zaw, Tsunami heights and damage along the Myanmar coast from the December Sumatra-Andaman earthquake, Earth, Planets and Space, Vol.5, pp. 3-5,. [1] O. Murao and F. Yamazaki, Development of fragility curves for buildings based on damage survey data of a local government after the 1995 Hyogoken-Nanbu earthquake, Journal of Structural and Construction Engineering, Vol.57, pp. 19-19, (in Japanese). [11] M. Shinozuka, M. Q. Feng, J. Lee, and T. Naganuma, Statistical analysis of fragility curves, Journal of Engineering Mechanics, Vol.1, pp. 1-131,. [1] Federal Emergency Management Agency, Multi-hazard Loss Estimation Methodology, HAZUS-MH MR3 (Earthquake Model), Technical Manual, pp. 5-35, 3. [13] Japan International cooperation Agency (JICA), The study on the urgent rehabilitation and reconstruction support program for Aceh province and affected areas in north Sumatra, Final Report (1), Vol.IV : Data book, 5. [1] T. Oie, S. Koshimura, H. Yanagisawa, and F. Imamura, Numerical modeling of the Indian Ocean tsunami and damage assessment in Banda Aceh, Indonesia, Annual Journal of Coastal Engineering, JSCE, Vol.53, pp. 1-5, (in Japanese). [15] S. Koshimura, T. Oie, H. Yanagisawa, and F. Imamura, Developing fragility functions for tsunami damage estimation using numerical model and post-tsunami data from Banda Aceh, Indonesia, Coastal Engineering Journal, No.3, pp. 3-73, 9. [1] I. Aida, Reliability of a tsunami source model derived from fault parameters, Journal of Physics of the Earth, Vol., pp. 57-73, 197. [17] K. R. Karim, and F. Yamazaki, A simplified method of constructing fragility curves for highway bridges, Earthquake Engineering and Structural Dynamics, Vol.3, pp. 13-1, 3. [1] M. Saatcioglu, A. Ghobarah, and I. Nistor, Performance of structures in Indonesia during the December great Sumatra earthquake and Indian Ocean tsunami, Earthquake Spectra, Vol., No.3, pp. S95-S319,. [19] Y. Namegaya and Y. Tsuji, Distributions of the Swept Away Houses in Banda Aceh City, Indonesia, due to the Indian Ocean Tsunami Estimated by Satellite Images, Annual Journal of Coastal Engineering, JSCE, Vol.53, pp. -9, (in Japanese). [] T. Iki, Field survey of Sanriku tsunami disasters, Bulletin of the Imperial Earthquake Investigation Committee, Vol.11, pp. 5-3, 19 (in Japanese). [1] F. Yamashita, Tragic histories of great Sanriku tsunami disasters, Seijisha, 13p. 19 (in Japanese). [] Earthquake Research Institute, Reports on the 3 March, Showa Sanriku tsunami, Bulletin of the Earthquake Research Institute, Tokyo Imperial University, Supplementary Vol.1, p., 193 (in Japanese). [3] Japan Meteorological Agency, Sendai Branch, Report of the May, Showa 35, Chilean earthquake tsunami, 191 (in Japanese). Journal of Disaster Research Vol. No., 9 7

Koshimura, S., Namegaya, Y., and Yanagisawa, H. [] Iwate Prefectural Government, Report of the recovery process from the 19 Chilean earthquake tsunami disaster, 51p., 199 (in Japanese). [5] T. Hatori, Damage probability of houses by tsunamis, Bulletin of the Earthquake Research Institute, Vol.59, pp. 33-39, 19 (in Japanese). [] N. Shuto, Evolution of tsunami disasters, Tsunami Engineering Technical Report, Tohoku University, Vol., pp. 1-1, 197a (in Japanese). [7] N. Shuto and C. Goto, Field survey of great Sanriku tsunami from Raga, Hiraiga, Shimanokoshi, Omoto and Shimokonari, Tsunami Engineering Technical Report, Tohoku University, Vol., pp. 39-5, 195a (in Japanese). [] N. Shuto and C. Goto, Field survey of great Sanriku tsunami from Okkirai, Tsunami Engineering Technical Report, Tohoku University, Vol., pp. -53, 195b (in Japanese). [9] F. Imamura, C. Goto, and N. Shuto, Study on Numerical Tsunami Forecasting System Accuracy of numerical models, Tsunami Engineering Technical Report, Tohoku University, Vol.3, pp. 3-7, 19 (in Japanese). [3] N. Shuto, J. Sayama, and K. Fujima, Field survey of great Sanriku tsunami from Ofunato, Tsunami Engineering Technical Report, Tohoku University, Vol., pp. 11-113, 197b (in Japanese). Name: Shunichi Koshimura Affiliation: Associate Professor, Graduate School of Engineering, Tohoku University Name: Yuichi Namegaya Affiliation: Postdoctoral Research Fellow, Geological Survey of Japan, National Institute of Advanced Industrial Science and Technology Address: Site 7, Higashi 1-1-1, Tsukuba 35-57, Japan Brief Career: 5-7 JSPS Research Fellow, Earthquake Research Institute, the University of Tokyo 7- Postdoctoral Research Fellow, Geological Survey of Japan, National Institute of Advanced Industrial Science and Technology Selected Publications: Y. Namegaya, Y. Tanioka, K. Abe, K. Satake, K. Hirata, M. Okada, and A. R. Gusman, In situ measurements of tide gauge response and corrections of tsunami waveforms from the Niigataken Chuetsu-oki earthquake in 7, Pure and Applied Geophysics, Vol.1, pp. 97-11, 9. Y. Namegaya and K. Satake, Tsunami generated by the 7 Noto Hanto earthquake, Earth, Planets and Space, Vol., pp. 17-13,. Academic Societies & Scientific Organizations: Japan Society of Civil Engineers (JSCE) Seismological Society of Japan (SSJ) Americal Geophysical Union (AGU) Address: Aoba --11-11, Aramaki, Aoba-Ku, Sendai 9-579, Japan Brief Career: - JSPS Research Fellow, National Oceanic and Atmospheric Administration -5 Research Scientist, Disaster Reduction and Human Renovation Institute 5- Associate Professor, Tohoku University Selected Publications: S. Koshimura, T. Oie, H. Yanagisawa, and F. Imamura, Developing fragility functions for tsunami damage estimation using numerical model and post-tsunami data from Banda Aceh, Indonesia, Coastal Engineering Journal, No.3, pp. 3-73, 9. S. Koshimura, Y. Hayashi, K. Munemoto, and F. Imamura, Effect of the Emperor seamounts on trans-oceanic propagation of the Kuril Island earthquake tsunami, Geophysical Research letters, Vol.35, L11, doi:1.19/7gl319,,. S. Koshimura, T. Katada, H. O. Mofjeld, and Y. Kawata, A method for estimating casualties due to the tsunami inundation flow, Natural Hazards, Vol.39, pp. 5-7,. Academic Societies & Scientific Organizations: Japan Society of Civil Engineers (JSCE) Institute of Social Safety Science Japan Associaiton for Earthquake Engineering (JAEE) Japan Society for Computational Engineering and Science (JSCES) Americal Geophysical Union (AGU) Name: Hideki Yanagisawa Affiliation: Company Member, Tokyo Electric Power Services Company Limited Address: Higashi-Ueno 3-3-3, Taito-ku, Tokyo 11-15, Japan Brief Career: -9 Post-Doctoral Research Fellow, Graduate School of Engineering, Tohoku University 9- Company Member, Tokyo Electric Power Services Company Limited Selected Publications: H. Yanagisawa, S. Koshimura, K. Goto, T. Miyagi, F. Imamura, A. Ruangrassamee, and C. Tanavud, Damage of mangrove forest by the Indian Ocean tsunami at Pakarang Cape and Namkem, Thailand, Polish Journal of Environmental Studies, Vol.1, No.1, pp. 35-, 9. H. Yanagisawa, S. Koshimura, K. Goto, T. Miyagi, F. Imamura, A. Ruangrassamee, and C. Tanavud, The reduction effects of mangrove forest on a tsunami based on field surveys at Pakarang Cape, Thailand and numerical analysis, Estuarine, Coastal and Shelf Science, Vol.1, pp. 7-37, 9. Academic Societies & Scientific Organizations: Japan Society of Civil Engineers (JSCE) Japan Society for Mangroves Journal of Disaster Research Vol. No., 9