ēȰÕʸt΂͵Ĝz 31 Ē620147 / Report of Tsunami Engineering Vol.316 20147 49 - 57 49

>H6HI:GY,6GC>CC9:Y;GDBY)NE=DDCY,>E=6Y6IYOJ$H=>B6Y>CY

Natt Leelawat*, Anawat Suppasri**, Mari Yasuda**, and Fumihiko Imamura**

YCIGD9J8I>DC much disaster prevention has been prepared, we cannot guard down the disaster awareness because Early warning is one of the priority actions anything can be happening. The comprehensive addressed by the Hyogo Framework for Action warning systems require the following actions: (UNISDR, 2007). The traditional disaster warning 4Detect the presence of a hazard5,4 Assess the system is likely to be the4 linear communication threat posed by that hazard5,4 Determine the systems5 (i.e., starting from monitoring and population facing risk from that hazard5,4 Inform detection systems, followed by data analysis and the population5,4 Determine appropriate prediction, and nally communication to public) protective actions that may be taken5,4 Direct the (Sorensen & Sorensen, 2007, p. 185). With the public to take those actions5, and4 Support the current various advanced technologies, there are actions taken by the public5 (Coppola 2011, p. many ways and sources where people can access 280281). In this study, we would like to zoom information. The public can receive warning in on the part of informing the population or so information from a multichannel approach called broadcasting. (Sorensen & Sorensen, 2007). No matter how

Fig. 1 IzuOshima Map Note. Source: Google Maps (2014a, b).

*Department of Industrial Engineering and Management, Graduate School of Decision Science and Technology, Institute of Technology, **International Research Institute of Disaster Science, Tohoku University, Japan 50 ēȰÕʸt΂͵Ĝz 31 Ē620147

In midOctober 2013, Typhoon Wipha (see Fig. 2). This issue catches the interest of (International Designation: 1326; Japanese many media and press, such as Ho lena (2013a, system: H25 Typhoon No. 26) extensively b), House of Japan (2013), Kyodo (2013), and attacked Japan. Among many typhoon Yamaguchi (2013). This study aims to understand encountered areas, IzuOshima (see Fig. 1), a the warning broadcasting system of this case 91.06km2 island, which is administered by the study through a business process management Tokyo Metropolitan Government, experienced (BPM) perspective. heavy rain, many landslides, and debris flows

Fig. 2 Typhoon Wipha3s Affected Area at IzuOshima. Note. Taken on January 20, 2014 at IzuOshima.

In Section 2, a review of the conceptual the physical interaction (a.k.a.4 Medium level5) background is described, followed by an (Perinforma, 2012). In order to understand the explanation of the research design and entire essence of the organization, the managerial methodology in Section 3. In Section 4, an perspective focuses on the ontological level, analysis of the case study is brie y summarized. while the infological level provides the analyzed Finally, the conclusions and suggestions are or calculated information and the datalogical level discussed in Section 5. provides the data used mainly by the operational perspective. The bottom level is the physical YDC8:EIJ6AY68@

Fig. 3 Interview at IzuOshima Town Of ce. Note. Taken on January 20, 2014 at IzuOshima

Y,6GC>CCH extraordinary natural phenomena5 (JMA, 2013a). For the meteorological disasters (i.e., heavy rain, According to JMA (2013a, c) the terminology heavy snow, high waves, storm, storm surge, of warnings in Japan has different criteria among snowstorm), there are three indices of criteria the disaster type. The summary is presented (JMA, 2013c): (1) rain (for heavy rain), (2) in Table 1. Since August 30, 2013, JMA has typhoon or similar (for heavy rain, high waves, launched the2 Emergency Warnings3 for alerting snowstorm, storm, storm surge), and (3) snow (for people to4 possible catastrophes caused by heavy snow). 52 ēȰÕʸt΂͵Ĝz 31 Ē620147

Table 1. Japan3s Disaster Warning Classi cation (JMA, 2013a, c) Disaster Type Warning Classi cation Description of Criteria Earthquake Emergency warning Seismic intensity of 6lower or more Warning Between seismic intensity of 5lower and 6lower Forecast Between seismic intensity of 3 and 5lower or more than magnitude of 3.5 Heavy rain Emergency warning 4Heavy rainfall with a level of intensity observed only one once every few decades is predicted in association with a typhoon or similar5 or4 Heavy rainfall is predicted in association with a typhoon expected to have a level of intensity observed only once every few decades or an with comparable intensity5 Warning 4Chance of catastrophe5 Advisory 4Possible development of serious adverse conditions5 Heavy snow Emergency warning 4Heavy snowfall with a level of intensity observed only once every few decades is predicted5 Warning 4Chance of catastrophe5 Advisory 4Possible development of serious adverse conditions5 High waves Emergency warning 4High waves are predicted in association with a typhoon expected to have a level of intensity observed only once every few decades or an extratropical cyclone with comparable intensity5 Warning 4Chance of catastrophe5 Advisory 4Possible development of serious adverse conditions5 Snowstorm Emergency warning 4A snowstorm is predicted in association with an extratropical cyclone expected to have a level of intensity observed only once every few decades5 Warning 4Chance of catastrophe5 Advisory 4Possible development of serious adverse conditions5 Storm Emergency warning 4A storm is predicted in association with a typhoon expected to have a level of intensity observed only once every few decades or an extratropical cyclone with comparable intensity5 Warning 4Chance of catastrophe5 Advisory 4Possible development of serious adverse conditions5 Storm surge Emergency warning 4A storm surge is predicted in association with a typhoon expected to have a level of intensity observed only once every few decades or an extratropical cyclone with comparable intensity5 Warning 4Chance of catastrophe5 Advisory 4Possible development of serious adverse conditions5 Tsunami Emergency warning Tsunami is over 3 m Warning Tsunami is between 1 m and 3 m Advisory Tsunami is between 20 cm and 1 m Forecast No tsunami to slight sea level changes Disaster Warning Broadcasting Process Management: A Case Study of Landslide from Typhoon Wipha at IzuOshima in 2013 53

Volcanic Emergency warning Residentialarea warning;4 Eruption or possibility of eruption eruption that may cause serious damage in residential areas and non residential areas nearer the crater5 or level 4 & 5 of volcanic alert Warning For nonresidential areas near the crater,4 Eruption or possibility of eruption that may severely affect places near residential areas (possible threat to life in such areas)5; for around the crater, 4Eruption or possibility of eruption that may affect areas near the crater (possible threat to life in such areas)5 or level 2& 3 of volcanic alert Forecast 4Volcanic ash emissions or other related phenomena may occur in the crater (possible threat to life in the crater)5 or level 1 of volcanic alert

Fig. 4 OCD of Warning Broadcasting System (Pattern #1) for nonearthquake, tsunami, volcanic eruption

Based on the interview data, we analyzed the executes the disaster monitoring. When the system following the DEMO modeling language. possible disaster is detected, the disaster monitor We drew the OCDs of the warning broadcasting manager requests a warning announcer of system. We found that there are two patterns of prefecture (i.e., actual actor: Tokyo Metropolitan the broadcasting system for this case. The first Government) to announce/relay a warning, one is the OCD of the normal disaster scenario depending on the criteria (see Table 1). In the case (as happened in midOctober 2013) (Pattern #1) of a disaster other than an earthquake, a tsunami, and the second one is the scenario of earthquake, or a volcanic eruption, it is necessary to request tsunami, and volcanic eruption (Pattern #2). Fig. an evacuation announcer of town (i.e., Disaster 4 shows the OCD of the Pattern #1, and Fig. 5 Prevention Division of town office) to make an shows the OCD of the Pattern #2. evacuation announcement. Before announcing, As shown in Fig. 4, there are two boundaries the evacuation announcer of town has to request in the system because we would like to show a town permission executor (i.e., actual actor: the actor roles in the townlevel evacuation Mayor or Deputy Mayor), who makes a decision broadcasting system (i.e., IzuOshima Town whether to announce an evacuation message or level). The larger boundary represents the not, to provide permission for broadcasting. Then, warning broadcasting system. First, the disaster the evacuation announcer of town can announce monitor manager (i.e., actual actor: JMA) and request people to evacuate in case of the 54 ēȰÕʸt΂͵Ĝz 31 Ē620147

Fig. 5 OCD of Warning Broadcasting System (Pattern #2) for earthquake, tsunami, and volcanic eruption

Table 2. Comparison between Pattern #1 and Pattern #2 Characteristics Pattern#1 Pattern#2 Actor role 4 actor roles 2 actor roles Transaction 5 transactions 3 transactions Responsibility Responsibility belongs to A2, A3, and Responsibility belongs to A2 A4 (especially A4 in the scope of town) Time More time consumption between the Less time consumption between the rst first initiator and the last executor be initiator and the last executor because cause there are more transactions which there are less transactions require decision making emergency warning level. generated huge mudslides along a 2km stretch In Pattern #2, as shown in Fig. 5, the beginning of mountains, which destroyed dozens of houses. is quite similar to the previous pattern, except for According to the interview data and report the latter part, where the prefectural evacuation (Tokyo Metropolitan Government, 2013), there announcer (i.e., actual actor: Tokyo Metropolitan were some issues learned from this case. First, Government) can execute an evacuation in practice, if we look at the medium used in announcement directly and automatically the town evacuation announcing (T3), Tokyo through the alarms installed in many places on Metropolitan Government used a fax machine to the island. The characteristics of two patterns are relay the message from JMA to the town office summarized in Table 2. in the evening. However, the person in charge It can be seen that the type of disaster is used left the of ce during that time, and unfortunately, to determine the pattern of the broadcasting it was a period of heavy rain. After 6 hours had process because the three severe disasters (i.e., passed, this message was received when the Chief earthquake, tsunami, and volcanic eruption) of Disaster Prevention came back to the office require immediate evacuation once any of them the next day (October 16th) at approximately occur. At the same time, other disasters require 1:30 a.m.; nevertheless, it was too late to issue more decisionmaking transactions at the town the evacuation at that time because it was in the level, even for the emergency warning level. middle of the night, and the rain was too heavy In our case study, when Typhoon Wipha to evacuate. Second, the actor of permission attacked Japan, the storm and heavy rain executor of the town actor role was not on the Disaster Warning Broadcasting Process Management: A Case Study of Landslide from Typhoon Wipha at IzuOshima in 2013 55 island due to a business trip; therefore, the town received the faxed message at the local office. permission execution could not be executed in a The suggestion is to change the assignment of the short time. As a result, there was no evacuation actor (as shift work) to this actor role for 24/7; announcement on the island. Landslides and then, the request can be received (and promised) mudslides caused 35 fatalities and 6 missing at all times. people (JMA, 2013b). Apart from the previous suggestion, which concerned the issue of the ontological level, Y>H8JHH>DCHY6C9YDC8AJH>DCH we also can change the physical level or the medium of communication. Many new mobile or The disaster warningevacuation broadcasting communication technologies can be selected for system in the Typhoon Wipha of IzuOshima use in this transaction. One possible approach is case has been analyzed by applying DEMO. to use technology that will allow the local of cer In addition, some suggestions are given in this to access the information anytime and anywhere. section. Along with Sorensen and Sorensen3s suggestion It can be understood that the current disaster (2007), because the advances in technology broadcasting system of IzuOshima consists during the information era has reached many of two patterns, based on the type of disasters. people, the trend of the disaster warning system While Pattern #2 (warning broadcasting can change to a multichannel approach, such system for earthquake, tsunami, and volcanic as mobile phones and devices (Sorensen & eruption) applies the automatic alarm, Pattern Sorensen, 2007). #1 (evacuation broadcasting system for non Lastly, as Coppola (2011) mentioned, making a earthquake, tsunami, and volcanic eruption) has decision whether to issue or announce a warning two other actor roles (i.e., the town evacuation is one of the most difficult steps for a disaster announcer and the town permission executor) and manager because it can create panic among two other transactions (i.e., the town evacuation the public or create a4 crying wolf5 problem. announcing and the town permission execution). Therefore, it requires time for consideration. By When comparing the two patterns (see Table 2), applying Pattern #2 to the2 emergency warning3 Pattern #2 requires less time due to the fewer level of all disaster types, it should reduce the numbers of actor roles and transactions. One time needed for the decisionmaking process. suggestion is the change of the business rule for cautioning. Instead of using an automatic alarm 8@CDLA:9<:B:CIH for only earthquakes, tsunamis and volcanic eruptions, the rule can be changed from the This study was supported by ACEEES Leading disastertypebased pattern to be based on the Program Educational Research Funds 2013 of level of impact or severity. The critical level of Tokyo Institute of Technology. The authors also disaster, which requires2 emergency warning3 would like to acknowledge Mr. Yoshihiko Kaneko level, can follow the Pattern #2 system, such that (Chief of Disaster Prevention Division of Izu the citizen can evacuate immediately. Oshima Town Office), Mr. Akimi Kaziya (JMA The completion of the transaction comprises of cer at IzuOshima Town Of ce), Mr. Toshiro request, promise, state, and accept (Dietz, 2006; Nakahayashi (Guide at IzuOshima), Mr. Shima Perinforma, 2012). From the OCD of Pattern Fujii (Tokyo Metropolitan Government); as well #1 (see Fig. 4), the town warning announcing as Prof. Junichi Iijima, Ms. Keiko Kitahara, lacked the2 promise3 step because no person and Iijima Laboratory of Tokyo Institute of 56 ēȰÕʸt΂͵Ĝz 31 Ē620147

Technology. go.jp/jma/en/Emergency_Warning/criteria.html Japan Meteorological Agency (JMA). (2013b, ':;:G:C8:H December). Member Report: Japan. The 8th Integrated Workshop/2nd TRCG Forum. Coppola, D. P. (2011). Introduction to Macau. International Disaster Management (2nd ed.). Japan Meteorological Agency (JMA). (2013c). Burlington, MA: ButterworthHeinemann. Relationships between criteria and indices. Dietz, J. L. G. (2006). Enterprise Ontology Retrieved Apr. 2014, from http://www. Theory and Methodology. Heidelberg: jma.go.jp/jma/en/Emergency_Warning/ SpringerVerlag. Relationships_between_criteria_and_indices. Google Maps. (2014a). [IzuOshima (in pdf Japanese)][Map]. Retrieved from https://www. Kamigaichi, O. (2004). JMA earthquake early google.com/maps/place/%E4%BC%8A%E8% warning. Journal of Japan Association for B1%86%E5%A4%A7%E5%B3%B6/@35.194 Earthquake Engineering, 4(3), 134137. 8844,139.6064894,9z/data=!4m2!3m1!1s0x60 Kyodo. (2013, October 17). The Japan Times. 1782d09e172939:0xe075b888a88821cb Despite deluge, no evacuation alert on island. Google Maps. (2014b). [IzuOshima (in Retrieved Mar. 2014, from http://www. Japanese)][Earth]. Retrieved from https://www. japantimes.co.jp/news/2013/10/17/national/ google.com/maps/search/%E4%BC%8A%E8 despitedelugenoevacuationalerton %B1%86%E5%A4%A7%E5%B3%B6/@34.7 island/#.Uyvkb1GSy0c 408671,139.3912632,25811m/data=!3m1!1e3 Leelawat, N., Suppasri, A., & Imamura, F. Hofilena, J. (2013a, October 23). Japan Daily (2013). Analyzing the essence of the disaster Press. Japan’s continue reliance on fax warning system in Japan. In Proceedings of technology to blame for lack of mudslide the Second International Education Forum warning?. Retrieved Jan. 2014, from http:// on Environment and Energy Science 2013. japandailypress.com/japanscontinued California. relianceonfaxtechnologytoblamefor Ozaki, T. (2011). Outline of the 2011 off the lackofmudslidewarning2338358/ Pacific coast of Tohoku Earthquake (MW Hofilena, J. (2013b, October 28). Japan 9.0) – Tsunami warnings/advisories and Daily Press. Japan PM Abe designates Izu observations –. Earth Planets Space, 63(7), Oshima as severe disaster area following 827830. Typhoon Wipha. Retrieved Apr. 2014, from Perinforma, A. P. C. (2012). The essence of http://japandailypress.com/japanpmabe organization Version 1.2. South Holland: Sapio designatesizuoshimaasseveredisaster Enterprise Engineering. areafollowingtyphoonwipha2838573/ Sorensen, J. H., & Sorensen, B. V. (2007). House of Japan. (2013, October 22). Mudslide Community processes: Warning and warning not relayed during Oshima Typhoon. evacuation. In H. Rodríguez, E. L. Quarantelli, Retrieved Apr. 2014, from http://www. & R. R. Dynes (Eds.), Handbook of Disaster houseofjapan.com/local/mudslidewarning Research (pp. 183199). New York, NY: notrelayedduringoshimatyphoon Springer. Japan Meteorological Agency (JMA). (2013a). Tokyo Metropolitan Government. (2013, Criteria for Issuance of Emergency Warnings. December). Effort for Oshima’s emergency Retrieved Sep. 2013, from http://www.jma. recovery (in Japanese). Tokyo: Tokyo Disaster Warning Broadcasting Process Management: A Case Study of Landslide from Typhoon Wipha at IzuOshima in 2013 57

Metropolitan Government. could have saved lives in Japan typhoon. United Nations International Strategy for Retrieved Apr. 2014, from http://www. Disaster Reduction (UNISDR). (2007). Hyogo valdostadailytimes.com/todaystopstories/ framework for action 20052015: building x2112899561/Evacuationcouldhavesaved the resilience of nations and communities to livesinJapantyphoon disaster. Geneva: UNISDR Yamasaki, E. (2012). What we can learn from Yamaguchi, M. (2013, October 18). Evacuation Japan3s early earthquake warning system. Momentum, 1(1).