The Role of Social Media in the 2011 East Japan Earthquake and Tsunami and Its Potential Future Application

The Role of Social Media in the 2011 East Japan Earthquake and Tsunami and its Potential Future Application

February 23, 2012

Graduate School of Global Environmental Studies Kyoto University

Master Course in Environmental Management Brett Douglas Martin PEARY

Table of Contents

Acknowledgements ...... vii List of Figures ...... ix List of Tables ...... xiii List of Acronyms ...... xv Executive Summary ...... xvii Chapter 1 Introduction ...... 1 1.1 BACKGROUND ...... 1 1.2 SIGNIFICANCE OF THE RESEARCH ...... 1 1.3 RESEARCH OBJECTIVES ...... 2 1.4 COMPONENTS OF RESEARCH THESIS ...... 2 1.4.1 Literature Review ...... 2 1.4.2 Collected Examples of Social media Use in the Disaster ...... 2 1.4.3 Framework ...... 3 1.4.4 Questionnaire Survey ...... 3 1.5 STRUCTURE OF THE THESIS ...... 3 Chapter 2 Literature Review ...... 5 2.1 SOCIAL MEDIA IN DISASTER ...... 5 2.1.1 Background on Social Media ...... 5 2.1.1.1 What Social Media Is ...... 5 2.1.1.2 The Rise Of Social Media ...... 6 2.1.1.3 Positive and Negative Issues Associated with Social Media ...... 8 2.1.2 Social Media in Recent Disasters ...... 9 2.1.2.1 The 7/7 Bombings in London ...... 9 2.1.2.2 The 2007 San Diego Fires ...... 10 2.1.2.3 Typhoon Ondoy/Ketsana ...... 11 2.1.2.4 The 2010 Haiti Earthquake ...... 11 2.1.2.5 2010 Yushu Earthquake ...... 12 2.1.2.6 Super-typhoon Megi ...... 12 2.1.2.7 2010 Eruptions of Mount Merapi ...... 13 2.1.2.8 2010 Canterbury (Christchurch) Earthquake ...... 13 2.1.2.9 Hurricane Irene ...... 13 2.1.2.10 2011 Virginia Earthquake ...... 13 2.1.3 Past Research on Social Media in Disaster ...... 14 2.2 DISASTER, COMMUNICATIONS AND SOCIAL MEDIA IN JAPAN ...... 17 2.2.1 Disaster in Japan ...... 17 2.2.1.1 Background ...... 17 2.2.1.2 The 1995 Great Hanshin Earthquake ...... 19 2.2.2 The Japanese Communication Landscape ...... 19 2.2.2.1 Overview of Usage of the Internet, Mobile phones and Social Media in Japan ... 19 2.2.2.2 Disaster Communication/Information systems ...... 20

iii

Chapter 3 Overview of the 2011 East Japan Earthquake and Tsunami and the Use of Social Media in the Disaster ...... 29 3.1 OVERVIEW OF THE DAMAGE ...... 29 3.2 THE USE OF SOCIAL MEDIA IN THE 2011 EAST JAPAN EARTHQUAKE AND TSUNAMI ...... 34 3.2.1 Overview of Social Media Use in the Disaster ...... 34 3.2.2 Coverage by the Mass Media ...... 36 3.2.3 Elements Related to the Examples of Social Media Use in the Disaster ...... 37 3.2.4 Previous Research Related to Social Media Use in the Disaster in Japan ...... 38 3.2.5 Examples ...... 39 3.2.5.1 Services and software ...... 40 3.2.5.2 Methods ...... 55 3.2.5.3 Group or Organization-oriented Social Media Use ...... 61 3.2.5.4 Type of Information Shared through Social Media ...... 64 3.2.6 Misinformation ...... 71 Chapter 4 Assessment of the Role of Social Media in the 2011 East Japan Earthquake and Tsunami ...... 79 4.1 Methodology ...... 79 4.1.1 Collected Examples of Social Media Use in the Disaster ...... 80 4.1.2 Analysis Framework ...... 81 4.1.2.1 Factors, Parameters, and Indicators ...... 81 4.1.3 Online Questionnaire Survey ...... 83 4.1.3.1 Types of Respondents to the Questionnaire Survey ...... 84 4.1.3.2 Layout of the Questionnaire ...... 87 4.1.3.3 Elements of the Questionnaire ...... 87 4.1.3.4 Method of Distribution ...... 88 4.1.3.5 Limitations of the Study ...... 90 4.2 Survey Results ...... 91 4.2.1 Results by user role ...... 91 4.2.1.1 Attributes of Respondents ...... 91 4.2.1.2 Source and Type of Information ...... 96 4.2.1.3 Motivation in Receiving and Sending Information ...... 101 4.2.1.4 Problems Encountered during Receiving and Sending of Information ...... 102 4.2.1.5 Evaluation of the Usefulness of Social Media in this Disaster and in Disasters in General 103 4.2.1.6 Areas of Information for Improvement and Support for Governmental Use of Social Media in Disasters ...... 104 4.2.2 Analysis Comparison by Location and Level of Affectedness ...... 105 4.2.2.1 Location ...... 105 4.2.2.2 Level of Affectedness ...... 110 4.3 Synthesis of Summary of Key Findings ...... 114 4.3.1 Main Results ...... 114 4.4 Key Findings ...... 120 4.5 Key Results ...... 121 Chapter 5 Discussion and Conclusion ...... 123 5.1 Discussion ...... 123

5.1.1 Reasons Organizations and Government Bodies Should Utilize Social Media for Disaster Preparedness, Relief, and Recovery ...... 123 5.1.2 The Flow of Information in Disasters ...... 125 5.1.3 Social Media Use from Inside and Outside of the Disaster ...... 126 5.1.4 Social Media use throughout the Disaster Cycle ...... 127 5.1.5 Social Media Tools ...... 127 5.1.6 Recent Trends and Changes which Might Affect Adoption of Social Media Use during Disaster ...... 128 5.1.7 Concerns, Barriers and Problems Related to the Use of Social Media in Disaster ...... 129 5.2 Conclusion and Policy Implications ...... 130 5.2.1 Conclusion ...... 130 5.2.2 Methods of Future Application of Social Media in Disasters by Organizations and Governments ...... 130 Appendices ...... I Appendix 1: Questionnaire survey on the use of social media in the 2011 East Japan Earthquake and Tsunami (English) ...... III Appendix 2: A survey on the use of social media in the 2011 East Japan Earthquake and Tsunami (Japanese) ...... V

Acknowledgements

I am deeply grateful to my academic advisors Professor Masami Kobayashi and Professor Hirohide Kobayashi of the Global Environmental Architecture Laboratory, Graduate School of Global Environmental Studies, Kyoto University for the opportunity to carry out my studies and for their guidance and continual help.

I would also like to express my appreciation to Professor Rajib Shaw of the International Environment and Disaster Management Laboratory (IEDM), Graduate School of Global Environmental Studies, Kyoto University for the opportunity to perform my research, his guidance and help in the formation of my research.

I would also like to express my gratitude to Dr. Yukiko Takeuchi for her help in formulating the ideas in my research and help with areas related to the Japanese language in my research.

I would like to deeply thank PHD candidates Jonas Joerin and Farah Mulyasari of the IEDM lab for the hours and hours they spent with me helping to polish my research and explain the research process.

I am deeply grateful to Professor Tracey Gannon, Professor Jane Singer, and Professor Andreas Neef for their constant encouragement, interest in my ideas, guidance, and for listening to me.

I would also like to thank Professor Chiho Ochiai and Ayako Fujieda for their help as well as the members of the International Environment and Disaster Management Laboratory and the Global Environmental Architecture Laboratory.

I am eternally grateful to the Ministry of Education, Culture, Sports, Science and Technology of the government of Japan for the generous scholarship, which has allowed me to perform my research and achieve new goals in my life. I will pay it back someday.

I am grateful for the support of my family, loved ones, and friends who have believed in me helped me come to this point, no matter how unique the road has been.

Lastly, I want to thank all the people who helped during this disaster, with social media or not, and recognize the great loss that is the basis of this research. I know no matter how much I learn, I can never truly understand.

vii

viii

List of Figures

Figure 1-1: Diagram of chapters ...... 4 Figure 2-1: Internet users, 2000-2010, world and by level of development ...... 7 Figure 2-2: Mobile-cellular subscriptions, 2000-2010, world and by level of development ...... 8 Figure 2-3: Tweet from KPBS News during the 2007 San Diego wildfires (Gibbs, 2007) ...... 10 Figure 2-4: Google Map mashup by KPBS News during the 2007 San Diego wildfires ...... 10 Figure 2-5: Overview of natural disasters in Japan from 1980 to 2010 ...... 17 Figure 2-6: Human exposure to hazard zones in Japan ...... 17 Figure 2-7: Map preview of Global Risk Data Platform for Japan showing areas of risk throughout Japan ...... 18 Figure 2-8: Japan’s risk profile ...... 18 Figure 2-9: Image of the J-ALERT SYSTEM ...... 20 Figure 2-10: Instructional image of the use of the saigai dengon dayaru ...... 22 Figure 3-1: Map of affected area ...... 29 Figure 3-2: Location of the main shock and aftershocks ...... 30 Figure 3-3: The tsunami sweeps through Natori, Miyagi Prefecture ...... 31 Figure 3-4: The tsunami striking Iwanuma, Miyagi Prefecture ...... 31 Figure 3-5: Dead and missing in the disaster ...... 32 Figure 3-6: Deaths by age in the disaster ...... 32 Figure 3-7: Area of the evacuation zone around Fukushima Daiichi Nuclear Power Plant ...... 33 Figure 3-8: A tweet with #prayforjapan asking for a donation to the Red Cross ...... 35 Figure 3-9: Website of Anpi Report, Anpi.tv ...... 36 Figure 3-10 : Helpfulness of social networks in the 2011 East Japan Earthquake and Tsunami . 39 Figure 3-11: Twitter, Facebook and Mixi use after the 2011 East Japan Earthquake and Tsunami by respondents ...... 39 Figure 3-12: Volume of @replies traveling in and out of Japan in a one-hour period before and after the earthquake ...... 42 Figure 3-13: Spread of information of worldwide retweets of Tweets originating in Japan for one hour after the earthquake ...... 42 Figure 3-14: An example of hashtag use during the disaster ...... 43 Figure 3-15: Portion of the Twitter blog post explaining hashtags for the disaster ...... 43 Figure 3-16: Portion of the Twitter blog post explaining how to use hashtags and GPS to get help ...... 43 Figure 3-17: Images of the Twitter blog post for mobile phone ...... 44 Figure 3-18: Image of a video graphic from produced by Facebook showing status updates related to the disaster ...... 45 Figure 3-19: Global Disaster Relief on Facebook ...... 46 Figure 3-20: Information related to the disaster provided by Facebook on users' News Feeds .... 47 Figure 3-21: A Mixi profile ...... 48 Figure 3-22: A Mixi group dedicated to the disaster with 346,333 members ...... 48 Figure 3-23: Image of Mixi's sign in page directing users to Google Person Finder and asking people to conserve energy ...... 49 Figure 3-24: Image of Mixi's list of groups related to the disaster ...... 49 Figure 3-25: “The East Japan Earthquake and Tsunami (Tohoku area Pacific Ocean Earthquake)@wiki” ...... 51 Figure 3-26: Olive wiki website with directions on how to make a mask from a shirt ...... 51 Figure 3-27: Image of a Twitter application for iPhone taken right after the earthquake with users commenting that an earthquake had occurred ...... 52 Figure 3-28: Softbank iPhone disaster message board application ...... 53

Figure 3-29: Image if the Sinsai.info website ...... 54 Figure 3-30: Image showing part of Safecast.org website with logo and radiation crisis map ...... 56 Figure 3-31: Radiation crisis map using Google Earth from Pachube.com ...... 56 Figure 3-32: An example of a social network group ...... 57 Figure 3-33: Mobile phone webpage info list, "Hang in there TOHOKU" ...... 58 Figure 3-34: Aggregator type website SAVE JAPAN! showing list Twitter hashtag keyword searches ...... 58 Figure 3-35: An image of the website Fanfan.jp soliciting tool and app suggestions from Twitter users ...... 59 Figure 3-36: Relief goods box label design ...... 59 Figure 3-37: Google Person Finder ...... 61 Figure 3-38: Live streaming of TBS News on Ustream after the disaster ...... 62 Figure 3-39: Image of NHK's website for affected and unaffected people to share messages with each other ...... 62 Figure 3-40: Image of NHK's website for sharing informational videos related to the disaster ... 63 Figure 3-41: Image of Peace Boat's Tohoku Kanto Emergency Relief Facebook page ...... 63 Figure 3-42: Slides from a Relief 2.0 presentation at TedxEarthquake 9.0 ...... 63 Figure 3-43: Image of a portion of the webpage for the group Japan Crisis Housing on Couchsurfing.org ...... 65 Figure 3-44: A portion of the Facebook group Beds in Osaka ...... 66 Figure 3-45: Image of the website Todoke showing a donation offer ...... 66 Figure 3-46: An image of the Twitter-sourced book "2:46 — Aftershocks: Stories from the Japan Earthquake" ...... 67 Figure 3-47: Image of a widget for the Go Setsuden project by Google showing actions that users can choose to do ...... 68 Figure 3-48: Image of website where translation service was provided by Babelverse ...... 69 Figure 3-49: An image of the "homemade" sign language broadcast on the program "Close Up Gendai" on NHK ...... 69 Figure 3-50: An image of the website of ilove.cat, a social media powered website that finds lost pets after the disaster ...... 70 Figure 4-1: Concept diagram of relationship of types of users of social media use in the 2011 East Japan Earthquake and Tsunami ...... 85 Figure 4-2: Concept diagram of types of users of social media use in the 2011 East Japan Earthquake and Tsunami, their larger groups and numbers of respondents for those groups ...... 86 Figure 4-3: Theoretical structure of the online questionnaire survey ...... 86 Figure 4-4: Pathways layout in the online questionnaire survey ...... 87 Figure 4-5: Image of the blog used to distribute the online questionnaire survey ...... 89 Figure 4-6: An Image of the Twitter account for the online questionnaire survey ...... 89 Figure 4-7: Image of the Facebook Page for the online questionnaire survey ...... 89 Figure 4-8: Image of tweets sent out through Twitter soliciting responses to the online questionnaire survey ...... 90 Figure 4-9: Language by user role ...... 92 Figure 4-10: Age by user role ...... 92 Figure 4-11: Gender by user role ...... 93 Figure 4-12: Location by user role ...... 94 Figure 4-13: Skill level with the Internet and social media by user role ...... 95 Figure 4-14: Level of affectedness by user role ...... 95 Figure 4-15: Most reliable source of information during the disaster ...... 97 Figure 4-16: Perceived most viable source of information for others in the 2011 East Japan Earthquake and Tsunami ...... 97 Figure 4-17: Types of information received during the disaster (multiple choice) ...... 98

Figure 4-18: Types of information sent during the disaster (multiple choice) ...... 99 Figure 4-19: Types of tools related to social media utilized during the disaster (multiple choice) 99 Figure 4-20: Original sources of received information (multiple choice) ...... 100 Figure 4-21: Original sources of sent information (multiple choice) ...... 100 Figure 4-22: Sources of received information besides social media (multiple choice) ...... 101 Figure 4-23: Reasons for utilizing social media to receive information (multiple choice) ...... 102 Figure 4-24: Reasons for utilizing social media to send information (multiple choice) ...... 102 Figure 4-25: Problems in receiving information through social media during the disaster (multiple choice) ...... 103 Figure 4-26: Problems in sending information through social media during the disaster (multiple choice) ...... 103 Figure 4-27: Level of helpfulness of social media during the disaster ...... 104 Figure 4-28: Overall impression of usefulness of social media during disaster in general ...... 104 Figure 4-29: Areas to be improved for receiving information in disasters (multiple choice) ...... 105 Figure 4-30: Support for government use of social media in disaster ...... 105 Figure 4-31: Relationship between location and most relied on source of information during the disaster ...... 106 Figure 4-32: Relationship between location and motivation ...... 107 Figure 4-33: Relationship between location and problems in receiving information ...... 107 Figure 4-34: Relationship between location and level of helpfulness of social media during the disaster ...... 108 Figure 4-35: Relationship between location and overall impression of the usefulness of social media during disasters in general ...... 108 Figure 4-36: Relationship between location and areas to be improved for receiving information in disasters ...... 109 Figure 4-37: Relationship between location and support for government use of social media in disasters ...... 109 Figure 4-38: Relationship between level of affectedness and most relied on source of information ...... 110 Figure 4-39: Relationship between level of affectedness and motivation ...... 111 Figure 4-40: Relationship between level of affectedness and problems in receiving information ...... 111 Figure 4-41: Relationship between level of affectedness and level of helpfulness of social media during the disaster ...... 112 Figure 4-42: Relationship between level of affectedness and overall impression of the usefulness of social media during disasters in general ...... 112 Figure 4-43: Relationship between level of affectedness and areas to be improved for receiving information in disasters ...... 113 Figure 4-44: Relationship between level of affectedness and support for government use of social media in disasters ...... 113 Figure 5-1: Diagram of the flow of information in disasters ...... 126

xii

List of Tables

Table 2-1: Overview of disasters in which social media was utilized ...... 14 Table 3-1: Summary of the 2011 East Japan Earthquake ...... 29 Table 3-2: Identified elements related to social media in the East Japan Earthquake and Tsunami ...... 37 Table 3-3: Summary table of services and software used and their characteristics ...... 54 Table 3-4: Summary table of organizational methods of the software and services used and their characteristics ...... 60 Table 3-5: Table of dominant information types shared through tools ...... 70 Table 4-1: Table outlining research conducted ...... 79 Table 4-2: Identified elements in relation social media in the disaster ...... 80 Table 4-3: Factors making up the framework for analysis and their corresponding parameters 82 Table 4-4: Table of indicators ...... 82 Table 4-5: Table outlining the online questionnaire survey conducted ...... 84 Table 4-6: Summary table of main results of analysis of user roles and factors ...... 116 Table 4-7: Summary of main results of analysis of location and factors ...... 118 Table 4-8: Summary of main results of analysis of level of affectedness and factors ...... 120

xiii

xiv

List of Acronyms

ANT Actor Network Theory

APAN All Partners Access Network

BBC British Broadcasting Corporation

CBC Canadian Broadcasting Corporation

CMC Computer Mediated Communication

EEW Earthquake Early Warning

EWB Emergency Warning Broadcast

FEMA Federal Emergency Management Association

ICT Information and Communication Technologies

ITU International Telecommunication Union

JMA Japan Meteorological Agency

JMIAC Japanese Ministry of Internal Affairs and Communication

MIC Ministry of Internal Affairs and Communications

MMD Mobile Media Data Labo

NHK Nihon Housou Kyoukai

PAGASA Philippine Atmospheric Geophysical and Astronomical Services

PBS Public Broadcasting Station

PHS Personal Handy Phone

SMS Short-Messaging Service

TISC Transnational Information Sharing Cooperation

UGC User Generated Content

UNISDR United Nations International Strategy for Disaster Reduction

USGS United States Geological Survey

VOIP Voice Over Internet Protocol

xvi

Executive Summary

Background and Purpose of the Study

During the 2011 East Japan Earthquake and Tsunami, newly popular social media such as Twitter and Facebook acted as a lifeline for directly affected individuals, a means of information sharing, and a way for people inside and outside Japan to volunteer and to provide information-based support to those affected individuals. Social media was used to perform vital relief functions such as safety identification, displaced-persons locating, damage information provision, support for disabled individuals, volunteer organization, fund-raising, and moral support systems as well as others. This study discusses the potential for involvement in disaster preparedness and response utilizing social media by public, civil society, and government organizations.

The flexibility of the tools and the ability for mass collaboration, or “crowdsourcing,” allowed for here-to-fore unimagined uses, in some cases accomplishing tasks that were not possible before. The use of social media in the disaster demonstrates the role it can play in disaster relief as well as its future potential through harnessing its power and applying it to create formerly unimagined disaster relief approaches. This use of social media in disasters is not a completely new phenomenon, but scale the disaster, the new popularity of the tools related to it, the particular situation of Japan’s communication landscape, and the scale of its use in the disaster are, and so there is a need to understand how social media was utilized in this disaster in particular.

Given the dramatic rise of the use of mobile phones and the Internet, the world is possibly on the precipice of a mass adoption of social media related communication as a way of life. Therefore, amongst the great tragedy and loss of the 2011 East Japan Earthquake and Tsunami, there is the opportunity to learn how social media was used in the disaster, its impact, and how it will likely be used in the future. These lessons can help to make more effective means of communication to reduce the negative impacts from disaster and better prepare for future ones.

The occurrence of the disaster in Japan, an industrialized nation with frequent disasters and a high level of social media use, presents a unique and rare insight into what may be the future role of social media in disasters as well. This is in contrast to the type of use of social media use occurring in disasters in developing nations. In addition, it may provide a model for NPOs and local and national governments to improve current disaster relief models.

To better understand this phenomenon, a research study was undertaken through a multi-part process. The first part of this process is a literature review, presenting background on social media, social media in recent disasters, past research related to social media in disasters, and other relevant information. Then data collection of examples of social media use in the 2011 East Japan Earthquake and Tsunami was undertaken. Following this, an online questionnaire survey on experiences related to social media use during the disaster was distributed in Japanese and English through social networks to individuals and organizations identified in the data collection process as well as to the general public. The framework of the online questionnaire survey was informed by the data collection process. The objectives of this research are:

xvii

• To identify problems associated with its use in disaster, such as the spread of misinformation and digital divide issues • To identify approaches of social media for disaster preparedness and response for NPOs and local and national government bodies for future utilization

These objectives are part of a greater goal to understand the development of social media use in disasters and identify its strengths and weaknesses, so that it can be improved and utilized more effectively in future disaster preparedness and disaster relief. To these ends, in addition, this research also aims to propose recommendations for pre- and post-disaster scenarios for the public, civil society, and government organizations.

Online Questionnaire Survey on Social Media Use in the 2011 East Japan Earthquake and Tsunami

Based on the collected examples of social media uses in the disaster, a framework for the online questionnaire survey targeted at social media users to obtain the opinions of individuals who used social media during the disaster was developed. The framework has two parts: an organizational structure based on the roles of users in the disaster and a method for measuring the responses of individuals fulfill these roles.

The first part, the organizational structure of the framework based on user roles, was created from identified roles of users of social media in the disaster. These roles are senders and receivers of information and more specifically non-user, non-sender, basic sender, volunteer sender, and manager sender.

The second part, the method of measuring the responses of the framework, consists of factors, parameters, and indicators. The factors are general areas for understanding to determine the role of social media in the disaster: information, motivation, problems, evaluation, and expectation. The parameters are specific groups of questions related to each of these factors and the indicators are specific sets of answers, aggregated to measure responses to the online questionnaire survey. The combination of these elements, the factors, parameters, and indicators, aim to give a well-rounded view of the role of social media in the disaster.

Assessment of the Role of Social Media in the 2011 East Japan Earthquake and Tsunami

Through the analysis of the online questionnaire survey and collected examples of social media use in the disaster a series of key results and key findings were produced.

Key findings show that over 60% of non-affected individuals, nearly 80% of indirectly affected individuals (Individuals who were not affected, but had family or friends were affected), and over 55% of directly affected individuals (Individuals who experienced the earthquake to a strong degree or had to evacuate due to the tsunami) responded that social media or the Internet was their most relied on source of information. While areas outside of the Tohoku region cited social media and the Internet combined at a rate near 70% (Kanto 71%, other areas in Japan 69%, areas outside of Japan 80%), individuals in the Tohoku region, the area affected by the tsunami, cited social media and the Internet combined at a much lower rate of 45%. Respondents in all areas in Japan evaluate “lack of trust in information” as the greatest problem associated with social media use in the disaster, with respondents in all locations rating it over 33%, whereas respondents outside of Japan rate it much lower at 14%. Users regardless of location (inside disaster areas and non-disaster areas) or level of affectedness (non-affected, indirectly affected, and directly affected) both overwhelmingly evaluated social media to be “helpful” or “extremely helpful” during the disaster at a combined rate of 94%. Support for government use of social media for respondents in the disaster affected areas (Tohoku and Kanto regions) is 95%, higher than those of other areas,

xviii

which is over 84%. Support for government use of social media in disasters is extremely high, above 80% for individuals in all locations, all levels of affectedness, and all user roles.

Using the results of the survey and the key findings d to the role of social media in the disaster, the results were analyzed and some key results were produced. The key results are as stated below:

• Social media and the internet were found to be reliable in the disaster to very high levels regardless of user role, location or level of affectedness and perceived reliability is even higher. • For directly affected individuals and individuals in the affected areas convenience and mass sending ability were the strongest reasons for utilizing social media during the disaster. • Problems with trustworthiness of information were experienced by all to a high degree. • Even though problems were experienced, users found social media to be extremely helpful in the disaster and beneficial in disasters in general to an overwhelming degree. • Users feel that all areas of information sharing through social media in the disaster need to be improved, infrastructure information to a particularly high degree. • Support for government use of social media in disasters is extremely high and is higher among directly affected individuals, individuals in disaster areas, and group senders. • The higher the level of participation in sharing of information through social media during the disaster, the more likely an individual is to received and share a broader amount of information, with a higher likelihood that that information is from credible sources.

Based on the key findings, in combination with the literature review and results from the data collection of examples of social media use in the disaster, results and the implications for the use of social media in disasters by public and governmental organizations were discussed. The discussion includes the merits and demerits of social media use in the disaster; the flow of information in disaster, the differences between social media use inside and outside of disasters; social media use throughout the disaster cycle; perspectives on social media tools in disasters; recent trends which might affect the adoption of social media use during disaster; concerns, barriers, and problems related to social media use in disaster; arguments for social media use in disaster preparedness, relief, and recovery by organizations and government bodies; and methods of future application of social media in disaster by organizations and government bodies

xix

Chapter 1 Introduction

1.1 BACKGROUND

Social media was used to perform vital relief functions such as safety identification, displaced-persons location, damage information provision, support for disabled individuals, volunteer organization, fund-raising, and moral support systems as well as others. The flexibility of the tools and the ability for mass collaboration, or “crowdsourcing,” allowed for here-to-fore unimagined uses, in some cases accomplishing tasks that were not possible before.

1.2 SIGNIFICANCE OF THE RESEARCH

Social media in past disasters has allowed for information dissemination and resources mobilization, as well as allow for public participation where users of social media can play vital roles as collaborators in disasters. It has been found that in some cases social media is a more efficient form of communication, demonstrating the possible governmental and public participation in disaster (Tseng, Chen & Chi, 2011). By understanding the role social media in the 2011 East Japan Earthquake and Tsunami a better understanding of how the public can participate as collaborators and how organizations and public bodies can participate more efficiently because of the unique situation surrounding this disaster, such as the scale the disaster, the new popularity of the tools related to it, the particular situation of Japan’s communication landscape, and the scale of its use in the disaster. Therefore there is a need to understand the role of social media utilization in this disaster in particular.

Other research related to the 2011 East Japan Earthquake and Tsunami has been performed by Acar and Murakami (2011), but their research specifically focuses on one social network, Twitter, whereas this research is intended to document and understand all types of social media use in the disaster. Therefore, this research is unique. In addition, this research has a

broader scope in terms of objectives and aims to present recommendations for organizations and local and national government bodies for effective utilization of social media in disasters.

1.3 RESEARCH OBJECTIVES

The objectives of this research are:

• To understand the role that social media played in the 2011 East Japan Earthquake and Tsunami, the type of information exchanged through it, motivations of its users, challenges associated with its use, whether users felt its use was valuable, and its potential uses in the future • To identify problems associated with its use in disaster, such as the spread of misinformation and digital divide issues • To identify approaches of social media for disaster preparedness and response for NPOs and local and national government bodies for future utilization

The above objectives are carried out through answering the following research questions:

• What types of information were exchanged through social media and at what rates? • What were the reasons for utilizing social media during the disaster? • What problems were encountered by users of social media during the disaster? • How do users of social media during the disaster evaluate their experience? • How do users of social media during the disaster think social media should be changed or utilized in the future?

1.4 COMPONENTS OF RESEARCH THESIS

1.4.1 Literature Review

Literature review was conducted on the background of social media, the use of social media in past disasters, past research on social media in disasters, and the situation regarding disaster, communication technologies, and social media use in Japan.

1.4.2 Collected Examples of Social media Use in the Disaster

Examples of social media use in the disaster were collected and organized for documentation as well as for the purpose of developing a framework for an online questionnaire survey, which was undertaken as part of this research. This data collection was performed over the period of time between March 11th, 2011 to September 13th, 2011.

1.4.3 Framework

Based on the collected examples of social media uses in the disaster, a framework for answering the research questions was developed. The framework has two parts: an organizational structure based on the roles of users in the disaster and a method for measuring the responses of individuals who fulfill these roles. This framework was used to design the online questionnaire survey for later analysis.

1.4.4 Questionnaire Survey

As part of this research an online questionnaire survey targeted at social media users was undertaken to collect the opinions of individuals who utilized social media during the disaster. The survey was made available for response on a webpage designed specifically for the survey over a one month period from September 13th, 2011, to October 13th, 2011. In order generate a better specified response, the survey was distributed to groups and individuals identified as having utilized social media in the disaster. Following this, general appeals were made through the social networks Twitter and Facebook by contacting users who had shared information related to the disaster. The survey was made available to all regardless of location or use of social media in the disaster. The survey was visited 1,538 times, and there were 206 respondents, 156 in Japanese and 50 in English.

1.5 STRUCTURE OF THE THESIS

The thesis consists of five chapters (See Figure 1-1). Chapter 1 is an introduction and lays out the background related to the study, the significance of the research, objectives, and methodology. Chapter 2 is a literature review consisting of explanations of the background of social media, an overview of social media use in past disasters, and an overview of communications and social media in Japan. Chapter 3 is an overview of the use of social media in the disaster with information on the disaster, examples of social media use in the disaster, and details on misinformation in the disaster. Chapter 4 is an assessment of the role of social media in the disaster and describes the research methodology, the process of collection of examples of the use of social media in the disaster, and results of analysis. Chapter 5 is discussion of the results and findings of the analysis as well as discussion of recommendations for pre- and post-disaster scenarios for the public, civil society, and government organizations. Followed by this is the conclusion.

Figure 1-1: Diagram of chapters

References

Acar, A. and Muraki, Y. (2011) Twitter for Crisis Communication: Lessons Learnt from Japan's Tsunami Disaster. International Journal of Web Based Communities. Retrieved October 31st, 2011 from http://www.eurekalert.org/pub_releases/2011- 04/ip-tan041511.php

Tseng, S., Chen, W., & Chi, C. (2011, June 21). Online Social Media in a Disaster Event: Network and Public Participation. Communications In Computer And Information Science: Digital Information and Communication Technology and Its Applications International Conference, DICTAP 2011. 167. 256-264.

Chapter 2 Literature Review

2.1 SOCIAL MEDIA IN DISASTER

2.1.1 Background on Social Media

2.1.1.1 What Social Media Is

In the simplest terms, social media is a set of applications and services that use the Internet to connect people. More specifically, Social media is a communication medium, made possible by the Internet, which combines dynamic, collaborative Internet-based tools, social networks, computers, and, increasingly, mobile devices. It allows users to connect to each other, exchange information and collaborate. It consists of social networks such as Twitter and Facebook, which act as a means of connection between users, and websites and applications that allow users to cooperate and create content, such as the websites Wikipedia and YouTube. Its use has grown in recent years along with the development and increased use of the Internet and mobile devices, such as smartphones. In addition, social media is constantly being redefined because of the evolving nature and the rapid change of the technologies.

Social media is media that “employ mobile and web-based technologies to create highly interactive platforms via which individuals and communities share, co-create, discuss, and modify user-generated content” (Kietzmann, Hermkens, McCarthy & Silvestre, 2011, p.241). It has been described as having the characteristics of participation, openness, conversation, community and connectedness (Mayfield, 2008, p. 2).

The term “social media” came about with the creation of social networking websites like MySpace and Facebook, but some tools, which are now considered to be part of social media, have been around for twenty years (Kaplan & Haenlein, 2010, p. 2). Social media has grown out of previous forms of user-publication based, web tools such as a discussion system called Usenet, which started in the late 1970’s, and blogs later in the 1990’s (Kaplan & Haenlein 2010). Social media is the content of Web 2.0, which is a term popularized by Tim O’Reilly of O’Reilly Media to describe the shift from an Internet made up of static web pages to a more complex and sophisticated one where users interact, collaborate, share information online and create content (O’Reilly, 2005).

With the development of Web 2.0, a number of tools that define it have evolved. Some basic forms of these tools, social media, are blogs, wikis, such as Wikipedia, podcasts, forums, content communities, such as YouTube and Flickr, and microblogging (Mayfield, 2008). These tools are all social media, part of the change to a Web 2.0 Internet, but social media is not limited to these alone.

Social networks, which is a form of social media and comprise a large portion of social media use were defined by Boyd and Ellison (2007) as such:

We define social network websites as web-based services that allow individuals to (1) construct a public or semi-public profile within a bounded system, (2) articulate a list of other users with whom they share a connection, and (3) view and traverse their list of connections and those made by others within the system. The nature and nomenclature of these connections may vary from website to website (p. 211).

They go on to explain the difference between “social network websites” and “social networking websites:”

While we use the term “social network website” to describe this phenomenon, the term “social networking websites” also appears in public discourse, and the two terms are often used interchangeably. We chose not to employ the term “networking” for two reasons: emphasis and scope. “Networking” emphasizes relationship initiation, often between strangers (p. 211).

Social media as a term has come to be widely used in recent years, yet it can be confusing, and still a large portion of the population is unfamiliar with it. In addition, due to the ever changing nature of the Internet and computer-related technologies, social media is constantly evolving and so what makes up social media can change. Therefore, it can have a variety of definitions. Kaplan and Haenlein support this and seek to clarify by stating that “there seems to be confusion among managers and academic researchers alike as to what exactly should be included under this term [social media], and how Social Media differ from the seemingly- interchangeable related concepts of Web 2.0 and User Generated Content” (2010, p. 60). They go on to say that social media is “a group of Internet-based applications that build on the ideological and technological foundations of Web 2.0, and that allow the creation and exchange of User Generated Content” (2010, p. 60).

Kaplan and Haenlein (2011) also clarify that social media differs from Web 2.0 and User Generated Content:

Web 2.0 is a term that was first used in 2004 to describe a new way in which software developers and end-users started to utilize the World Wide Web; that is, as a platform whereby content and applications are no longer created and published by individuals, but instead are continuously modified by all users in a participatory and collaborative fashion (p. 60-61).

And that:

User Generated Content (UGC) can be seen as the sum of all ways in which people make use of Social Media. When Web 2.0 represents the ideological and technological foundation, User Generated Content (UGC) can be seen as the sum of all ways in which people make use of Social Media. The term, which achieved broad popularity in 2005, is usually applied to describe the various forms of media content that are publicly available and created by end-users (p. 61).

Since many of these services and their definitions did not exist even a few years ago, and given the rapid adoption and evolution of social media, new forms of social media and the definition will likely 1change over time.

2.1.1.2 The Rise Of Social Media

Beginning around 2005, social media has grown in popularity to become a part of the daily lives of many people in recent years. Since the tools and content that make up social media are dispersed, it can be difficult to measure usage, but Facebook, the largest social network, which started in 2004, only became open to the public in late 2006 (Abram, 2006), as of September 2011, has 800 million active users, 50 percent of whom access the website daily (Facebook, 2011). Other social networks boast large numbers as well; Twitter, which functions in a very different way than Facebook and can be used with SMS (Short-Messaging Service) text messages through mobile phones, started in mid 2006 (Mashable, 2011) and has about 200 million users (Shiels, 2011). In addition, Google has entered social networking

with Google+ in June 2011 (Kaste, 2011), and as of October 14th, 2011, has 40 million users (Svelik, 2011).

The use of Internet and mobile technologies, which act as pathways and access points for social media, is also rising. According to the International Telecommunication Union (ITU, 2011), an agency of the United Nations that specializes in information and communication technologies, over the ten years from 2000 to 2010, the rate of Internet users world wide rose from less than ten per 100 people to nearly 30 per 100 people, with the rate rising from around 25 per 100 people to nearly 70 per 100 people in the developed world, and from less than ten per 100 people to 21 people per one hundred people in the developing world (See Figure 2-1).

Figure 2-1: Internet users, 2000-2010, world and by level of development Source: ITU Along with the growth in users of the Internet, users of mobile phones has also increased. According to the ITU, over 90% of the world’s population has available access to 3G mobile networks (2010). The ITU also says that 5 billion people have cellphone subscriptions and cellphones are “de facto ubiquitous” (“Cell phones ubiquitous,” 2011). Data from the ITU (2011) shows that mobile–cellular subscriptions are 114.2 to every 100 inhabitants in developed nations, 70 to every 100 inhabitants in developing nations and 78 to every 100 world wide (Figure 2-2). While much social media requires Internet data to be utilized, and in developing nations only one in every 20 people have a high-speed mobile subscription, compared with one in two people in developed nations, developing nations had the sharpest rise in adoption of high-speed mobile subscriptions (“Cell phones ubiquitous,” 2011). Still, some social networking websites like Twitter can be used with out an Internet data plan by using SMS, so all mobile phone users could be social media users potentially.

Figure 2-2: Mobile-cellular subscriptions, 2000-2010, world and by level of development Source: ITU The use of mobile-cellular phones that can use social media through mobile data networks is also increasing. As of 2011, 85% of mobile-cellular devices were able to access the mobile web (MobiThinking, 2011). Smartphones, which can connect to the Internet using applications and are commonly used with social media are becoming more popular also. Mobile-cellular-based Internet access is expected to overtake access from personal computers by 2015 (MobiThinking, 2010).

2.1.1.3 Positive and Negative Issues Associated with Social Media

Social media allows people to connect to and share information with their friends, family, and other users with similar interests. Social networks, the pathways, combined with the dynamic, web-based tools and user-generated information that make up social media have changed the way that a large amount of people communicate with friends and family, organizations and government. People can communicate, cooperate, exchange, interact, and manipulate data in ways that were never possible before. It has already changed the way that millions of people around the world communicate on a daily basis.

This potential has already been seen throughout the world in the past couple years. Although not originally designed to do so, its use has increasingly been felt in non-“social” areas, some credit it with helping to spur protests in Iran after the disputed election in 2009 (“Iran’s Twitter revolution,” 2009). In June of that year, the US State Department asked Twitter delay a network upgrade so that protesters in Iran could continue to use it to organize while demonstrations were going on (Grossman, 2009). Social media, Facebook in particular, is also being heralded as allowing organization that spurred the Arab Spring, the successive revolutions throughout the Arab world beginning in the Spring of 2011 (Huang, 2011).

Recognizing the potential of these tools, applications that leverage the power of the Internet and social media for social change have developed. One such case is the Ushahidi platform, which was originally designed to map occurrence of violence in Kenya after an election to map the location of violence using SMS and Google Maps so users could know what areas were dangerous. The non-profit company that created the tool, and has the same name, has gone on to use its “crisis-mapping” platform to aid in several disasters, such as the 2010 Haiti earthquake (Giridharadas, 2011).

Social media has great potential, but it also has issues for concern. Issues such as misinformation, digital divide (the gap between users and non-users, usually related to economics or age), privacy, identity theft, and misinformation are of concern when speaking

of social media. In order to minimize the negative impacts of these issues, further research is necessary.

2.1.2 Social Media in Recent Disasters

Social media has been used increasingly in disasters, natural and man-made in recent years. The increase in use of the Internet and the increase in use of network-connected mobile devices has lead to more people having access to social media. This increasingly larger group of individuals armed with instant access to the Internet means that when a disaster happens they are likely to use that avenue of communication because it is available to them at that moment. The way that the technology is used varies based on the type of disaster, location, situation, and resources available.

Individuals who make use of social media in disasters are limited by the types of services, tools, and number of users that they have available to them at the moment when the disaster occurs. Due to the fact that the change to Web 2.0 is gradual and that new innovative web services develop over time, we can see a difference between the early use of social media in disaster and later, more sophisticated uses in the ways problems are approached.

Experiences show that sometimes social media is the only functioning communication method in some disasters (Acar and Murakami, 2011). In addition, the availability of these new tools and a need for information during crisis spawns new organizational methods to accomplish tasks, some that were not possible or imaginable before the availability of a specific tool and the need to communicate in that specific situation. When disasters happen, users of social media use it to fulfill a need, and sometimes that need can require that they use the social media that they have at hand in a way that was not previously imagined by the designer. Luckily, social media’s eclectic and flexible nature lends itself to being “mashed up” and so these crises can creates greater innovation.

2.1.2.1 The 7/7 Bombings in London

When the coordinated suicide attacks in London, UK occurred on July 7th, 2005, it was arguably the first time that social media had been widely used in any kind of disaster. At the time, newly popular websites, Wikipedia and Flickr, along with recently available photo- capable mobile phones became tools in the hands of witnesses of the attacks (“Cell Phones Capture,” 2005). Photos taken by people on the scene were sent to others who using computers uploaded them to Flickr, a photo sharing website created the year before. In addition to the photos taken by people on the scene, the website showed images of news reports on television, as well as images expressing condolences to people affected by the attack.

Images were also sent by SMS from camera phones to the British Broadcasting Corporation (BBC) and were broadcasted around the world. Over 1000 pictures and 20 videos of first hand accounts were sent to the BBC in the hours following the disaster (Barnes & Carrell, 2005). The photos were used in the coverage of The New York Times (Cowell, 2005). This event popularized the idea of the “citizen reporter” (Huck, 2005), which is the idea that with new technologies such as blogs and camera-capable mobile phones that average persons could be “reporters” helping to cover stories when they happen to be on the scene.

Wikipedia, an online crowd-sourced encyclopedia, was being used in a similar matter, with hundreds of individuals editing it to update it with the latest known information. In the first day after its creation the Wikipedia page about the attack, “7 July London Bombings,“ was updated 2581 times (Massa, 2010), nearly once ever 30 seconds (Kuhagen, 2011).

It was possible to get a detailed picture of the disaster by looking at the first hand images, the photos of the news, and the Wikipedia page on the bombings, without ever seeing broadcast radio, television, or newspaper, which have been the main methods of communicating disaster up until that point.

2.1.2.2 The 2007 San Diego Fires

In October of 2007, a little over a year after the microblogging website Twitter was created, wildfires that eventually burned over 400,000 acres and 1500 homes broke out in southern California (Flaccus, 2007). Two San Diego residents began posting real-time, status updates on Twitter from multiple sources such as radio, television news, SMS, and other online sources on information, on as evacuations, meeting points, supply locations and other information (Poulsen, 2007).

At the same time, the website of the local public radio station, KPBS, was brought down because of too much traffic, 36 times the normal amount, from individuals looking for info on the state of the wildfires (Poulsen, 2007). In addition, several of the radio station’s towers were burned in the fire (Gibbs, 2007). The online editor of the website had been happening to be experimenting with Twitter and so employees of the radio station began to use Twitter to share information with residents (Figure 2-3). They also made a Google Map “mash-up,” which is a way that users of the Google Maps service can take data from various sources and place it on the maps to interact with the data, with locations of shelters, supplies and danger zones (Figure 2-4) (Poulsen, 2007).

Figure 2-3: Tweet from KPBS News during the 2007 San Diego wildfires (Gibbs, 2007) Source: KPBS’s Twitter account https://twitter.com/#!/kpbsnews

Figure 2-4: Google Map mashup by KPBS News during the 2007 San Diego wildfires Source: KPBS via http://www.nathangibbs.com/2007/10/23/san-diego-fire-updates/

2.1.2.3 Typhoon Ondoy/Ketsana

In September 2009, during typhoon Ondoy, which created the heaviest amount of rainfall in four decades and resulted in subsequent flooding that submerged 80% of Metro Manila, individuals inside and outside of the disaster used Twitter and other social media tools to perform relief functions and “transformed themselves from content disseminators into active rescue and relief workers” (Morales, 2010, p. 23). Individuals used Twitter to give updates on their safety and the situation around them and later communication changed into collaborative action. This action took the form of making online spreadsheets where individuals could enter up-to-date information on the disaster situation, such as information on evacuation centers, emergency rescue numbers, drop-off websites for donation, and lists of people in need of relief, as well as the creation an organization called Philippine Aid by Filipino bloggers in order to set up a PayPal account for online donations to help support disaster victims. The organization collected USD $12,500. Up until this point no system for online donations existed and this drove the Philippines Red Cross to create a PayPal account for donations in response. A crisis map was also set up using Google Maps that showed distress calls in real time. The crisis map ended up being an “invaluable resource” for relief workers in the field (Morales, 2010, p.23).

2.1.2.4 The 2010 Haiti Earthquake

On January 12th, 2010, a 7.0 magnitude earthquake hit Haiti’s capital Port-au-Prince, in which the damage amounted to 220,000 deaths, over 300,000 injuries, 180,000 collapsed homes, 1.5 million people made homeless and 3.5 million affected (“Haiti Earthquake Facts,” 2011). This is one of the recent disasters where the use of social media has been widely documented. When the disaster occurred, individuals from the non-profit company Ushahidi, which created “crisis-mapping” in response to post-election violence in Kenya, created a software platform to be used in Haiti and began to organize. They created a crisis-mapping system, and SMS- based, alternative 911, since none existed at the time, that uses information from the ground from Haitian people sent by SMS to a special number, 4636. SMS received by the team based in Boston, Massachusetts in the US, were aided by 12,000 Creole and French translators in 49 countries who were organized through a request sent out by the company through Facebook. With the help of the translators the team had a 5 to 10 second translation time for each of the 40,000 messages they received (PBS, 2011).

Through the crisis map the team can see, in general, what areas are facing what kind of problem, as well as respond to specific requests for supplies, medical aid or housing. The system was also used to track a cholera outbreak which began six months after the earthquake. By mapping SMS asking about medical help for cholera, the team can tell what areas still have a cholera outbreak (PBS, 2011).

The US State Department, recognizing the potential of the crisis-mapping system, organized with the Red Cross, the US Coast Guard, and Ushahidi. Then local radio was used, which 90% of Haitians use, to inform people that if they need help, they should text the 4636 number. In Haiti, 80% of people have mobile phones. There are no specific numbers on how many lives were saved, but the US state department claims lives were saved because of experience such as rescuing a woman giving birth in the middle of the street (PBS, 2011).

Since this time, the Ushahidi platform has been used in disasters in Chile, Pakistan, and Japan, and the United Nations asked the Ushahidi to make a team that can always be on call for when a crisis hits. The team was first used when the uprising in Libya kicked off in February of 2011. Sarah Farmer of the United Nations Global Pulse Group, which uses technology to track humanitarian crises said:

A team going into Libya would had nothing to go on. There was very little picture of what was happening on the ground. It's [the Ushahidi crisis-mapping system] changed the way that we saw what was going on (PBS, 2011).

Also during the crisis resulting from the earthquake, a new tool called the Transnational Information Sharing Cooperation (TISC), which grew out of a file sharing system for countries with a low level of infrastructure developed by the Pentagon called All Partners Access Network (APAN), was released and utilized by the US military’ Southern Command (Hodge, 2011). The tool was scheduled to be tested in a simulated hurricane in Haiti and the Dominican Republic around the time that the earthquake hit and so the tool went live to help in the earthquake relief. The tool is, in essence, a social network or “community” that becomes a hub for information sharing, made up of organizations and individuals, currently 1700 in Haiti, related to the relief effort (Pierce, 2011). The network is an unstructured, online community of web portals set up by the organizations and uses wikis, forums, SMS, and email (Redmon, 2010).

Due to the platform’s success during the relief efforts, the US military is looking to use crowdsourcing methods, accomplishing tasks through collaboration with large groups of people over the Internet much like the way Wikipedia functions, to aid in the disaster relief, much in the same way that they Ushahidi platform does. By converting crowdsourced data sent by individuals on the ground with SMS into meaningful information, the TISC platform can extend the type of goals it can accomplish and improve its efficiency with medical situations and public health by through “biosurveillance,” a method of using crisis-mapping to monitor infectious diseases. They are also looking to enhance the information with data- rich tools, such as a map that allows users to walk through an area visually, similar to the functionality of Google Street View, so that users can maneuver roads to a destination to see what pathways are blocked or open before leaving (Pierce, 2011). The APAN platform won the 2010 Forrester Groundswell Award for successful social technology (Redmon, 2010).

The ability of social media platforms like the Ushahidi and TISC to exchange information, organize and view that information to help affected people and relief organizations understand a constantly changing situation is changing the way that we approach disasters, natural and man-made. Only in recent years has infrastructure, in terms of levels of mobile phone usage and social media sophistication, existed, so these tools are at very initial stages of development. In the near future these systems will become more sophisticated and efficient, and change the way disaster relief and preparedness is thought about.

2.1.2.5 2010 Yushu Earthquake

On April 14th, 2010, in Yushu, Qinghai province, China, a 7.1 magnitude earthquake struck at 7:49 a.m. local time. Thousands of people were injured, 2,698 were killed, and 270 went missing. There was a large response on social media, in particular on Sina-Weibo, a Twitter- style microblogging website that is popular in China. Huang, Ou, Zhang and Zhang (2011) analyze 94,101 microblog posts and 41,817 re-posts related to the earthquake from Sina- Weibo during a 48-day period immediately after the earthquake.

2.1.2.6 Super-typhoon Megi

In October 2010, the Philippine Atmospheric Geophysical and Astronomical Services (PAGASA) launched a Twitter account just before a category 5 super typhoon, Megi, the tenth typhoon of the year, was due to hit land The organizations sent out tweets related to the typhoon via twitter and the information was picked up by traditional mass media and reported on. After a month, the organization had 28,000 followers on Twitter and later, before another typhoon, Ketsana, was due to hit the Twitter followers on their account increased by 3,000.

As of November 15th, 2010, the organization has 42.364 followers and Twitter and over 100,000 supporters on Facebook. In the Philippines where 20 typhoons hit a year, the social media accounts of the organization are able to keep people constantly updated as to disaster conditions. The organization says it is the most cost-efficient to make announcements to the public.

2.1.2.7 2010 Eruptions of Mount Merapi

In Indonesia in November of 2010, 300 people had been killed by lava flow caused by the eruptions of Mount Merapi on Java island, when a community-based organization, Jalin Merapi, which started as a community radio station to broadcast updates about the volcano’s eruptions, used Twitter to get the word out and organize. The organization has sent out 700 volunteers to locations that have not been able to receive government aid and has them report their needs to the organization and the community by using Twitter. In one case, a Twitter status update, or “tweet” was sent out saying that some coordinators in the organization had no way to transport relief food aid that was in a neighboring town, and within 10 minutes over a dozen vehicles showed up at the location in response to the update. In another, the organization sent out tweets saying that they need help to prepare 30,000 meals and because of volunteer response from Twitter users, the meals were prepared in four hours. As of November 22nd, 2010, the organization’s Twitter account had 35,000 followers (“Indonesians beat slow,” 2010). Alanna Shaikh of United Nations Dispatch points out in a blog post that the organization is mobilizing local resources and not “guiding an influx of foreign aid,” and that “the ability of social media to connect disaster survivors with local benefactors may decrease, over time, the need for international relief in response to disasters” (2010).

2.1.2.8 2010 Canterbury (Christchurch) Earthquake

At 12:51 p.m. on February 22nd, 2011, Christchurch, New Zealand experienced a 6.3 magnitude earthquake, causing 181 deaths (New Zealand Police, 2011). During the crisis Twitter was utilized by government, relief organizations, and residents, and a coordinated flow of recovery information was able to be sent out via Twitter to the residents. The quick recognition of an official hashtag denoting official information is seen to have been effective in sharing information. Twitter is seen as having played a notable role in the disaster, but also as having spread misinformation, so emergency authorities and media experts held a symposium to discuss how to better use social media in disasters (“Authorities learn to,” 2011). The symposium discussed topics such as development of an “Emergency 2.0 Wiki,” which comes out of a Social Media for Emergency Management project from that area’s Government 2.0 Taskforce. The project states that “citizen information seekers become information providers” when using social media in disasters (Collins, 2011).

2.1.2.9 Hurricane Irene

The Federal Emergency Management Association (FEMA), a governmental organization in the United States, encouraged people potentially affected by hurricane Irene in August of 2011 to use SMS or social networks to keep in touch with friends and family instead of calling and to make sure mobile phones are charged in preparation for the hurricane (“Irene Update 20,” 2011).

2.1.2.10 2011 Virginia Earthquake

When an 5.3 magnitude earthquake occurred on the eastern coast of the US on August 23rd, 2011, mobile phone service stopped working for a couple hours due to mass call load. Mobile

phone operators commented that no mobile phone towers were damaged and that the loss of service was due to too many users calling simultaneously, demonstrating the advantages of sharing information over SMS or social media as opposed to calling (Ford, 2011). In addition, some Twitter users in the northeast, close to New York City, were able to read tweets about the earthquake from users in Washington D.C. 30 seconds before they felt the earthquake, which originated in Virginia, because the information moves through the Internet quicker than the seismic waves of the earthquake (Ford, 2011).

Table 2-1: Overview of disasters in which social media was utilized Disasters in which social media Details on how social media was utilized in the disasters was used One of the first examples of social media use in a man-made disaster. Newly popular cell The 7/7 Bombings in London phone cameras were widely used. Flickr, a photo sharing site and Wikipedia were used. Popularized the ideas of the citizen reporter. Twitter was newly popular. Users in and outside of the fires used Twitter through SMS. Information on safe locations and supplies was aggregated by outside users and used to The 2007 San Diego make mash-ups with Google Maps. A large TV station relied on Twitter when its website crashed. Twitter was widely used. Online spreadsheets were used to share information on the disaster situation. PayPal was first used for online donations inspiring the Red Cross to Typhoon Ondoy/Ketsana adopt the practice. A crisis map was created with Google Maps and was called “an invaluable resource” by relief workers. The Ushahidi crisis map platform was widely used, helping to create an ad-hoc 911 system. 12,000 translators were recruited through Facebook, making for a 5 to 10 second The 2010 Haiti Earthquake turnaround time for incoming SMS messages. The map was used to track cholera outbreaks 6 months after the disaster. The system was widely used by relief organizations. A Twitter-clone site called Sina-Weibo was widely used to exchange information after 2010 Yushu Earthquake the earthquake. Information is limited because of language limitations with Chinese. The Philippine Atmospheric Geophysical and Astronomical Services (PAGASA) launched a Twitter account just before the typhoon. Their tweets were broadcasted Super-typhoon Megi traditional mass media. After a month, the organization had 28,000 followers. There are 20 typhoons a year in the Philippines and the organization says it is the most cost-efficient to make announcements to the public. A community radio station created to respond to an ongoing lava flow began using Twitter to organize. They were able to send 700 volunteers to places where government 2010 Eruptions of Mount Merapi aid was not getting. The organization had 35,000 followers and would make requests for help for driving and cooking and receive mass amounts of support from the community. Demonstrates the possibility of reducing dependency on foreign aid for communities. The government used Twitter to send a coordinated flow of recovery information to 2010 Canterbury (Christchurch) residents. An official hashtag keyword was instituted. Misinformation was also spread Earthquake and 2011 June using Twitter during the disaster. A symposium after the disaster about social media in Christchurch earthquake the disaster discussed creating an “Emergency 2.0 Wiki.” The US Federal Emergency Management Association (FEMA) encouraged people to use Hurricane Irene SMS and social networks to keep in touch with family and friends instead of calling by phone so as not to jam networks. Twitter users in New York City and other locations saw tweets about the earthquake, 2011 Virginia Earthquake which originated in Virginia state, up to 30 seconds before it was felt, showing that information moves faster through networks than the earthquakes themselves.

2.1.3 Past Research on Social Media in Disaster

Since the use social media is an extremely recent phenomenon, the areas of study most closely related to it such as computer science and communications are the fields in which with topic has been most dealt with. These areas commonly use the terms Information and Communication Technologies (ICT) and Crisis-informatics. Amongst the published research related to social media and disaster, two approaches have been taken: one, analysis of events in past disasters, in which social media was utilized, and, two, proposals and recommendations for system design or architecture to improve existing examples of social media use in disasters. Some of the research of events in past disasters aim to gather data to understand its role in the disaster. Potts (2009) analyzes strengths and limitation of systems of the use of what she calls

as “social software” in disaster, natural and man made over the past several years in order to point out important points to researchers and practitioners. In order to trace participant activities she uses Actor Network Theory (ANT), which she says suggests “that we should examine human and non-human actors as equal agents of action, giving neither of them priority” (p. 1). The events that she analyzes are The 2004 Indian Ocean earthquake and tsunami, the 2005 London bombings, 2005 Hurricane Katrina, the 2007 Virginia Tech shootings, and finally the 2008 Mumbai attacks. She points out that, in earlier disasters, connections via channels provided by the mainstream media were utilized, but in more recent disasters participants use various systems to locate data, validate it as information and share it as knowledge. She points out that current systems are ones that are being used in unintended ways therefore, she asks researchers to participate in the process of collecting and verifying data in these disasters so that they can better understand the constraints of the systems they are researching. Adding that creating new systems that have single uses is not a preferable as creating technologies that unite content so it can be validated, speaking in the context of designing systems. Tseng, Chen and Chi (2011) preformed research seeking to understand public participation online in a disaster event, 2009 Typhoon Morakot, which struck Taiwan, through a survey that consisted of reviewing electronic documents online after the disaster and interviewing eight major microbloggers who use that social networks Twitter and Plurk. The researchers identify two major functions of social media, described as “information dissemination and resources mobilization” and come to three conclusions, which are that the social networks allowed for public participation and engagement, users of the social networks played vital roles as collaborators in the disaster and that social media utilization is more efficient disaster backchannel communication mechanism and thus demonstrates that possible governmental and public participation in disaster. Qu, Huang, Zhang, and Zhang (2011) undertook a study of microblogging on the Sina- Webibo social network, a Twitter clone, after the 2010 Yushu earthquake in China using multiple analysis methods to understand the roles played by microblogging systems in disasters as so to better harness that power to enable better disaster response as well as understand how Chinese users, having a non-western socio-cultural system, in particular use microblogging after disaster. The analysis consisted of the content of the status updates of the microblogging service, trends of topics, and the information spreading process. The analysis was carried out based on data related to the earthquake taken over a 48-day period immediately following the earthquake, in total 94,101 microblog posts and 41,817 reposts. The study found four major purposes for sharing data during the disaster: situation update, opinion expression, emotional support, and calling for action. It also found that different types of messages trend and follow different patterns and at different times during the disaster, in particular, calls for action were spread easier. Updates of the situation were also found to spread faster, although a large amount of them were second-hand messages, so the authors stressed the need to have systems to distinguish between first-hand and second-hand information as well as authorized information. Starbird, Palen, Hughes, and Vieweg (2010) studied the use of Twitter in the 2009 Red River Valley flooding in the US and Canada, identifying mechanisms of information production, distribution, and organization. The aim of their research is to theorize about the type of communication that they identify Twitter as, computer mediated communication (CMC), and to describe the relationship of it to disaster events as to predict the potential of the tool so that it can be used to mitigate disaster damage. The data used in the research was acquired over a 51 day period surrounding the flooding by using the Twitter Search API to save the terms “red river” and “redriver.” This produced 13.153 tweets and 4,983 tweets from unique authors. The second part of the analysis involved collecting the data streams from each of those unique users through a script developed by the authors. The data was then coded and analyzed using special software designed to be used with CMC data sets called E-Data

Viewer to find the locations of the tweets and the affiliation of the Twitter users. The study finds that official information remains important and that it is complimented, not usurped, by information shared by the public. They state that the lesson of the research is not that agencies need to be everywhere online but know that people have the capacity to re-use data for their local needs. In an related research, Vieweg, Hughes, Starbird, and Palen (2011) analyze data from two concurrent disaster events, the previously mentioned 2009 Red River Valley Floods and the 2009 Oklahoma Grassfires to understand what contribution Twitter makes to situational awareness. The method for data collection is the same as the one previously described over the 51 day period, only also having been also done with the terms “Oklahoma,” “okfire,” “grass fire,” and “grassfire.” The analysis of the two events identifies features of information generated during emergencies and helps to create a framework for designing and implementing information extraction by software systems so that they systems can be used by the public and emergency responders to improve situational awareness. There is one example of research from the 2011 East Japan Earthquake and Tsunami. Acar and Murakami (2011) undertook a study on the use of Twitter in the tsunami disaster. The study was conducted two weeks after the earthquake, using open-ended questionnaires sent to a randomly selected group of twitter users as well as analysis performed on tweets form the disaster-hit areas. They found that directly affected individuals’ status updates tented to be related to “their unsafe and uncertain situation” (p. 392) and that individuals in remote areas used Twitter to let others know that they are safe. They also found that status updates in the disaster hit areas were warnings, help requests, and reports about the situation and that official local authority Twitter accounts, which were set up after the disaster, were particularly useful because the information was well followed and retweeted extensively, especially in the case of tsunami warnings. The research also found that there were many retweets that were inaccurate information as well as that there were many calls for help that were lies or misplaced. In addition, users were concerned by these types of misinformation, but there were not enough status updates from government and mass media to combat the misinformation. The researchers make recommendations for use of social media in disaster that say users should have more responsibility for their tweets, that users should realize it is a public method of communication, and that information sources should be clear in updates. They go on to say that more research needs to be done to see if official hashtags keywords could help improve the use of Twitter in disasters (Acar and Murakami, 2011). This thesis’ research is similar to the research here due to the topic of the 2011 East Japan Earthquake and Tsunami, yet there are many differences in terms of scope and methodology of the research. Acar and Murakami’s research uses data taken from Twitter and analyzes it to find patterns of use, in the process discovering problems with unreliable tweets and makes recommendations about how to minimize those problems. The research does not focus specifically on one social media service, such as Twitter, as Acar and Murakami’s does. It uses examples of all kinds of social media use in the disaster. It also does not analyze a large amount of data collected from a specific source. In addition, this thesis aims to make policy recommendations for organizations and government bodies. The study of social media use in disasters from a disaster management perspective is an area in which little research has been undertaken. Research aimed at creating policy recommendations for organizations and governmental bodies is also an area that has not been delved into.

2.2 DISASTER, COMMUNICATIONS AND SOCIAL MEDIA IN JAPAN

2.2.1 Disaster in Japan

2.2.1.1 Background

Japan is a country with a long history of natural disaster and is rated second in the world in terms of natural hazard risk (Maplecroft, 2011). From 1980 to 2010, there were 157 natural disasters, with 8,568 people killed, 3,361,979 people affected, and over USD $200 billion in damage (See Figure 2-5). In terms of human exposure to disaster, Japan ranks first in the world for cyclones and earthquakes and second for tsunami (See Figure 2-6). It also ranks eighth in the world for land slides (2009 Global Assessment Report on Disaster Risk Reduction).

Figure 2-5: Overview of natural disasters in Japan from 1980 to 2010 Source: 2009 Global Assessment Report on Disaster Risk Reduction http://www.preventionweb.net/english/countries/statistics/risk.php?iso=jpn

Figure 2-6: Human exposure to hazard zones in Japan Source: 2009 Global Assessment Report on Disaster Risk Reduction http://www.preventionweb.net/english/countries/statistics/risk.php?iso=jpn

Figure 2-7: Map preview of Global Risk Data Platform for Japan showing areas of risk throughout Japan Source: 2009 Global Assessment Report on Disaster Risk Reduction http://www.preventionweb.net/english/countries/statistics/risk.php?iso=jpn Although Japan is at high risk for disasters, it also has a high capacity to combat the effects of major disaster due to its high level of socio-economic resilience to withstand their effects compared with other countries (Maplecroft, 2011). This socio-economic resilience includes aspects such as economic strength, strong governance, well established infrastructures, disaster preparedness and strong building regulations (Maplecroft, 2011). Due to this, even though Japan’s ranks second in terms of natural hazard risk, Japan’s overall risk is lower. Its risk, calculated by the probability of disasters and negative effects of these possible disasters, in terms of mortality and economic loss, make Japan 7th out of 89th for risk associated with cyclones, 6th out of 76th for risk associated with tsunami, and 23rd out of 153rd for risk associated with earthquakes (See Figure 2-8).

Figure 2-8: Japan’s risk profile Source: UNISDR http://www.unisdr.org/partners/countries/jpn Japan is also signatory to the Hyogo Framework for Action, which is focused on building resilience of nations and communities to disaster and is under operation from 2005 to 2015 (UNSDIR, 2005). The Hyogo Framework for Action, adopted by 190 member states, is the first plan that attempts to detail required actions for reducing disaster impacts from various sectors and actors and has five priority actions to do so (UNSDIR, 2011). The World Conference on Disaster Reduction, at which the Hyogo Framework for Action was adopted, was held in Kobe, Hyogo Prefecture, Japan almost 10 years to the day after the 2005 Great Hanshin earthquake that occurred in the same area in which 6,434 people lost their lives. For this reason, the Hyogo Framework for Action has a great symbolism in Japan and for the Japanese people.

2.2.1.2 The 1995 Great Hanshin Earthquake

One of the largest disasters in Japan in recent history was the 1995 Great Hanshin earthquake, which occurred on January 17, 1995, at 05:46 JST (Fukushima, 1995). The earthquake registered 6.8 magnitude (USGS, 2010) and a 7 on the Japanese Japan Meteorological Agency (JMA) scale (City of Kobe, 2009). The epicenter of the earthquake was in Awaji Island (City of Kobe, 2009), but the dense metropolitan city of Kobe, 20 kilometers away, was highly impacted, with 4,571 out of 6,434 of the deaths from the resulting from the earthquake being there (City of Kobe, 2009). The total financial damage from the earthquake was over USD $100 billion (Kimura, 2011).

Over 250,000 buildings were destroyed or partially destroyed (Kimura, 2011). In addition a one kilometer stretch of the Hanshin Expressway Route 43 collapsed. The area which was affected by the disaster had a population around 3.6 million. After the disaster 30% of residents were in temporary housing, with 47% of them being over 65 years old, while the ratio of the population over 65 was 15% (Kimura, 2011).

Subsequent fires also created much damage throughout the area. In Nagata ward, chemicals used in first floor shoe factories, for which the area was known for, helped spur on the fires. Relief from the government was slow due to roads being impassable because of damage. In addition aftershocks lasted for several days after the initial earthquake.

Following the disaster, more than 1 million people came to volunteer to help after the disaster (Imada, 2003). This event is considered to be the emergence of volunteerism in Japan. The Japanese government made January 17 a national "Disaster Prevention and Volunteerism Day," and the week following it "Disaster Prevention and Volunteerism Week" to remember the events and encourage volunteerism in Japan (Government of Japan Cabinet Office, 1995).

2.2.2 The Japanese Communication Landscape

2.2.2.1 Overview of Usage of the Internet, Mobile phones and Social Media in Japan

Japan, as a developed nation, has a high amount of communications infrastructure and is known around the world for being technologically advanced in the areas of communication for its products. Japan has a high amount of traditional forms of mass communication, such as televisions, radios and telephones. Mobile phone usage in Japan is considerably high. According to the Japanese Ministry of Internal Affairs and Communications, as of December 2007, 105 million contracts for mobile phones and PHS (Personal Handy Phone) and 16 million IP phones (2008). By household, this comes to 96% of households having cellular phones (MIC, 2011) and 88.5% of individuals (Kamino, 2009). Japan has four mobile phone service providers NTT Docomo, KDDI AU, Softabank and Willcom. Nearly all mobile phones in Japan are mobile Internet capable and many of these users access mobile Internet with the devices.

Personal computer use and use of the Internet with computers is also very common. As of 2008, 86% of households owning one and the rate of Internet use by household as of 2008 is 91%, up from 34% in the year 2000 (MIC, 2011).

With such high rates of mobile phone, mobile Internet, computer and Internet use, services accessed through the Internet, such as social media is also very high. Japan has several popular social networks, all which slightly differ from one another. The most popular social networking site is called Mixi. It is followed by Twitter and then Facebook. There are also mobile phone game based social network sites, such as Gree and Mobage, which are extremely popular.

Mixi began in 2004 and as of January 2011, has 21 million users (Tabuchi, 2011). Usage of Twitter has grown drastically since 2009, after the Twitter-fueled post-election demonstrations in Iran (“Japanese Twitter Users,” 2009). As of March 2010, Japan has the largest number of users on Twitter of any country in the world (Jenssen, 2010) and as of August of that year Twitter had 16 million users in Japan (Russell, 2010), with the number having since grown. Japanese users tweet (posting short status update messages) at double the rate of users from other countries, making up 13% of the content on Twitter (Jenssen, 2010). In contrast, Facebook, the most popular social network in the world, as of January 2011 only had 2 million users in Japan (Preston, 2011). Likely related to the increased use of social networking in during and after the 2011 East Japan Earthquake and Tsunami, the number has grown to 5 million (“Japan Facebook Statistics,” 2011).

Social media use in Japan is growing, but there are barriers to its growth due to cultural differences. Many Japanese users feel uncomfortable having their identity online (Tabuchi, 2011). Surveys showed that almost 90% of social media users in Japan are reluctant to use their real names and that nearly half of users to not have connections to people that they know personally in social networking sites (Tabuchi, 2011).

2.2.2.2 Disaster Communication/Information systems

Since natural disasters are commonplace in Japan several systems have been developed for disaster situations. Some systems are used to notify of impending disasters and others are to be used after disasters to confirm the safety of family members or friends. Bousai musen, or “disaster prevention wireless,” is a system of using speaker announcements to notify of danger from disaster. The system operation differs based on range, from local to national. The system is currently being upgraded to work digitally on a national scale using satellites for instant notification. The new system is called J-ALERT (See Figure 2-9) (JMIAC, 2007). The new J-ALERT system, in the same way as Bousai musen, can send out notifications of impending tsunami or actionable information related to a disaster after it has occurred, but the J-ALERT system can be used nationally, as opposed to Bousai musen.

Figure 2-9: Image of the J-ALERT SYSTEM Source: Ministry of Internal Affairs and Communication http://www.fdma.go.jp/neuter/topics/houdou/190110- 2/190110houdou.pdf The Earthquake Early Warning (EEW) system is designed to send automatic alerts to television and radio, as well as other devices such as mobile phones, several seconds before an earthquake occurs by detecting p-wave preliminary tremors to allow people to brace for

shaking or to stop moving vehicles. The system is enacted and monitored by the Japan Meteorological Agency.

When one or more seismometers detects p-waves in an area, the information is automatically sent to the Japan Meteorological Agency and then from there sent to advanced users, such as mobile phone providers and television and radio stations. When EEW signals are sent to radio and television stations, they are subsequently automatically sent to out to televisions and radios, triggering an Emergency Warning Broadcast (EWB). The Emergency Warning Broadcast is recognizable by two chimes and then an announced or written message saying that an earthquake will occur and to prepare. Three of the main mobile phone carriers, NTT Docomo, KDDI AU, and Softbank, provide EEW notifications to users via their networks, although not all of the phones are capable of receiving the alerts. Smartphones such as the iPhone can receive the EEW notifications through installable applications from third party providers. Other devices such as radios are also available that can receive the notifications.

Another system used in disaster situations is a call-in message service for confirming the safety of family and friends, called saigai dengon daiyaru in Japanese (See Figure 2-10). In order to use the system, individuals call in through landline phones or mobile phones to the number 171 and register their home phone number or any other landline number that others might know and leave a 30 second message to their family members or friend. Then an individual looking for that person can call into the service on a landline or a mobile phone and enter the phone number of the person they are trying to contact and, if that person has registered and left a message, the person can listen to the message (“Anpi kakunin,” 2011).

The service also has an online component that can be accessed by computer or mobile phone. Each of the major mobile phone companies provide and address to access the system. The system essentially works in the same manner except for two differences: mobile phone numbers may also be registered and voice messages cannot be left so the user enters a short up to 100 character message (“Anpi kakunin,” 2011).

Figure 2-10: Instructional image of the use of the saigai dengon dayaru Source: Nikkei Keizai Shimbun http://www.nikkei.com/news/headline/article/g=96958A9C93819499E3E3E2E28B8DE3E3E2E1E0E2E3E3E2E2E2E 2E2E2 References

2009 Global Assessment Report on Disaster Risk Reduction (GAR). (2009, May 17) Japan – Risk Profile. Retrieved from http://www.preventionweb.net/english/countries/statistics/risk.php?iso=jpn

Abram, C. (2006, September 26). Welcome to Facebook, everyone. The Facebook Blog. Retrieved from: http://blog.facebook.com/blog.php?post=2210227130

Acar, A. and Murakami, A. (2011). Twitter for crisis communication: lessons learned from Japan's tsunami disaster. International Journal of Web Based Communities. 7 (3). 392-402.

Anpi kakunin no dengon daiyaru, tsugi no houhou de. (2011, March 13). Nikkei Keizai Shimbun. Retrieved from http://www.nikkei.com/news/headline/article/g=96958A9C93819499E3E3E2E28B8 DE3E3E2E1E0E2E3E3E2E2E2E2E2E2

Aoki, E., Kikuchi, T., Korida, K., Yoshiyama, N., Shibata, Y., Takahashi, M., & Takenaka, M. (2011). 2011 International Conference on Complex, Intelligent and Software Intensive Systems (CISIS). 691-696.

Authorities learn to ‘tweet’ in disasters. (2011, March 30). Computer World. Retrieved from http://www.computerworld.com.au/article/381497/authorities_learn_tweet_disasters/ #closeme

Barnes, A. & Carrell, S. (2005, July 10). How 3G phone technology created instant history. The Independent. Retrieved from: http://www.independent.co.uk/news/uk/crime/how-3g-phone-technology-created- instant-history-498279.html

Boyd, D., Ellison, N. B. (2007, December 17). Social Network Sites: Definition, History, and Scholarship. Journal of Computer-Mediated Communication. 13 (1). 210-230. Retrieved from: http://onlinelibrary.wiley.com/doi/10.1111/j.1083- 6101.2007.00393.x/full

Cell Phones Capture London Blasts. (2005, July 8). Wired. Associated Press. Retrieved from http://www.wired.com/science/discoveries/news/2005/07/68143

Cell phones ubiquitous globally, but 1 in 8 now have mobile Internet. (2011, September 15). The Christian Science Monitor. Retrieved from: http://www.csmonitor.com/Business/Latest-News-Wires/2011/0915/Cell-phones- ubiquitous-globally-but-1-in-8-now-have-mobile-Internet

City of Kobe (2009, January 1). The Great Hanshin-Awaji Earthquake: Statistics and Restoration Progress [PDF]. Retrieved from http://www.city.kobe.lg.jp/safety/hanshinawaji/revival/promote/img/january.2009.pdf

Collins, S. (2011, April 5). Using social media in times of crisis. Social Media Today. Retrieved from http://socialmediatoday.com/stephencollins/283892/using-social- media-times-crisis

Cowell, A. (2005, July 7). After Coordinated Bombs, London Is Stunned, Bloodied and Stoic. The New York Times. Retrieved from http://www.nytimes.com/2005/07/07/international/europe/07cnd-explosion.html

Disasters Emergency Committee. Haiti Earthquake Facts and Figures. Retrieved October 21, 2011 from http://www.dec.org.uk/haiti-earthquake-facts-and-figures

Facebook. (2011). Statistics. Retrieved October 19, 2011. Retrieved from: http://www.facebook.com/press/info.php?statistics

Flaccus, G. (2007, October 24). 1,500 homes lost; $1B loss in San Diego area. The Seattle Times. Retrieved from http://seattletimes.nwsource.com/html/nationworld/2003971082_wildfires24.html

Fukushima, G. S. (1995, March). The Great Hanshin Earthquake. Japan Policy Research Institute. Retrieved from http://www.jpri.org/publications/occasionalpapers/op2.html

Ford, R. (2011, August 23). Earthquake: Twitter Users Learned of Tremors Seconds Before Feeling Them. The Hollywood Reporter. Retrieved at http://www.hollywoodreporter.com/news/earthquake-twitter-users-learned-tremors- 226481

Gibbs, N. (2007, October 23). San Diego Fire Updates. Retrieved from http://www.nathangibbs.com/2007/10/23/san-diego-fire-updates/

Giridharadas, A. (2011, March 13). Africa’s Gift to Silicon Valley: How to track a Crisis. The New York Times. Retrieved from: http://www.nytimes.com/2010/03/14/weekinreview/14giridharadas.html

Government of Japan Cabinet Office. (1995, December 15). Bosai to borantia no hi oyobi bosai to borantia shuukan nitsuite. Retrieved from http://www.bousai.go.jp/volunteer/html/detail_kakugi.html

Grossman, L. (2009, June 17). Iran Protests: Twitter, the Medium of the Movement. Time. Retrieved from: http://www.time.com/time/world/article/0,8599,1905125,00.html

Hodge, N. (2011, January 22). Military may Crowdsource Haiti Relief. Wired. Retrieved from http://www.wired.com/dangerroom/2010/01/military-may-crowdsource-haiti-relief/

Huang, C., Ou, Y., Zhang, Jun & Zhang, P. (2011). Microblogging after a major disaster in China: a case study of the 2010 Yushu earthquake. In Proceedings of the ACM 2011 conference on Computer supported cooperative work (CSCW '11). ACM, New York, NY, USA, 25-34. DOI=10.1145/1958824.1958830 http://doi.acm.org/10.1145/1958824.1958830

Huang, C. (2011) Facebook and Twitter key to Arab Spring. The National. Retrieved October 11, 2011 from: http://www.thenational.ae/news/uae-news/facebook-and-twitter-key- to-arab-spring-uprisings-report

Huck, P. (2005, July 11). ‘We had 50 images within an hour’. The Guardian. Retrieved from: http://www.guardian.co.uk/technology/2005/jul/11/mondaymediasection.attackonlon don

Huiji, G., Barbier, G. & Goolsby, R. (2011). Harnessing the Crowdsourcing Power of Social Media for Disaster Relief. : Intelligent Systems, IEEE. 26 (3). 10-14.

Imada, M. (2003). The voluntary response to the Hanshin Awaji earthquake: a trigger for the development of the voluntary and non-profit sector in Japan. In Osborne (Ed.), The voluntary and non-profit sector in Japan: the challenge of change.(pp. 40-51). Ed. London: RoutledgeCurzon. Print.

Indonesians beat slow disaster relief by tweeting. (2010, November 11). Reuters. Retrieved from http://www.reuters.com/article/2010/11/22/us-indonesia-volcano-twitter- idUSTRE6AL1Q820101122

International Telecommunication Union. (2010, October 19). ITU estimates two billion people online by end 2010. Retrieved from http://www.itu.int/net/pressoffice/press_releases/2010/39.aspx

International Telecommunication Union. (2011). Measuring the Information Society. Geneva, Switzerland. Retrieved from: http://www.itu.int/ITU- D/ict/publications/idi/2011/index.html

Iran’s Twitter revolution [Editorial]. (2009, June 16). The Washington Times. Retrieved from: http://www.washingtontimes.com/news/2009/jun/16/irans-twitter-revolution/

Irene Update 20: Make a plan to keep in touch. (2011, August 27). Federal Emergency Management Association. FEMA Blog. Retrieved at http://blog.fema.gov/2011/08/irene-update-20-make-plan-to-keep-in.html

Japanese Ministry of Internal Affairs and Communication (JMIAC). (2007, January 10). Zenkoku shunji keihou shisutemu (J-ALERT) ni yoru ichibu no jyouhou no hasshin kaishi. Retrieved from http://www.fdma.go.jp/neuter/topics/houdou/190110- 2/190110houdou.pdf

Japanese Twitter Users. (2009, July 7). Shibuya 246. Retrieved from http://shibuya246.com/2009/07/07/japan-twitter-users-gachapin/

Japan Facebook Statistics. Socialbakers. Retrieved November 4th, 2011 from http://www.socialbakers.com/facebook-statistics/japan

Jenssen, C. L. (2010, March 29). Event Report: Green Leaders Forum #2: Green Media & Web2.eco. Web2.eco. (YouTube video). Retrieved from http://equalc.com/events/japan/107-glf2-event-report-photos-video

Kamino, E. (2009, August 31). Keitaidenwa, PHS no jinkou fukyuu ritsu, 88.5% ni – soumusho matome. CNET Japan. Retrieved from http://japan.cnet.com/news/com/20399136/

Kaplan, A., & Haenlein, M. (2010, January-February). Users of the world, unite! The challenges and opportunities of Social Media, Business Horizons, 53 (1). 59-68. Retrieved from: http://www.sciencedirect.com/science/article/pii/S0007681309001232

Kaste, M. (2011, June 29). Facebook’s Newest Challenger: Google Plus. National Public Radio. Retrieved from: http://www.npr.org/2011/06/29/137507567/facebooks- newest-challenger-google-plus

Kietzmann, J. H., Hermkens, K, McCarthy, I. P. Silvestre, B. S. (2011, May-June) Social media? Get serious! Understanding the functional building blocks of social media, Business Horizons, 54 (3), 241-251. Retrieved from: http://www.sciencedirect.com/science/article/pii/S0007681311000061

Kimura, M. (2011, January 12). Recovery Efforts after the Great Hanshin-Awaji Earthquake [Presentation]. Superintendent of Emergency Management Hyogo Prefectural Government.

Kuhagen, S. (2011, July 27). Wikipedia? In a judicial opinion? (Pt. II). Retrieved from: http://scottkuhagen.com/2011/07/27/wikipedia-in-a-judicial-opinion-pt-ii/

Lang, G. and Benbunan-Fich, R. (2010). The Use of Social Media in Disaster Situations: Framework and Cases. International Journal of Information Systems for Crisis Response Management (IJISCRAM). 2(1), 11-23

Liu, S. B., Palen, L., Sutton, J., Hughes, A., & Vieweg, S. (2008). In Search of the Bigger Picture: The Emergent Role of On-Line Photo Sharing in Times of Disaster. Proceedings of the 5th International Community on Information Systems for Crisis and Response Management Conference (ISCRAM 2008).

Mashable. Twitter. Retrieved Oct 19, 2011from http://mashable.com/category/twitter/

Massa, P. (2010, September 24). History Unfolding [Video file]. Video posted to http://vimeo.com/15247745

Maplecroft. (2011, August 11). China, India, Philippines and Indonesia have highest economic risk from natural disasters. Retrieved from http://maplecroft.com/about/news/natural_hazards_2011.html

Mayfield, A. (2008, January 8) What is social media? iCrossing. Retrieved from: http://www.icrossing.co.uk/fileadmin/uploads/eBooks/What_is_Social_Media_iCross ing_ebook.pdf

Ministry of Internal Affairs and Communications (MIC). (2008, February 28). Status of Numbers of Subscribers to Telecommunications services. Retrieved from http://www.soumu.go.jp/joho_tsusin/eng/Statistics/pdf/080227_1.pdf

Ministry of Internal Affairs and Communications (MIC) (2011). Chapter 11 Information and Communication / Science and Technology. Japan Statistical Yearbook 2011. Retrieved from http://www.stat.go.jp/english/data/nenkan/1431-11.htm

MobiThinking. (2011, July). Global mobile statistics 2011. Retrieved on Oct. 19, 2011 from http://mobithinking.com/mobile-marketing-tools/latest-mobile-stats

Morales, X. Y. Z. G. (2010, April 26). Networks to the Rescue: Tweeting Relief and Aid During Typhoon Ondoy. (Unpublished master’s thesis). Georgetown University, Washington D.C.

New Zealand Police (2011, September 8th). List of deceased. Retrieved from http://www.police.govt.nz/list-deceased

O’Reilly, T. (2005, September 30). What is Web 2.0. Oreilly.com. Retrieved from http://oreilly.com/web2/archive/what-is-web-20.html

Palen, L. (2008, November 8). Online Social Media in Crisis Events. Educause Quarterly. 31 (3). 76-78

PBS (2011, May 13) Crisis Mapping [Television series episode]. Need to Know. Public Broadcasting Station. Retrieved from http://video.pbs.org/video/1923247005#

Pierce, David. (2010, January 21). Pentagon’s Social Network Becomes Hub for Haiti Relief. Wired. Retrieved from http://www.wired.com/dangerroom/2010/01/pentagons-social- network-becomes-hub-for-haiti-relief/

Poulsen, K. (2007, October 23). Firsthand Reports From California Wildfires Pour Through Twitter. Wired. Retrieved from http://www.wired.com/threatlevel/2007/10/firsthand- repor/

Potts, L. (2009, 19 July). Peering into Disaster: Social Software Use from the Indian Ocean Earthquake to the Mumbai Bombings. Professional Communication Conference, 2009. IPCC 2009. IEEE International. 1-8.

Preston, J. (2011, March 13). After Quake and Tsunami, Japanese Citizens Flock to Social Networks for Information. The New York Times. Retrieved from http://mediadecoder.blogs.nytimes.com/2011/03/13/after-quake-and-tsunami- japanese-citizens-flock-to-social-networks-for-information/?ref=media

Qu, Y., Huang, C., Zhang, P. & Zhang, J. (2011). Microblogging after a Major Disaster in China : A Case Study of the 2010 Yushu Earthquake. Proceedings of the ACM 2011 conference on Computer supported cooperative work (CSCW 2011). 25-34

Qu, Y., Huang, C., Zhang, P. & Zhang, J. (2011). Harnessing Social Media in Response to Major Disasters.

Redmon, A. (2010, November 9). Department of Defense’s APAN Community Wins Acclaimed Industry Award for Successful Social Technology. Retrieved from http://telligent.com/company/news/b/press_releases/archive/2010/11/09/department- of-defense-s-apan-community-wins-acclaimed-industry-award-for-successful-social- technology.aspx

Rogstadius, J., Kostakos, V., Larado, J., & Vukovic, M. (2011). Towards Real-time Emergency Response using Crowd Supported Analysis of Social Media. Proceedings of the 29th international conference on Human factors in computing systems (CHI 2011) .

Russell, J. (2010, August 23). Japan overtakes Indonesia as biggest Twitter user in Asia. Asian Correspondent. Retrieved from http://asiancorrespondent.com/39290/japan- overtakes-indonesia-as-biggest-twitter-user-in-asia/

Santos, K. (2010, November 15). Philippines: typhoon warnings shared on Twitter. Global Information Network. Retrieved from: http://search.proquest.com/docview/769208542

Shaikh, A. (2010, November 22). Tweeting Mount Merapi: A new Kind of Disaster Response. United Nations Dispatch (web log post). Retrieved from http://www.undispatch.com/twitter-mount-mirapi

Shiels, M. (2011, March. 28) Twitter co-founder Jack Dorsey rejoins company. BBC News. Retrieved from: http://www.bbc.co.uk/news/business-12889048

Starbird, K., Palen, L., Hughes, A. L. & Vieweg, S. (2010). Chatter on the red: what hazards threat reveals about the social life of microblogged information. CSCW '10 Proceedings of the 2010 ACM conference on Computer supported cooperative work. 241-250.

Sutton, J., Palen, L., & Shklovski, I. (2008). Backchannels on the Front Lines: Emergent Uses of Social Media in the 2007 Southern California Wildfires. Proceedings of the 5th International Community on Information Systems for Crisis and Response Management Conference (ISCRAM 2008).

Svetlik, J. (2011, October 14.). Google Plus has 40m users. CNET UK. Retrieved from: http://crave.cnet.co.uk/software/google-plus-has-40m-users-50005633/

Tabuchi, H. (2011, January 9). Facebook Wins Relatively Few Friends in Japan. The New York Times. Retrieved from http://www.nytimes.com/2011/01/10/technology/10facebook.html?_r=1&scp=1&sq= mixi&st=cse

United Nations International Strategy for Disaster Reduction (UNSDIR). (2005, February 10). Hyogo Framework for Action 2005-2015*: Building the Resilience of Nations and Communities to Disasters.

United Nations International Strategy for Disaster Reduction (UNSDIR). Hyogo Framework for Action. Retrieved November 6, 2011 from http://www.unisdr.org/we/coordinate/hfa

United Nations International Strategy for Disaster Reduction (UNSDIR). Japan. Retrieved November 6, 2011 from http://www.unisdr.org/partners/countries/jpn.

United States Geological Survey (USGS). (2010, January 5). Significant Earthquakes of the World: 1995. Retrieved from http://earthquake.usgs.gov/earthquakes/eqarchives/significant/sig_1995.php

Vieweg, S., Hughes, A. L., Starbird, K., & Palen, L. (2010). Microblogging during two natural hazards events: what twitter may contribute to situational awareness. Proceedings of the 28th international conference on Human factors in computing systems (CHI 2010). 1079-1088.

Chapter 3 Overview of the 2011 East Japan Earthquake and Tsunami and the Use of Social Media in the Disaster

3.1 OVERVIEW OF THE DAMAGE

On March 11th, 2011, a 9.0 magnitude earthquake occurred off the eastern coast of Japan at 2:46 p.m. Japanese Standard Time (JST). The earthquake occurred 130 kilometers to the east at a depth of 24 kilometers. On the Japanese scale for measuring earthquakes, the JMA Seismic Intensity Scale, which is from one to seven, the 2011 East Japan earthquake recorded an intensity of 7 in the city of Kurihara in Miyagi prefecture (Figure 3-1) and a level of 6+ in 28 other cities in Japan. The earthquake is the largest ever recorded in Japan. The earthquake has been named “The 2011 off the Pacific coast of Tohoku Earthquake” by the Japan Meteorological Agency (“The 2011 off,” 2011).

Figure 3-1: Map of affected area Source: USGS

Table 3-1: Summary of the 2011 East Japan Earthquake Date and Time 11 March 2011 14:46 JST (05:46 UTC) Magnitude 9.0 (international scale) (interim value; the largest earthquake recorded in Japan) Hypocenter N38.1, E142.9 (130km ESE off Oshika Peninsula) Depth 24km (interim value) Reverse fault type with WNW-ESE compressional axis Mechanism (by CMT analysis) Earthquake Issued 8.6 seconds after the detection of the first P-wave at the nearest seismic Early station

Warning 7 (Max) Kurihara City of Miyagi Prefecture 28 cities and towns (including Wakuya Town, Tome City, JMA Seismic 6+ Osaki City, Natori City) in Miyagi, Fukushima, Ibaraki, Intensity and Tochigi Prefectures 6- or weaker Observed nationwide from Hokkaido to Kyushu Source: The Japan Meteorological Agency (JMA) After the earthquake several aftershocks were recorded. Within 2 hours, 3 aftershocks with a level if 7 on the JMA seismic scale were recorded in other surrounding prefectures. There were a total of 49 aftershocks with a magnitude greater than 6 on the JMA seismic scale (Figure 3-2) in surrounding areas. The largest aftershock was in Ibaraki Prefecture at 3:15 p.m. JST (“Mega Disaster,” 2011).

Figure 3-2: Location of the main shock and aftershocks Source: JMA, (2011)

Immediately after the earthquake, and prior to the tsunami, a tsunami warning was issued at 2:46 p.m. JST. At 3:31 p.m., a major tsunami struck the eastern coast line of northern Japan at a levels of 3 to 10 meters in many places. The highest recorded height was 38.9 meters at Miyako City in Iwate Prefecture (“Daishinsai,” 2011). Tsunami warnings were also issued in other countries on the Pacific ocean. (Parashar, et al. 2011).

Figure 3-3: The tsunami sweeps through Natori, Miyagi Prefecture Source: CBC News http://www.cbc.ca/news/world/story/2011/03/11/japan-quake-tsunami.html

Figure 3-4: The tsunami striking Iwanuma, Miyagi Prefecture Source: CBC News http://www.cbc.ca/news/world/story/2011/03/11/japan-quake-tsunami.html

The earthquake itself resulted in about 230 deaths and from the subsequent tsunami there are, as of September 30th, 2011, 15,815 dead and 3,966 missing (Figure 3-5) (Veraeck and Daniell, 2011). The deaths are spread out over 18 prefectures (“Damage Situation,” 2011). There were around 6,000 injuries related to the tsunami. More than 66% of the deaths (Figure 3-6) were of persons over the age of 60 (Veraeck and Daniell, 2011). This might be due to the fact that the area has a large aged population because of a general aging and decline in population and an exodus of younger people for economic and educational opportunities. The figure could also reflect the fact that older individuals have a harder time of evacuating during disaster.

Figure 3-5: Dead and missing in the disaster Source: Earthquake-Report http://earthquake-report.com/2011/10/02/japan-tohoku-earthquake-and-tsunami-catdat- 41-report-october-2-2011/

Figure 3-6: Deaths by age in the disaster Source: Earthquake-Report http://earthquake-report.com/2011/10/02/japan-tohoku-earthquake-and-tsunami-catdat-41-report-october-2-2011/

After the disaster millions of people lacked food, water and electricity. 310,00 people were living in emergency shelter or with relatives and 24,000 were stranded in temporary shelter such as automobiles. In addition, 1.5 million households were without water and 1.9 households were without electricity (“Millions Of Stricken,” 2011).

The tsunami caused an unfathomable amount of damage to property. As of September 30th, 2011, over 900,000 buildings had suffered damage related to the earthquake and the tsunami (Veraeck and Daniell, 2011). There was serious damage to infrastructure, transportation systems, electrical service, fuel supply chains, and communication services. The socio- economic loss from the disaster is estimated at US $100 billion to $500 billion (Veraeck and Daniell, 2011). The then prime minister, Naoto Kan, said that the disaster was the Japan’s most difficult crisis since the end of World War II (“Anxiety in Japan,” 2011).

The earthquake caused 11 nuclear power reactors to shut down. (“Timeline: Japan,” 2011). 46 minutes later, the tsunami hit the Fukushima Daiichi Nuclear Power Plant and Fukushima Daini Nuclear Power Plant (Fackler, 2011). A 15 meter wave from the tsunami overtopped the 5 meter sea walls designed to protect the power plant, causing damage (Johnston, 2011). The Fukushima Daiichi Nuclear Power Plant has 6 reactors or units. At the time, reactors 4, 5, and 6 were shut down for routine inspections (“Timeline: Japan,” 2011). The tsunami destroyed backup diesel generators that the plant was running on at four of the six reactors The last remaining generator allowed reactors 5 and 6, which were shut down for maintenance, to maintain their cooling systems, but reactor 4 was damaged by water from the tsunami (Fackler, 2011).

Due to this the plant could not properly cool and so meltdowns occurred at reactors 1, 2 and 3. These meltdowns caused three hydrogen explosions and this led to three explosions, which leaked radioactive material into the environment (“Timeline: Japan,” 2011) The failure of the cooling systems at the power plant caused the government to declare a state of emergency at the Fukushima Daiichi Nuclear Power Plant (McCurry, 2011) On March 15th, the Prime Minister asked all people within a 20 kilometer radius of the Fukushima Daiichi Nuclear Power Plant to evacuate and for people living between a 20 to 30 kilometer radius to stay inside their homes (Figure 3-7) (“Japan’s PM,” 2011). Subsequently, 200,000 residents were evacuated from the Fukushima area (“Fukushima Nuclear Evacuees,” 2011).

Figure 3-7: Area of the evacuation zone around Fukushima Daiichi Nuclear Power Plant Source: Wikipedia http://en.wikipedia.org/wiki/File:Towns_evacuated_around_Fukushima_on_April_11th,_2011.png

Other nuclear power plants were also affected by the earthquake but were able to use backup generators to continue functioning. The Fukushima Daini Nuclear Power Plant reactors also shut down during the earthquake (“Japan Initiates,” 2011). The amount of leakage was significant enough that Japan’s Nuclear and Industrial Safety Agency said the disaster was a level 5 disaster, which the UN International Atomic Energy Agency classifies as a crisis causing “severe radiation deaths” (Derbyshire, 2011)

At the time, there was great fear about the extent of radiation leakage and the effects of the radiation. Throughout the country, and especially in Tokyo people began stocking up on goods like food and gasoline to the point that there were shortages (Idei, 2011). Access to organized and accurate information was also a problem because the situation unstable and changing by the minute (Birmingham, 2011)

3.2 THE USE OF SOCIAL MEDIA IN THE 2011 EAST JAPAN EARTHQUAKE AND TSUNAMI

When the earthquake and tsunami occurred, newly popular social media and mobile devices allowed for individuals inside and outside of the disaster to access and share information. In some cases, these individuals were directly affected persons who needed vital information to protect themselves. In other cases, these individuals were indirectly affected persons who saw the disaster, saw a need, and decided to act by offering up a service, in many cases cooperating with other people. And in yet other cases, a vast many of them, social media allowed for people to connect with others, affected or not, donate money or goods and share their moral support.

3.2.1 Overview of Social Media Use in the Disaster

The examples of social media use in the disaster can broadly be divided into the following groups:

• Social media use for personal benefit

o Directly affected individuals using social media on their own

o Indirectly affected individuals using social media on their own

• Social media use for the benefit of individuals with no personal connection

o Social media for moral support or donation

o Social media use with a group

Examples of theses types of social media uses are as follows. These are simplified, broad generalizations to give the reader a basic understanding of the types of usage that occurred. More detailed descriptions will be provided later on in the text.

• Social media use for personal benefit

o Directly affected individuals using social media on their own

In these situations, usually a person in an affected area uses a mobile phone or smartphone and, via 3G data, accesses a social network website while phone networks are jammed from calling, where through their previously established relationships he or she is able to get or send information. The individuals may have constraints such as not having electricity, so they most likely rely on mobile phones or smartphones. In many of the cases of this, individuals are attempting notify others such as their family members and friends that they are alright or find out if their family members or friends are alright. Much of this kind of use is through the social networks Twitter and Facebook. At the time of writing, Twitter very popular in Japan and Facebook tends to be used by foreign residents.

There were many articles in the news and on the Internet after the disaster that describe experiences similar to the ones previously mentioned. One such article describes how a family in the US, who had two sons in Japan, learned of one of the son’s safety:

He was able to post a Facebook message soon after the event, even though phone service was not available to make calls. This provided such relief to his worry- stricken parents. (Garret, 2011).

The author has heard similar stories from a handful of individuals.

o Indirectly affected individuals using social media on their own

In this research, indirectly affected individuals are defined as having friends or family that are possibly directly affected while the individuals themselves are not affected. In many cases, individuals send messages via social networks asking if friends or family is alright. In such cases, since they are in other locations (otherwise they would be affected themselves) they may have access to electricity, computers and reliable mobile networks, so instead of only sending out simple messages via mobile device applications, they are able to look at much more dynamic content and use multiple services, such as maps or Skype (a Voice-Over- Internet Protocol phone service) to get information.

In an example of this type of use, an individual writes that his friend, a man from the US, went missing in Japan during the disaster. He talks about how the man’s friend’s and family coordinated through Facebook and Twitter, and used blogs, people-finding websites, satellite images, and maps, to help track his friend down, eventually finding where he was and making contact with him (Shepard, 2011).

• Social media use for the benefit of individuals with no personal connection

In many cases when individuals used social media in the disaster, they are using it to get information related to their safety, but there are a vast number of people who have no personal connection yet want to give moral support to those who are directly affected or indirectly affected by the disaster, or want to “do something” to help support those people.

o Social media for moral support or donation

Twitter, Facebook, and Mixi became avenues for people to express their feelings about the disaster and their concern for the people affected by the disaster. One example of this type of expression, was how on Twitter a hashtag (a way of identifying topic keywords) #prayforjapan became instantly popular with millions of individuals worldwide, starting in the English speaking world and then being grabbed onto by people in Japan (Nishimura, 2011). The hashtag was often paired with messages asking for donations to the Red Cross via an SMS number, 90999 (See Figure 3-8), which when sent to automatically makes a USD $10 donation to the Red Cross through the billing system of the mobile service provider, raising almost USD $3 million in less than a week (Duke, 2011).

Figure 3-8: A tweet with #prayforjapan asking for a donation to the Red Cross Source: Twitter user @OfficialCharice http://www.charicemania.com/2011/03/12/announcements/prayers-for-japan- officially-missing-you-charice/ o Social media use with a group

With such dynamic and sophisticated tools at the fingertips of millions of people, it is only natural that seeing a need, many of these people will try to work together to perform some sort of aid function. During the disaster, there were many examples of groups of people, formal groups and groups of individuals, spontaneously cooperating, performing or attempting to perform some function vital to directly and indirectly affected individuals. Much of this type of aid is done by groups specifically organized to use social media to accomplish a goal or by groups of social media users who instantly began to work together to perform a task, but more and more traditional aid organizations and even governments are

beginning to recognize the advantages of social media in accomplishing their goals, therefore use of social media amongst these organizations is also increasing. These groups’ use of social media in the disaster tend to be extremely varied because the way that the groups use social media can be complex with different actors performing different functions, and the aid that the groups seek to provide can be highly specific and specialized.

One of the many examples of this type of social media use is the website Anpi Report (anpi.tv), created by a web developer, Yusuke Wada. Wada recognized that many people on Twitter were sharing people-finding information (anpi means “safety” in Japanese), but that it was hard to search for information on specific people, so he created a website where people could search for specific people using information from Twitter. Unfortunately, the information could only be input by hand one at a time, so Wada used Facebook and Twitter to look for people to help him enter the data. 250 spontaneous volunteers, connected through social media, input the tweets they found on Twitter about people’s safety into the website that Wada created. In the end over 100,000 people used the website to look for information on friends and family members (“Ima, watashitachi,” 2011). As of October 28th, 2011, the website has reports on the status of 8019 individuals.

Figure 3-9: Website of Anpi Report, Anpi.tv Source: Anpi.tv http://blog.u2plus.jp/wp-content/uploads/2011/05/anpireport.jpg

3.2.2 Coverage by the Mass Media

After the earthquake and tsunami occurred, social media was beginning to be used and the mass media quickly picked up on it. Immediately, there were videos and pictures of the earthquake and the tsunami on the Internet, in many cases from people with video-capable phones and smartphones, showing the tsunami from several locations in a way that has never happened before (Aamoth, 2011). There were also a multitude of articles, mostly about how people close to the disaster were able to let their friends and family members know that they were alright or get vital information even though the telephone lines were jammed, that

Twitter in particular had acted as a lifeline in the disaster. Many news articles touted the benefits of Twitter and Facebook, with journalist or commenters saying that they had used it to let family members know they were alright or that they personally knew someone who used it.

3.2.3 Elements Related to the Examples of Social Media Use in the Disaster

Traditional forms of mass communication are rather simple. For example, in the case of radio, a radio station transmits a signal which is picked up by listeners who listen using a radio. The same for television, and even newspapers in some sense.

The way that social media functions is complex. A simplistic explanation is that users connect to the Internet using devices that are able to access the Internet, and they receive information from some source. There are a great many kinds of social media, with each of the services or software using different ways of operating or connecting, each with different reasons for existing. Each of these services or software might even require a specific program or specific type of hardware to connect to it.

In general, the applications that a user utilizes, the websites that they visits, the ways that they exchanges information, and the other people they connect with can vary greatly. In addition, these variables can multiply because users can use multiple services or tools to utilize the social media, meaning social media can be used by utilizing one specific tool or by utilizing several different services and tools together, each with their own different way of functioning. To understand the process of how an individual or organization utilizes social media, a complex explanation of each of the individual components that they use might be necessary. Therefore, in order to group and clearly explain the multiple examples of social media use, the following elements have been identified.

Table 3-2: Identified elements related to social media in the East Japan Earthquake and Tsunami Elements related to social media in Groups the disaster Social networks • Twitter • Facebook • Mixi Services and software used SMS

Wikis Webpages or blogs Smartphone applications Maps Crowdsourcing • Crisis-mapping Social-network-based group type Organizational methods of the Info-list type software or services used Aggregator type Innovation type Design type Raw-social-media-data-mining type Large organization type Group or organization oriented • Mass-media type social media use • Direct-relief type

Government-social-media type General disaster information Safety confirmation Fundraising Infrastructure status notification/regional facility status Type of information shared Housing provision through social media Goods provision Moral support Resource saving Volunteer recruitment Special needs support Safety conformation

3.2.4 Previous Research Related to Social Media Use in the Disaster in Japan

Research by Acar and Murakami (2011) of Kobe City University of Foreign Studies conducted a survey two weeks after the earthquake with Twitter, tracking status updates of individuals in the affected area. The aim of the survey was to understand the benefits of social networks for individuals directly affected by disaster and others who were not affected but using Twitter to find news.

The research found that directly affected individuals’ status updates tented to be related to “their unsafe and uncertain situation” and that individuals in remote areas used Twitter to let others know that they are safe. They also found that status updates in the disaster hit areas were warnings, help requests, and reports about the situation and that official local authority Twitter accounts, which were set up after the disaster, were particularly useful because the information was well followed and retweeted extensively, especially in the case of tsunami warnings. The research also found that there were many retweets that were inaccurate information as well as that there were many calls for help that were lies or misplaced. In addition, users were concerned by these types of misinformation, but there were not enough status updates from government and mass media to combat the misinformation (Acar and Murakami, 2011).

The researchers make recommendations for use of social media in disaster that say users should have more responsibility for their tweets, that users should realize it is a public method of communication, and that information sources should be clear in updates. They go on to say that more research needs to be done to see if official hashtags keywords could help improve the use of Twitter in disasters (Acar and Murakami, 2011).

In another research study related to the disaster and social media, the Mobile Marketing Data Labo (MMD Labo) (2011) conducted a survey over the Internet over a five day period from April, 22nd to the 29th, 2011, with 1,891 individuals on the use of social media in the disaster on mobile devices. The survey consisted of a series of questions about usage of the social networks Twitter, Facebook, and Mixi, asking if individuals used the networks before or starting after the disaster and if the networks were helpful during the disaster.

Responses show that 63.9% of Twitter users said that information obtained through Twitter was helpful, with 34.7% of Facebook users saying information from Facebook was helpful and 26.0% of Mixi users saying information from Mixi was helpful (See Figure 3-10) (Mobile Media Labo, 2011). Responses also show that users thought that each social network was not helpful for anything at the following rates: 30.7% for Twitter, 49.4% for Facebook, and 57.9% for Mixi (See Figure 3-10).

Figure 3-10 : Helpfulness of social networks in the 2011 East Japan Earthquake and Tsunami Source: MMD Labo http://mmd.up-date.ne.jp/news/detail.php?news_id=799 Of the respondents, 42.1% were Twitter users, with 1.9% saying they started Twitter because of the disaster specifically. Facebook users made up 21.9% of the respondents, with 0.9% starting beginning to use Facebook after the disaster. Mixi users made up 50.8% of respondents, with 50.7% of them having used Mixi before the disaster (See Figure 3-11) (Mobile Media Labo, 2011).

Figure 3-11: Twitter, Facebook and Mixi use after the 2011 East Japan Earthquake and Tsunami by respondents Source: MMD Labo http://mmd.up-date.ne.jp/news/detail.php?news_id=799 The results also show that of the respondents that had not used Twitter, 26.9%, said that they have an interest in using Twitter, so there is potential for an increase in the use of Twitter (Mobile Media Labo, 2011).

3.2.5 Examples

As previously stated, the elements related to social media use in the disaster are tools (hardware, software, services), methods (crowdsourcing, crisis-mapping, etc.), groups

(organizations, mass media, governments, etc.), and the type of information being shared. These different elements can be combined in various ways with limitless possibility. Therefore, the following list does not aim to be exhaustive, only representative. Each element has an explanation and at least one example of its use in the disaster. Due to the fact that many examples of social media use in the disaster can use multiple tools, have multiple types of organizational methods and share multiple types of information, one specific example could be representative for multiple tools, methods, etc.

The combinations of these elements is what gives social media its flexible and ever- developing qualities, therefore understanding that fact is of prime importance when trying to understand how social media was used in general in the disaster. In the next section, the author gives a brief description of the elements and their attributes.

3.2.5.1 Services and software

o Services and software used ! Social networks • Twitter • Facebook • Mixi ! Email ! SMS ! Wikis ! Webpages or blogs ! Smartphone applications ! Maps

Social networks

Twitter

Of the examples of social media during that disaster that were found online, Twitter appears to be the most important tool in exchanging information and helping people to connect. Twitter started in mid 2006 (Mashable, 2011) and has about 200 million users as of July 2011 (Shiels, 2011). Twitter is a microblogging website, a blog-like website with very short messages of no more than 140 characters. Users post short messages to their profile, which other users can view. If a user is interested in the viewing all the future status updates of a particular user, her or she may “follow” that user.

When a user is followed, that person’s status updates, or “tweets” will appear in chronological order in the timeline, the main portion of the website or application used. The timeline is made up of all the tweets that a person follows. Every time a user follows another user, the followed user’s tweets will appear in the timeline of the user. The relationship between users can be one-way, meaning that the person followed does not have to follow a person that follows him or her, or the relationship can be two way.

Due to the possibility of one-way following relationships, the website tends to differ from other social network websites in that the connections on other websites tend to be made up of already existing personal connections. Other websites tend to work on exclusivity in the sense that individuals can only have access to an individual’s information if that individual agrees to give access and allows the creation of a two-way relationship. Twitter works in an open manner, although users can make their tweets private, but the use of this feature is rare.

Therefore, Twitter connections are varied; users can follow famous individuals, companies or total strangers as well as friends with a prior personal connection.

Users post messages to the website by via various means: SMS, mobile phones through a web browser, smartphone applications, through a computer’s web browser or a desktop application. Twitter was originally designed to be used with SMS, which has limited use in Japan because of restrictions by mobile phone providers.

Communication between users in Twitter is done by an @reply, which is the method of sending a message to another user by posting a tweet with the symbol @ and a users name. Tweets can also be shared by “retweeting,” which is when a user sends out a tweet to his or her followers that originally came from someone that they are following. This way one original message can be amplified and shared many, if not thousands, of times over. Twitter also uses “hashtag” keywords, such as #prayforjapan, one which was used during the disaster, to enable users to gather around a topic by searching the keyword and seeing all the tweets that have that hashtag.

Twitter is very popular throughout the world and has been used in many disasters, as described in the previous chapters, but Twitter was virtually unknown in Japan even two years ago in 2009, while it had already become popular in many other countries. Twitter was introduced and became instantly popular in Japan after the post-election protests in Iran (“Japanese Twitter Users,” 2009) and was featured heavily in Japanese mainstream media (Akimoto, 2009). As of March 29th, 2010, Japanese users on Twitter account for 6% of the user base, the largest group by country in the world, and make up 13% of the content, almost double the rate for that of other countries (Jenssen, 2010). As of August 2010, there were over 16 million Twitter users in Japan, about 10% of the population (Russell, 2010). A survey in 2010, showed that 44% of Twitter users have tweeted from their mobile phones (Tabuchi, 2010). In May 2010, Softbank, one of the mobile service providers in Japan announced that it was adding Twitter functionality to 14 models of its new, non-smartphone, mobile phone line (Tabuchi, 2010).

With Twitter’s new popularity in Japan, when the 2011 East Japan Earthquake and Tsunami happened, there were many Japanese people who were using Twitter and a large amount of them using it from their mobile phones. Although, not exclusively in Japan, Twitter’s user during and following the disaster grew by one third, to 7.5 million (Hosaka, 2011). For the whole day of the disaster, the daily tweet count was 177 million tweets, 37 million more than the daily average (Smith, 2011).

Twitter later released data showing the influx of information on the website saying:

On Twitter, we saw a 500% increase in Tweets from Japan as people reached out to friends, family and loved ones in the moments after the March 2011 earthquake. This video shows the volume of @replies traveling into and out of Japan in a one-hour period just before and then after the earthquake [See Figure 3-12]. Replies directed to users in Japan are shown in pink; messages directed at others from Japan are shown in yellow (Twitter, 2011).

Figure 3-12: Volume of @replies traveling in and out of Japan in a one-hour period before and after the earthquake Source: Twitter and Twitter user @miguelrios http://www.flickr.com/photos/twitteroffice/5885172082/ Another video from Twitter shows worldwide retweets of tweets originating in Japan, for one hour after the earthquake (See Figure 3-13). In the video the senders’ original tweets are shown in red and tweets retweeted by their followers in the hour after the event are shown in green (“Spread of Information, 2011).

Figure 3-13: Spread of information of worldwide retweets of Tweets originating in Japan for one hour after the earthquake Source Twitter and Twitter user @miguelrios http://www.flickr.com/photos/twitteroffice/5884626815/ During the disaster, Twitter users around the world used the hashtags #tsunami and #prayforjapan and Twitter users inside Japan used hashtags such as #jishi (meaning “earthquake” for earthquake information), #j_j_helpme (for aid and rescue), #hinan (meaning “evacuation” for evacuation information), #anpi (meaning “safety” for safety conformation of individuals) and #311care (for medical information) (“Analysis and paper.” 2011).

An example of hashtag use is this example from a Twitter user @gusan who says “I was watching the stream of this hashtag and couldn’t help crying. Everybody in the world is tweeting for Japanese. #prayforjapan” (See Figure 3-14) (Nishimura, 2011).

Figure 3-14: An example of hashtag use during the disaster Source: Twitter user @gusan http://www.collabo311.com/2011/07/297/ During the disaster Twitter published a blog post in Japanese and English with an explanation of how to use Twitter during the disaster and the hashtags for different topics (See Figure 3-15 and Figure 3-16). Twitter also published a mobile phone website with the same information (http://twtr.jp/earthquake) (See Figure 3-17) (“Touhoku chihou,” 2011). This is significant because although only a portion of the population in Japan has smartphones, which can use applications for Twitter, nearly all mobile phones are Internet capable, so access to a mobile phone based Twitter website would make Twitter available to virtually anyone with a mobile phone, which is almost 115 million mobile phones for a population of 126 million (CIA, 2009)

Figure 3-15: Portion of the Twitter blog post explaining hashtags for the disaster Source: Twitter http://blog.jp.twitter.com/2011/03/blog-post_12.html

Figure 3-16: Portion of the Twitter blog post explaining how to use hashtags and GPS to get help Source: Twitter http://blog.jp.twitter.com/2011/03/blog-post_12.html

Figure 3-17: Images of the Twitter blog post for mobile phone Source: Twitter http://blog.jp.twitter.com/2011/03/blog-post_12.html In other countries, much of this kind of communication is done through Twitter via SMS. Until very recently in Japan, one mobile phone provider, Softbank, provided SMS service, with other providers, such as NTT Docomo and AU KDDI using separate messaging systems that can only be used within that network. Mobile email is used for communication between the users of different mobile service providers. Due to this, Twitter does not provide SMS based numbers for status updates for using Twitter in Japan. Twitter users who access Twitter on mobile devices use applications specifically designed for Twitter or they use Twitter’s mobile web interface. These applications and web pages use significantly more data than an individual SMS, and so, Japan not having had SMS-based Twitter communication, presents an obstacle to communication during disasters which users of Twitter in many other countries do not have. If Twitter provides a SMS based number for Japan, it is likely that this SMS- based Twitter use will be used during future disasters.

Many of the individuals and groups that used social media in the disaster used Twitter as a method of connecting to other social media tools. During the course of the author’s research, many users on Twitter who stated that they began Twitter after the disaster were encountered.

Facebook

Facebook, the third largest social network in Japan, is another social network that was widely used during the 2011 East Japan Earthquake and Tsunami. It started in 2004 and only became open to the public in late 2006 (Abram, 2006). As of September 2011, it has 800 million active users worldwide, 50 percent of whom access the website daily (Facebook, 2011). Facebook is a second-generation of social network sites that have innovative features that let users connect easier than first generation social networks like MySpace and Friendster. In

Japan, a first-generation social network, Mixi, is widely popular and has features that are specially geared toward Japanese users, where as Facebook lacked Japanese language support for quite some time and lacked features that would allow Japanese users to enter their names in both Chinese characters and roman characters, so Facebook’s pogress in Japan has been slow. In January of 2011, only 2% of people who use the Internet in Japan used Facebook, about 2 million people, as opposed to 20 million users of Mixi (Preston, 2011), but the number is rising. As of October 2011, the number of Facebook users in Japan had grown to 5 million (“Japan Facebook Statistics,” 2011).

Cultural differences also present barriers to adoption of Facebook in Japan. Many Japanese users do not feel comfortable using their real names online (Tabuchi, 2011a). Facebook asks for real names when signing up and the company stresses the importance of “real world connections” in its network (Tabuchi, 2011a). A survey in Japan by the Mobile Marketing Data Labo showed that 89% of 2,130 respondents were reluctant to use their real names online, which differs from the real-life relationship based connections of users from the US (Tabuchi, 2011a). In another survey by Microsoft, over 50% of respondents in Japan said that not even one of the connections to others in social networks was a close friend (Tabuchi, 2011a).

Although having yet to take off in Japan, Facebook is the largest social network in the world and, as such, it played a great role in communication after the disaster. Many people around the world learned about the disaster through Facebook and used Facebook to express their support as well as donate to the relief effort. The social network also played a large role in helping friends and family of directly affected individuals get in touch with their loved ones. Facebook and its group functionality was also used by many organizations and spontaneous volunteers to help with the relief effort in terms of direct support and informational support. It provided the means of communication for the groups to function. In addition, due to the fact that Facebook is very popular outside of Japan, many foreign residents rely on Facebook for communication with friends and family back in their countries of origin, and so when the disaster occurred it became the most logical method for communication. There are countless articles and comments in mass media surrounding the event about how Facebook became a lifeline in the disaster.

Facebook, the company, created an animated graphic using data made from the status updates of users after the disaster (See Figure 3-18 ). The video shows the location of some of the 4.5 million status updates from 3.8 million users who used the words “Japan,” “earthquake” or tsunami” after the disaster (Kessler, 2011). A spokesperson from Facebook stated that the social network was being used as a tool for people to get in touch with friends and family in the crisis (Fergusson, 2011).

Figure 3-18: Image of a video graphic from produced by Facebook showing status updates related to the disaster

Source: http://www.bbc.co.uk/news/technology-12730297 Some of the many examples of individuals notifying their family members and friends of their safety are similar to the experiences of Nicolas Savio, a student who having just landed in Tokyo when the earthquake happened, said that since the phone lines were down he relied on email and social media for communication (Blackburn, 2011). Another woman, Patty Johnson, whose son lives in Tokyo, said she was worried for hours until she found out that her son and his pregnant wife were ok through Facebook, saying “Thank God for Facebook” (Stephens, 2011). Some Japanese students attending a satellite campus for Showa University, in Boston Massachusetts, USA, were able to contact their families through Facebook and Twitter, with an administrator at the school commenting, “With Facebook and Twitter, it was the key for them to find out within a half a day if their parents and friends and family were OK or not” (Shea, 2011). Another man in the US, whose father was in Japan when the disaster happened, says he loves the power of social media because he can get updates from his father and information from media outlets at the same time (Jacobson, 2011). In addition, in some cases, foreign journalists covering the disaster used Facebook, such as Facebook’s chat, as a means to communicate and interview affected individuals (Hornyak, 2011).

Individuals who are not directly affected or indirectly affected used Facebook and its group functionality to express support and share information. One group, Japan Earthquake, in less than 12 hours, had 3,000 “likes” (By clicking “like” on a group page, updates from that group then appear in a user’s News Feed) (Gaudin, 2011). The website Anpi Report, made after the disaster specifically for confirming the safety of individuals affected by the disaster, used Facebook as one of the methods to recruit the 250 volunteers who helped to enter data on individuals’ whereabouts (Ima, watshitachi, 2011).

Facebook, as an organization, also helped out in the disaster by providing a centralized page for disaster information called Disaster Relief (See Figure 3-19). As of October 2011, the page is being followed, “liked,” by over 680,000 people. For users in Japan, Facebook also put information on blackouts and train service above users’ News Feeds, the first page users see when they log into Facebook (See Figure 3-19).

Figure 3-19: Global Disaster Relief on Facebook Source: https://www.facebook.com/DisasterRelief

Figure 3-20: Information related to the disaster provided by Facebook on users' News Feeds Source: Facebook Nobuyuki Hayashi, a prominent Japanese tech journalist and consultant, said that after the disaster a large amount of people in Japan joined Twitter to exchange information but rumors and misinformation about a fire that was said to create toxic rain spread through Twitter and so some people turned to Facebook because the flow of information is more trusted and the conversation style is topic based (“Japan disaster,” 2011). Facebook also was reported to have misinformation problems such as donation scams (Sherman, 2011).

Mixi

Mixi is a Japanese social network website that started in 2004. As of January 2011, Mixi has 21 million users (Tabuchi, 2011a). Mixi began earlier than the other current popular social networks in Japan, Twitter and Facebook, which popularized short status updates, so Mixi has traditionally been more similar to an older social network, MySpace, but recently it has added functionality similar to Twitter and Facebook, so it has many features of them both. Mixi users tend to not use their real names or photos (Tabuchi, 2011a).

Figure 3-21: A Mixi profile Source: Marketingpilgrim.com http://www.marketingpilgrim.com/images/mixi.jpg Mixi has status updates similar to Twitter and profiles and groups or “communities” similar to Facebook Groups. These were both used in the disaster, but the use of the groups appears to be more common. There are many “communities” on Mixi dedicated to the disaster. One group has 346,333 users (See Figure 3-22).

Figure 3-22: A Mixi group dedicated to the disaster with 346,333 members Source: Yahoo Blogs user Shigeto1953 http://blogs.yahoo.co.jp/shigeto1953/archive/2011/4/16 Mixi, as a company, also took actions to help in the disaster. At the log in page of the website after the disaster, large images with links to Google Person Finder were placed as well as messages asking users to conserve energy (See Figure 3-23). Mixi also provided a blog post with a list of all the communities related to the disaster and links to other websites and services. The list had four categories: general, affected person’s support information, regional and “other.” The groups had very high membership numbers because Mixi prohibited the making of new groups during the disaster (“Jishin – Mikushi,” 2011). This was probably to make sure that more than one group for the same topic was not made, which often happens with Mixi communities. This style of approach is unique amongst the examples of the social networks.

Figure 3-23: Image of Mixi's sign in page directing users to Google Person Finder and asking people to conserve energy Source: http://www.howmuch-ad.com/2011/04/mixi-google.html

Figure 3-24: Image of Mixi's list of groups related to the disaster Source: Mixi http://news.livedoor.com/article/detail/5410243/ Mixi is a very helpful tool and it clearly has been of help to many people, but it has a few limitations in disasters. Mixi is also only offered in Japanese, locking out most non-Japanese persons. Also, when making an account on Mixi, a Japanese mobile phone number must be registered for safety reasons and to prevent duplicate accounts, but this effectively locks out people who do not live in Japan including Japanese citizens who might live abroad. The barriers to non-Japanese speakers and individuals not in Japan might create walls the restrict the flow of information, which might make other social networks a more attractive choice to share information during a disaster, especially for Japanese people that have friends or family that do not speak Japanese.

Although email itself is not considered social media due to the fact that it existed before the advent of Web 2.0, it can act as a form of communication with other forms of media, so it should be mentioned here as a tool. In addition, SMS, which is a tool often used with social media is not used between mobile service providers in Japan and so its use is limited, thus Japanese mobile phones and smartphones use email to communicate between service providers. Therefore, it is likely that many people relied on mobile phone email during the disaster. There are accounts of chain emails with incorrect information being spread around after the disaster (“Saigaiji no cheenmeeru,” 2011).

SMS

As previously stated, SMS is not used widely in Japan. Mobile service providers have an internal messaging system for use between users of the same service provider. At the time of the disaster, and in the case of Softbank, this messaging system is SMS, so Softbank users can use SMS to communicate with individuals using Softbank and others abroad. For communication between different mobile service providers, mobile email addresses are used.

Twitter, which is arguably the most helpful form of social media in disasters, uses SMS, except for inside Japan for the previously mentioned reason, but it does not use email. Therefore, users of Twitter in Japan for mobile use must rely on applications such as ones used by smartphones or mobile web browser to view Twitter’s mobile web page. Applications and mobile web pages use significantly more data than individual SMS messages, and so this could create unnecessary burden on networks in disaster, which would mean that less vital disaster information would be able to get through. This could delay communication at vital times. The major mobile carriers in Japan, Docomo, KDDI, SoftBank Mobile and eAccess (Emobile), announced that inter-carrier SMS would begin in Japan starting July 13th, 2011. This date is after the March, 11th, 2011 and so inter-carrier SMS was not used during the disaster. Given this development, if Twitter provides a local SMS number for Japan, SMS and Twitter could be used together in the case of another disaster, which would be beneficial because of its low data use.

SMS was in other countries during the disaster to exchange information, so it is valuable in that sense. Also, SMS was used to collect donations from mobile phone users around the world. In the case of the US, the Red Cross used the number 90999 to receive USD $10 donations, raising almost USD $3 million in less than a week (Duke, 2011).

Wikis

Wikis are websites that can be communally edited. Wikipedia is the most widely know wiki. During the disaster, wikis were widely used by groups sharing information on the disaster, usually in the form of lists for people looking for information on how to receive or give aid during the disaster. Wikis are a powerful tool for this kind of task, because several people can add information from various sources on one page at the same time or different times. Prior to the advent of wikis, to make such a list would require it to be done by one person or to coordinate constantly to exchange information and have one person manage it.

One strong point of a wiki is that it is static. For example, in Twitter information flows by in a timeline so it can be hard to locate past information. Many helpful links may be there but they are not in an organized order and if users are not there at the right time they might not see the links. Using a wiki, a user could take many of the links to organizations and information that they found on Twitter or another website, organize it in a wiki with the help of other people and then share that link on Twitter and thereby have a very detailed, helpful list. Individuals

who edit and add information to wikis tend to be more experienced with social media and computers than users who do not.

During the disaster, many lists of information related to the disaster, what are identified here as “info-lists” were made with wikis. The most comprehensive wiki with information related to the disaster is the “higashinihon daijishin (touhoku chihou taiheiyouoki jishin)@wiki,” translated as “The East Japan Earthquake and Tsunami (Tohoku area Pacific Ocean Earthquake)@wiki” (http://www46.atwiki.jp/earthquakematome/) (See Figure 3-25). It has an extremely large amount of information in Japanese and English related to the disaster such as information on the details of the earthquake, tsunami and nuclear accident, disaster prevention, electricity, tsunami warnings, hospitals, shelters, people finding, etc.

Figure 3-25: “The East Japan Earthquake and Tsunami (Tohoku area Pacific Ocean Earthquake)@wiki” Source : http://www46.atwiki.jp/earthquakematome/ There are also topic specific wiki websites related to the disaster. One such example is the website Olive (https://websites.google.com/website/oliveinenglish/), a wiki specifically made for design-oriented solutions to the problems created by the disaster. The website is in Japanese, Korean, Chinese, and English and has directions for various solutions for disaster related problems, such as how to make a mask out of a shirt or how to make an aluminum can stove (See Figure 3-26).

Figure 3-26: Olive wiki website with directions on how to make a mask from a shirt Source: https://websites.google.com/website/oliveinenglish/nichi-youhin/safety_mask

Webpages or blogs

Webpages themselves are not considered social media, but a website that is a social media website is made up of webpages, and when users share information through social media it is done so with links to webpages, which could be newspaper articles, webpages with photos, people finding databases, virtually anything. Therefore, webpages are one of the building blocks of social media.

Blogs are sometimes considered to be social media because with the advent of Blogger, which is a product of Google, it became easier for individuals who were new to using computers and the Internet to be able to publish on their own on the Internet. During the disaster, many info-lists were made by individuals using blogs. These lists were often self compiled, differing from the wiki-based info lists, which are group compiled. An example of a info-list on a blog is this one on Greenz.jp (http://greenz.jp/2011/03/12/jishin/).

Smartphone applications

Smartphones, which are mobile phones that can use applications like a computer, such as the iPhone and Android platform phones, have become newly popular in Japan. Many of the phones rely on mobile Internet data to function. During the disaster, applications that run on the iPhone and Android platform phones and interface with Twitter Facebook and other websites were used by smartphone users. In many cases, the mobile phone lines were jammed because of an excess of calls, but individuals with smartphones were able to use 3G mobile Internet data and these applications to get and share information. Therefore, these applications and devices can be extremely helpful in a disaster.

Figure 3-27: Image of a Twitter application for iPhone taken right after the earthquake with users commenting that an earthquake had occurred Source: http://d.hatena.ne.jp/Batayan3/20080921/1222003029 The mobile service provider Softbank released a disaster message board the day after the disaster that allows people to register that they are alright and check on the safety of others. The application lets users preregister people to auto email if the user logs in after a disaster. That way, just by registering that one is safe, the application will notify friends and family, without having to write an email or select individuals to send to (“Saigai you,” 2011).

Figure 3-28: Softbank iPhone disaster message board application Source: http://itunes.apple.com/jp/app/id425650996?mt=8 One application called “Yurekuru kouru” sends out notifications about impending earthquakes. The mobile service provider NTT Docomo has a similar service but the other mobile service providers do not, so this is a way that users can get the service even if they do not use NTT Docomo..

There are also first aid applications that can be helpful in disaster. Some applications have maps that can be stored inside the phone and used even when there is no Internet data available. With the applications, GPS can still be used without having access to Internet data and show a user where they are on a map. In the case of the tsunami, where roads were no longer recognizable, this functionality could be helpful.

Maps

Maps such as Google Maps and Google Earth are powerful tools in disaster when combined with other information. Sinsai.info was a website built during the disaster that uses the Ushahidi platform, a crisis-mapping platform described in the section about the 2010 Haiti earthquake (See Figure 3-29). The website overlays data gathered from individuals throughout the affected area, the country, and the world using social media, email, mobile phone email, SMS and other forms of communication. The website can show information on where to find shelter, what areas have lost electricity, evacuation locations etc. This type of information gathering is called “crisis-mapping” and would not be possible without map services like Google Maps and Open Maps.

Figure 3-29: Image if the Sinsai.info website Source: http://www.sinsai.info/

Table 3-3: Summary table of services and software used and their characteristics Services and software Characteristics used Social Networks Act as network between tools. Vary greatly by type. - Twitter Popular in Japan. Highly public. Concise and fast info. Info is easily shareable and user driven. High level of anonymity. One-way relationships possible. Character amount limits. Had problems with rumors. - Facebook Popular amongst foreign community. Less public. Not so popular in Japan. Low level of anonymity. More likely to have personal friends. Resending shared info not always straight forward. Has groups. Good for group networking. - Mixi Popular in Japan. Japanese language only. Sharing info is less fluid. New group creation related to the disaster was locked to centralize info. Directed users to Google Person Finder. Email Main text-based type of communication for cell phones in Japan. Sharing of info possible, but more difficult than social networks. Tracking origin info difficult because not public. Accounts of chain mails with incorrect information widespread. SMS Cell phone based. At the time in Japan, not for cross-network use so use was minimal. Used widely outside of Japan to collect donations. Highly valuable in disasters in general. Character amount limits. Wikis Crowdsourced. Information rich. Accessible by even low level users. Collaboration usually is by higher level users. Commonly used as Info list. Innovative. Webpages or blogs Commonly used. Common for info lists, but usually created by one person, differing from wikis. Information shared in social media links to them. Smartphone applications Must have smartphone to utilize. Offers GPS functionality. Mobile. High flexibility. Not commonly used by lower level users. Maps When aggregating other information, easily useable to understand a complex situation using crisis-mapping. Requires a computer to use easily. Usable on smartphones but somewhat difficult.

3.2.5.2 Methods

Organizational methods of the software or services used

• Crowdsourcing

o Crisis-mapping

• Social-network-based group type

• Info-list type

• Aggregator type

• Innovation type

• Design type

• Raw-social-media-data-mining type

Crowdsourcing

Crowdsourcing is the idea of using the Internet and related technologies to accomplish a task by using the collective efforts of a large group of people who are dispersed over a large area. The most well know crowdsourcing endeavor is Wikipedia. During the 2011 East Japan Earthquake and Tsunami, much of the non-affected and indirectly affected use of social media used crowdsourcing to accomplish tasks. Many info-lists are crowdsourced using wikis. Sinsai.info, which uses the Ushahidi platform to create crisis maps functions on the idea of crowdsourcing. Translation can also be done through crowdsourcing. The concept can be applied to many different types of tasks.

Crisis-mapping

During the disaster the Sinsai.info, which runs the Ushahidi platform (more detailed info in the sections titled Maps and 2010 Haiti Earthquake), seemed to be the main crisis-mapping website. Other websites have crisis-mapping and Google Earth, which is able to aggregate information from organizations, was also a way that crisis-mapping was used.

After the radiation leakage and subsequent crisis a website for tracking radiation leakage called Safecast.org was created (See Figure 3-30). It lets residents upload Geiger counter reading to the website and uses data from specially outfitted cars that drive around the affected area and take readings. The organization, started one week after the radiation leakage as of June 30th, had USD $57,000 using a crowdsource-style donation website called Kickstarter (Shreffler, 2011).

Figure 3-30: Image showing part of Safecast.org website with logo and radiation crisis map Source: http://blog.safecast.org/ Another website, Pachube created a radiation crisis map using Google Earth from information from Safecast.org and other radiation crisis-mapping sources (See Figure 3-31).

Figure 3-31: Radiation crisis map using Google Earth from Pachube.com Source: http://community.pachube.com/node/611#3d

Social-network-based group type

Websites like Facebook and Mixi, which have group functionality, make it easy to make a group for a specific purpose, add users, and share information. During the disaster, people all around the world made Facebook groups, and people in Japan made Facebook groups and Mixi communities to share information on the disaster. Due to the fact that the group functionality in these websites tend to be computer-based as opposed to mobile phone-based much of the work in sharing information was done by non-affected and indirectly affected individuals and made available for affected individuals.

Although, Facebook and Mixi tend to be used as a way to connect with individuals with which the users have preexisting relationships, the group feature is an easy way for individuals who do not know each other to connect and cooperate. Providing information by this means is one way that individuals who are not affected and have no relation to the disaster can be helpful.

Figure 3-32: An example of a social network group Source: Facebook

Info-lists

Info-lists are lists of resources, such as people finding websites, locations of shelters, translation services, etc. all compiled in one location for directly affected and indirectly affected individuals. These can be on blogs, news websites, or wikis.

There are also mobile phone webpage info lists like “Hang in there TOHOKU” (http://www.geocities.jp/saveeq311/) (See Figure 3-33). These can be helpful because after the disaster occurred 3G mobile phone Internet data was still useable in many places and the vast majority of cell phones are not smartphones, so mobile webpages like “Hang in there TOHOKU” would be viewable by these devices.

Figure 3-33: Mobile phone webpage info list, "Hang in there TOHOKU" Source: http://www.geocities.jp/saveeq311/

Aggregator type

Aggregator type information is information that comes from several websites or several sources and is displayed in one location usually using some sort of search parameter. Many websites that aggregate information are set up to show lists of tweets using specific hashtag keywords from Twitter. One such website is SAVE JAPAN! (http://savejapan.simone- inc.com/) (See Figure 3-34).

Figure 3-34: Aggregator type website SAVE JAPAN! showing list Twitter hashtag keyword searches Source: http://savejapan.simone-inc.com/index.html

Innovation type

After the disaster, a few projects, realizing the creative response that was happening using social media, created websites or services where Twitter users could suggest new and innovative ways to help in the relief efforts. One service, set up by the environmental website Greenz.jp, used the service bla.bo, a Twitter hashtag keyword aggregator to collect ideas (http://www.bla.bo/c/greenzjp/mkf5493/). Another, Fanfan.jp, created a website to let user share the tools and applications they were using so that they could be intergrated into a smartphone application called “Tasuke ai Japan” that was created after the disaster as a central location for disaster relief information on a smartphone (See Figure 3-35) (http://review.fansfans.jp/campaigns/fansfansjp/20110412/) .

Figure 3-35: An image of the website Fanfan.jp soliciting tool and app suggestions from Twitter users Source: http://review.fansfans.jp/campaigns/fansfansjp/20110412/ Design type

Another type of innovative website that was created after the disaster was one that use graphic design or design in construction of objects to help in the relief effort. These websites connect to users and get ideas using social networks like Twitter and Facebook. Ideaforlife.jp is one such website and it has creative ideas such as labels designed for boxes that ship relief aid goods, which show images of the type of good contained and places to enter quantity so receivers know exactly what they are getting (See Figure 3-36). Previously mentioned Olive is another example of this type of website.

Figure 3-36: Relief goods box label design Source: http://www.ideaforlife.jp/

Raw-social-media-data-mining type

More complex forms of social media use, which rely organized groups to function, can use data mining of social media websites as a way to gather useful information that can be of use to affected and indirectly affected individuals. Anpi Report, a searchable people finding website, is made up of tweets about the safety of individuals that was found and entered into the Anpi Report database by volunteers. By taking raw information from social media that is accessible on the Internet and using motivated volunteers who “just want to do something,” it is possible to create useful tools that can be useable by directly affected and indirectly affected individuals.

Table 3-4: Summary table of organizational methods of the software and services used and their characteristics Organizational methods of Characteristics the software and services used Crowdsourcing Method of accomplishing tasks with mass group collaboration, spanning location. Projects tend to be highly innovative. Usually higher level users participate. Wikipedia is an example. - Crisis-mapping A form of crowdsourcing. Ushahidi and Sinsai.info are examples. Has ability to aggregate and visually represent mass amounts of information. Highly useful. Information gathered by individuals inside and outside of the disaster is made available for affected individuals. Highly utilized in the 2010 Haiti earthquake. Social-network-based group Exists within social networks. Easy to create and join. type Many spontaneous volunteers became involved through these types of groups. Easy way to share information and coalesce around an issue. Info-list type Very common during the disaster. Most simple way to share information. Commonly done by one individual on blogs and done by groups on wikis. Simply a list of sites and resources. Aggregator type Collects data from several places usually using a search word. Commonly used to display Twitter keyword hashtags about a particular topic. Helpful for getting a mass amount of information on one topic or location, for example the keyword hashtag #save_miyagi would be specific to the city of Miyagi. Innovation type A way for social media users to share ideas to create new ways of offering information-based support. Utilized by users who recognize that social media is innovative in an effort to exploit that aspect. Design type Similar to innovation type. Often used by artists. Seek to solve specific problems, often using tangible objects. Raw-social-media-data- A method of taking information shared on social media mining type and reorganizing, often with the help of a large group of people, to accomplish a task such as safety identification.

3.2.5.3 Group or Organization-oriented Social Media Use

• Large-organization type

o Mass-media type

o Direct-relief type

• Government-social-media type

Large organization type

By using the above mentioned tools and methods, groups of people connected by social media can accomplish tasks that were not before possible. During the disaster, there were groups that were spontaneously organized through social media that performed vital tasks for affected and indirectly affected individuals. Also, there were many pre-existing groups, organizations, companies and local and national governments that utilize social media to share information and accomplish goals.

Google is an example of a large organization with massive resources, that when connected to social media, can provide vital disaster information, such as Google’s Person Finder attempted to do so during the disaster. Google had created its Google Person Finder service (See Figure 3-37) for the earthquake in Haiti in 2010. During this disaster in Japan, since many people affected by the disaster did not have access to the Internet, Google asked people to take pictures of the lists of missing persons at shelters, which were uploaded to Google’s photo service Picasa and then copied into the Google Person Finder database by 100 Google employees (Tabuchi, 2011b). This particular relief would not have been possible without social media and the vast resources and creativity of Google.

Figure 3-37: Google Person Finder Source: http://japan.person-finder.appspot.com/?lang=en

Mass media type

Existing mass media organizations can use their resources and information in extremely beneficial ways during disaster by sharing information via social media. During the 2011 East Japan Earthquake and Tsunami, Television-based broadcasts were made available on the Internet through a live-streaming broadcast service called Ustream (See Figure 3-38), which the mobile service provider Softbank owns a 30% stake in (Tabuchi, 2010). By providing

these television broadcasts online, social media users can view broadcastss on smartphones over the Internet, special English language and other language translation broadcast can be made and people all over the world can get first hand information from Japanese television networks over the Internet without having access to terrestrial broadcasts, which in order to view would require that the individual be in Japan.

Figure 3-38: Live streaming of TBS News on Ustream after the disaster Source: Ustream Another example of this kind of mass media use of social networking is the website Liason Hisaito (See Figure 3-39 and Figure 3-40) (http://www.nhk.or.jp/hisaito/), made by NHK, Japan’s national news organization, which is a way for affected and unaffected individuals to send messages to each other as well as a collection of informational videos for affected and unaffected people.

Figure 3-39: Image of NHK's website for affected and unaffected people to share messages with each other Source: http://www.nhk.or.jp/hisaito/

Figure 3-40: Image of NHK's website for sharing informational videos related to the disaster Source: http://www.nhk.or.jp/hisaito2/chie/index.html

Direct relief

Social media use can be beneficial in direct relief. In the 2011 East Japan Earthquake and Tsunami, organizations like Peace Boat used social media to recruit volunteers, organize and keep interested persons up to date on the efforts of the organization (See Figure 3-41).

Figure 3-41: Image of Peace Boat's Tohoku Kanto Emergency Relief Facebook page Source: Facebook Another direct relief style of support made possible by social media is that of the organization Relief 2.0. The organization, which was active in disaster relief after the disaster, focuses on listening to the needs of the people who are affected and strives to meet the needs by using independent units in the field equipped with mobile phones who use social media to ask individuals outside of the affected area to help meet those needs.

Figure 3-42: Slides from a Relief 2.0 presentation at TedxEarthquake 9.0 Source: http://www.youtube.com/watch?v=BTzEaEYUY2A

Government social media type

During the disaster, many local governments turned to local media to get information out. In some cases, increased access to local government websites caused the websites to crash and so social media like Twitter could be used as a back up method to get information to residents.

In one example, two days after the earthquake, the prime minister’s office in Japan created a Twitter account, @Kantei_Saigai,, for announcements about evacuations and press conferences, and then an English account five days later, so local foreign residents and foreign organizations could be kept up to date. The account had 7,000 followers four hours after creation (Peters, 2011).

3.2.5.4 Type of Information Shared through Social Media

• General disaster information

• Safety confirmation

• Fundraising

• Infrastructure status notification/regional facility status

• Housing provision

• Goods provision

• Moral support

• Resource saving

• Volunteer recruitment

• Special needs support

Social media is used to exchange information and so during the disaster some specific types of disaster related information were exchanged. The following list is an attempt to identify the major types.

General disaster information

This type of general information related to the disaster appears to be the most common since many people affected and unaffected wanted to properly grasp the situation of the disaster. Due to the fact that the situation in the disaster changed so rapidly and that new information was becoming available all the time, there was a large amount of this type of information being shared constantly. Much of this information comes from news organizations, governmental sources or directly affected individuals and then is shared over and over through out social media.

Safety confirmation

Individuals confirming their safety via social media or asking about the safety of friends or family members seemed to be extremely common. Many websites such as Google Person Finder, buji.me and Anpi Report were set up to help find people who may or may not have been affected by the disaster.

Fundraising

Social media was also commonly used with fundraising activities. Due to the outpouring of goodwill toward Japan and affected people by people around the world, donations were one way that people felt they could do something to help. Social media was used to solicit donations via SMS or through applications like iTunes (Haydon, 2011).

Another form of fundraising that depended on social media was charity publishing fundraising made from tweets and letters from affected individuals and unaffected individuals looking to express support. One of the most famous books is called “2:46 Aftershocks: Stories from Japan Earthquake,” which came out of a website called Quakebook,org. The book has writings from Yoko Ono, William Gibson and others. 100% of the proceeds go to the Japanese Red Cross Society (“About,” 2011).

Infrastructure status notification/regional facility status

Websites like “Urayasa dansui jouhou“ (http://stnde.com/urayasu/), which was designed to gather information on infrastructure damage related to flooding due to the earthquake, used Twitter and a Mixi community to get information from residents on the status of flooding and damage around the town of Urayasu in Chiba. The website also has aggregated Twitter streams around hashtag keywords related to the website and a Google Map showing the damage. A Twitter sharing widget on the website shows that 2,571 people have tweeted about the website as of October 30th, 2011.

Housing provision

After the disaster, a few organizations made efforts using social media to find housing for people displaced by the earthquake, tsunami or radiation leakage. One of the groups was created using the group functionality in a social networking website called CouchSurfing, which was designed to allow individuals who enjoy traveling to places to stay free of charge with other individuals who enjoy traveling. The website says there are 3 million couch surfers of whom 31,426 have used the website in the last week (October 31st, 2011) to meet other travelers or find places to stay. The group that grew out of the disaster, called Japan Crisis Housing (See Figure 3-43), as of October 31st, 2011, has 728 members

Figure 3-43: Image of a portion of the webpage for the group Japan Crisis Housing on Couchsurfing.org Source: http://www.couchsurfing.org/group.html?gid=39703 Another group born out the disaster, is the Facebook group “Beds in Osaka” (See Figure 3-44). Unlike the CouchSurfing group, which is a social network for finding places to stay, Beds in Osaka was spontaneously created by concerned individuals that “just wanted to do something.”

Figure 3-44: A portion of the Facebook group Beds in Osaka Source: https://www.facebook.com/pages/BEDS-in-Osaka/188015291240350 It seems that these groups might have not been utilized to a very high degree. Reasons might be that government offered shelter was adequate in terms of numbers, individuals had friends or family they could stay with, Japanese people might feel uncomfortable staying with people they don’t know, and many of these groups have a large non-Japanese member base, which might cause communication and cultural barriers. This type of relief support might be more successful in other countries.

Goods provision

After the disaster, many people wanted to help by sending supplies to the affected area, but because of logistical problems this was not possible, so organizations clearly stated that only monetary donations could be accepted and that money would be used to procure supplies in bulk. Due to the inability of average people to send goods, an organization called “Todoke” (See Figure 3-45) was created by some innovative individuals. The idea behind Todoke was that individuals could suggest what kinds of supplies they would like to donate and then through social media other individuals could add their support to types of supplies they would also like to donate. Then companies with large resources would purchase the goods on behalf of the individuals who had wanted to donate the supplies.

Figure 3-45: Image of the website Todoke showing a donation offer Source: http://todoke.org/

Moral support

The outpouring of moral support throughout social networks and social media was vast. For days after the disaster, besides information specific to the disaster about the only type of communication to be found on social networks was messages of support for Japan and those affected by the disaster. Another form of moral support that used social media after the disaster, which is made up of the first form, was the previously mentioned charity-publishing type. “2:46 Aftershocks: Stories from Japan Earthquake” (See Figure 3-46), which is more commonly known as “quakebook” because of the use of the hashtag keyword #quakebook in its making through Twitter, is an example of this type of moral support.

Figure 3-46: An image of the Twitter-sourced book "2:46 — Aftershocks: Stories from the Japan Earthquake" Source: http://quakebook.org/about/

Resource saving

The earthquake and tsunami caused the Japanese government to shut down 15 of its 54 nuclear reactors following the disaster. Due to this, society as a whole in Japan must make drastic efforts to reduce energy usage with the government requiring big power users to reduce peak consumption by 15 percent and encourage residents to make efforts to help, in ways such as by reducing air conditioning use, making sure lights are not left on, etc. (“In, Japan,” 2011). Amongst this backdrop, Google created a project called “Go Setsuden” (See Figure 3-47) that uses social media so that an individual can share the ways that they are helping to reduce energy consumption and encourage others to do the same. The project also allows people to make suggestions on better ways to save energy, which can easily be shared through social media (“Go Setsuden, 2011).

Figure 3-47: Image of a widget for the Go Setsuden project by Google showing actions that users can choose to do Source: http://googlejapan.blogspot.com/2011/06/go.html

Volunteer recruitment

During and after the disaster, volunteer recruitment using social media from organizations such as the previously mentioned Peace Boat was common, as well as volunteer recruitment for spontaneously created projects or groups such as the previously mentioned “Beds in Osaka” or “Quakebook.” It seemed that after the disaster there were a great many people who “just wanted to do something” to help even if they did not know anyone affected by the disaster. This spirit of giving can be accessed and utilized after a disaster is as new type of virtual donation that can perform real world tasks and have real world effects.

Special needs support

In any disaster there will be people who have special needs that cannot be met by the methods that are used to provide information and support to the vast majority. During the disaster, social media allowed people to communicate, cooperate and help to solve some of the problems presented to individuals with special needs.

After the disaster happened, there were many foreign residents who do not speak Japanese fluently and had trouble communicating. Social networks Twitter and Facebook allowed individuals with limitations with Japanese language to receive information in their native languages. Also, a social media powered website called Babelverse, which is focused in translation services, provided a place for volunteer translators to sign up and, through free phone calls donated by the VOIP company Skype, connect translators to individuals who needed language support during the disaster (See Figure 3-48).

Figure 3-48: Image of website where translation service was provided by Babelverse Source: http://blog.tropo.com/2011/03/18/babelverse-using-the-phone-for-real-time-japanese-translation-for-crisis- workers/ Another example of special needs support is “Tezukuri shuwa nyuusu,” a “homemade” Japanese sign language broadcast created during the disaster (See Figure 3-49). When the Fukushima Daiichi Nuclear Reactor began leaking radiation, all television stations were streaming live updates and press conferences from the government. At this point in time, it was not known how large the threat was and many thought that Tokyo was at great risk, so there was reason to be constantly aware of the situation. Unfortunately, during this period, there was no sign language service provided on television and many deaf individuals began complaining on social media that they were being left behind. Due to this, a few individuals, such as a professor and a magazine editor who began communicating through Twitter, with the help of many other people, created a sign language broadcast sent over a Japanese video website, Nico Nico Douga. Over 13,000 people watched the broadcast online (“Ima, watashitachi,” 2011). The efforts of the people involved combined with the social media tools available made it possible for those deaf individuals who were left in out of the loop of traditional mainstream media to understand the current risk to them.

Figure 3-49: An image of the "homemade" sign language broadcast on the program "Close Up Gendai" on NHK Source: http://cgi4.nhk.or.jp/gendai/kiroku/detail.cgi?content_id=3022

Another organization, ilove.cat, used social media to help find pets that were lost in the disaster (See Figure 3-50).

Figure 3-50: An image of the website of ilove.cat, a social media powered website that finds lost pets after the disaster Source: http://ilove.cat/ja/2379 The types of information shared through social media during the disaster depends on the tools used, the goal of the user and the level of experience of the user. Tools that are commonly used as private communication such as email or SMS are likely to have safety confirmation information and possibly information related to goods provision or housing provision. Public methods of information sharing such as social networks and webpages are more likely to have, primarily, general disaster information, fundraising information, moral support information. In addition, these public methods of communication are also used to exchange other types of information for directly affected individuals, such as safety conformation, infrastructure status information, housing provision etc., but this type of information, which is specific to affected individuals, is exchanged at a lower amount compared with other general disaster information due to the fewer number of people needing the information specific to affected individuals as compared with the general populace. The following table is a representation of the dominant information types shared through tools used with social media (See Table 3-5).

Table 3-5: Table of dominant information types shared through tools Webpages or Smartphone Twitter Facebook Mixi SMS Email Wikis blogs apps Maps General disaster O O O O O O O O O information Safety O O O O O O conformation

Fundraising O O O O O

Infrastructure status O O O O notification/regio nal facility status Housing O O provision Goods provision O O O

Moral support O O O O O

Resource saving O O O O Volunteer O O O O O recruitment Special needs O support

3.2.6 Misinformation

Social media was used to exchange information during the disaster by individuals to receive vital information related to the disaster and by some to help in the relief effort with, arguably, positive results in many cases, yet at the same time social media helped to cause confusion in some cases due to incorrect information or purposefully spread misinformation.

Social networks such as Twitter and Facebook and other social media can be helpful in disaster, but due to its public nature and lack of a system for verification of information, rumors, lies and other type of information can be spread through it just as any other type of information. According to a study by Acar and Murakami (2011) on the use of Twitter in the 2011 East Japan Earthquake and Tsunami, in which the researchers tracked the tweets of individuals in the affected area two weeks after the disaster, unreliable retweets on Twitter was the biggest problem that users of Twitter had during the disaster because the users repeated inaccurate information. In particular, they found that there were calls for help that were misplaced or lies.

They also found that many of the users were concerned about the number of rumors. The researchers state that users suggested solutions for introducing official hashtag keywords or adding features that let users trace the sources of information. They find that users should have more responsibility for their tweets, everyone should realize that Twitter is a public communication tool, and information sources should be made clearer in updates. Acar and Murakami also go on to say that further research is need to show if official hashtag keywords during disasters would be able to solve the problems witnessed in this disaster.

In one case of rumors circulating through Twitter, the creator of Pokemon, an animated TV program, was said to have been killed in the tsunami (“Pokemon,” 2011). Other untrue cases of famous persons being killed in the tsunami were also reported (Montenegro, 2011). In another, misinformation saying that, due to a fire near Tokyo, rain would be toxic (“Japan disaster,” 2011).

In Twitter, the most widely used social network during the disaster, it can be difficult to find older information because Twitter is mainly used in the present, so finding past incorrect information can be difficult if not impossible. On the other hand, since it is almost entirely public, just as quickly as incorrect information is spread, it can be corrected. Reliability of information in social networks is definitely a problem as Acar and Murakami show although more research needs to be done in this area.

It is possible that types of social media that have a higher level of anonymity, such as Twitter, have a higher level of misinformation than one with a lower level of anonymity, such as Facebook, which stresses real the use of real names and real world connections. There are reports of some individuals began to use Facebook instead of Twitter to receive information after the disaster, due to misinformation on Twitter (“Japan disaster,” 2011). Though, Facebook also has problems with misinformation as reports of donation scams show (Sherman, 2011). Therefore, research needs to be performed to see if this assumption is true.

Other types of social media also had problems with in accurate information. Google’s Person Finder, which Google created for finding people after the disaster had false reports of the deaths of some individuals, deceiving their families into thinking that they were deceased (“Sick messages, 2011). In addition, during the disaster, chain emails were being sent around with incorrect information about radiation from the nuclear reactors. Due to the fact that email is non-public form of communication it is very hard to combat this kind of misinformation.

References

244,000 foreigners leave Japan in week after quake+. (2011, April 15). Breitbart. Retrieved from http://www.breitbart.com/article.php?id=D9MK40BO0&show_article=1

Aamoth, D. (2011, March 11). Japan Captures Tsunami, Quake Footage with Camera Phones. Time. Retrieved from http://techland.time.com/2011/03/11/japan-captures-tsunami- quake-footage-with-camera-phones/

About. Quakebook.org. Retrieved October 30th, 2011 from http://quakebook.org/about/

Abram, C. (2006, September 26). Welcome to Facebook, everyone. The Facebook Blog. Retrieved from: http://blog.facebook.com/blog.php?post=2210227130

Akimoto, A. (2009, October 17). Japan Public TV NHK Intensely Promotes Twitter. Asiajin. http://asiajin.com/blog/2009/10/17/japan-public-tv-nhk-intensely-promotes-twitter/

Akimoto, A. (2011, June 15). Japan to finally get cross-carrier SMS. The Japan Times Online. Retrieved from http://www.japantimes.co.jp/text/nc20110615ak.html

Analysis and paper presented at TechCamp Jakarta on social media use after the March 11, 2011 earthquake in Japan. TechCampGlobal. Retrieved October 28, 2011 from http://wiki.techcampglobal.org/index.php?title=Analysis_and_paper_presented_at_Te chCamp_Jakarta_on_social_media_use_after_the_March_11,_2011_earthquake_in_J apan

Anxiety in Japan Grows as Death Toll Climbs. (2011, March 14). CNN. Retrieved from http://edition.cnn.com/2011/WORLD/asiapcf/03/13/japan.quake/index.html?iref=NS 1

Birmingham, L. (2011, March, 14) Tokyo: Frustration, Fear and Fierce Resolve. Time. Retrieved from http://www.time.com/time/world/article/0,8599,2058698,00.html

Blackburn, 2011. (2011, March 11). Japan Earthquake and Tsunami Drive Social Media Dialogue - ABC News http://abcnews.go.com/Technology/japan-earthquake-tsunami- drive-social-media-dialogue/story?id=13117677#.TrOJOnHwC9E

Central Intelligence Agency (2009). The CIA World Factbook: Japan. Retrieved from https://www.cia.gov/library/publications/the-world-factbook/geos/ja.html

Damage Situation and Police Countermeasures associated with 2011 Tohoku district – off the Pacific Ocean Earthquake (2011, October 24). Japanese National Police Agency. Retrieved from http://www.npa.go.jp/archive/keibi/biki/higaijokyo_e.pdf

Daishinsai no tsunami, Miyako de 38.9m… Meiji sanriku uemawaru. (2011, April 15). Yomiuri Shimbun. Retrieved from http://www.yomiuri.co.jp/science/news/20110415- OYT1T00389.htm

Derbyshire, D. (2011, March 19) Japan earthquake and tsunami: Fukushima nuclear plant radiation leak could kill people. Mail Online Retrieved from http://www.dailymail.co.uk/news/article-1367684/Japan-earthquake-tsunami- Fukushima-nulear-plant-radiation-leak-kill-people.html

Duke, A. (2011, March 18) Sandra Bullock sends $1 million for Japan. CNN. Retrieved from http://edition.cnn.com/2011/SHOWBIZ/celebrity.news.gossip/03/17/bullock.japan.do nation/

Facebook. (2011). Statistics. Retrieved October 19, 2011 from: http://www.facebook.com/press/info.php?statistics

Fackler, M. (2011, June 1). Report Finds Japan Underestimated Tsunami Danger. The New York Times. Retrieved from http://www.nytimes.com/2011/06/02/world/asia/02japan.html

Fergusson, R. (2011, March 16). Social media sites help loved ones connect after quake and tsunami. E & T Magazine. Retrieved from http://eandt.theiet.org/news/2011/mar/japan-socialmedia.cfm

Fukushima nuclear evacuees to undergo mental health checks. (2011, October 18). The Mainichi Daily News. Retrieved from http://mdn.mainichi.jp/mdnnews/news/20111018p2g00m0dm016000c.html

Gaudin, S. (2011, March 11). Twitter, Facebook become lifelines after Japan quake. Computerworld. Retrieved from http://www.computerworld.com/s/article/9214213/Twitter_Facebook_become_lifelin es_after_Japan_quake

Garrett, B. (2011, March 11). Facebook and Twitter help families during Japan Earthquake and Tsunami. Examiner.com. Retrieved from http://www.examiner.com/social- media-in-knoxville/facebook-and-twitter-help-families-during-japan-earthquake-and- tsunami

Glionna, J. M. (2011, March 16). Japan nuclear crisis: Radiation fears boost sales of surgical masks in Tokyo. Los Angeles Times. http://articles.latimes.com/2011/may/16/world/la-fg-japan-surgical-masks-20110515

Go setsuden purojekuto wo kaishi shimashita. Google Japan Blog. Retrieved from http://www.nytimes.com/2011/09/26/opinion/in-japan-the-summer-of-setsuden.html

Haydon, J. Red Cross amps up donations through iTunes. Retrieved October 30th, 2011 from http://www.johnhaydon.com/2011/03/red-cross-amps-up-donations-through-itunes/

Hornyak,T. (2011, March 11). Twitter, Facebook vital as Japan cell networks jam. CNET. Retrieved from http://news.cnet.com/8301-17938_105-20042108-1.html

Hosaka, T. (2011, April 9). Tohoku Disaster Sparks Innovative Social Media Responses. The Japan Times. Retrieved from http://search.japantimes.co.jp/mail/nn20110404f2.html

Idei, Y. (2011, March 16) Smoke, Confusion And Fear As Japanese Nuclear Crisis Mounts. Forbes. Retrieved from http://www.forbes.com/sites/yasidei/2011/03/16/chaos-at- japan-nuclear-plant-ends-in-evacuation/

Ima, watashitachi ga dekiru koto. (2011, March 29). [Television broadcast] NHK Close Up Gendai. Retrieved from http://cgi4.nhk.or.jp/gendai/kiroku/detail.cgi?content_id=3022

In Japan, the Summer of Setsuden [Editorial]. The New York Times. Retrieved from http://www.nytimes.com/2011/09/26/opinion/in-japan-the-summer-of- setsuden.html?_r=1

Jacobson, K. (2011, March 12). Social media proves to be reliable communication in a time of crisis. NWCN.com. Retrieved from http://www.nwcn.com/home/?fId=117811539&fPath=/news/local&fDomain=10227

Japanese Twitter Users. (2009, July 7). Shibuya 246. Retrieved from http://shibuya246.com/2009/07/07/japan-twitter-users-gachapin/

Japan disaster sparks social media innovation. (2011, April 1). Dawn.com. Retrieved from http://www.dawn.com/2011/04/01/japan-disaster-sparks-social-media- innovation.html

Japan Facebook Statistics. Socialbakers. Retrieved November 4th, 2011 from http://www.socialbakers.com/facebook-statistics/japan

Japan initiates emergency protocol after earthquake. (2011, March 11). Nuclear Engineering International. Retrieved from http://www.neimagazine.com/story.asp?sectioncode=132&storyCode=2059127

Japan: iodine supplies dwindle amid radiation fears. (2011, March 16). Channel 4 News. Retrieved from http://www.channel4.com/news/japan-iodine-supplies-dwindle-amid- radiation-fears

Japan's PM urges people to clear 20-km zone around Fukushima NPP (Update-1). (2011, March 15) RIA Novosti. Retrieved from http://en.rian.ru/world/20110315/163008635.html

Jishin – Mikushi, Mixi no jishin kanren komyunitii matome peeji wo kaisetsu. (2011, March 13). Livedoor News. Retrieved from http://news.livedoor.com/article/detail/5410243/

Johnston, J. (2011, April 10). The Moment The Tsunami Hit Fukushima Nuclear Power Plant [Video]. The 9 Billion. Retrieved from: http://www.the9billion.com/2011/04/10/moment-japan-tsunami-hit-fukushima- nuclear-plant/

Kessler, S. (2011, March 15). How Facebook Users Reacted to the Japan Earthquake & Tsunami. Mashable. Retrieved from http://mashable.com/2011/03/14/how-facebook- users-reacted-to-the-japan-earthquake-tsunami-animated-graphic/

McCurry, Justin (2011, 11 March). Japan Declares 'Nuclear Emergency' after Quake. The Guardian. Retrieved from http://www.guardian.co.uk/world/2011/mar/11/japan- declares-nuclear-emergency-quake

McMahon, F. (2011, July 2). Twitter map shows traffic surge after Japan tsunami. Storyful Retrieved from http://storyful.com/stories/1000005029

Millions Of Stricken Japanese Lack Water, Food, Heat. (2011, March 14). NPR. Retrieved from http://www.npr.org/2011/03/18/134527591/millions-of-stricken-japanese-lack- water-food-heat

Mobile Marketing Data Labo. (2011). Higashinihon daishinsaigo no sousharu media riyou nikansuru jitai chousa [Data file]. Retrieved from http://mmd.update.ne.jp/news/detail.php?news_id=799

Montenegro, C. (2011, March 14). Japanese stars survive quake, tsunami — but not Twitter. GMA News. Retrieved from http://www.gmanetwork.com/news/story/215237/scitech/japanese-stars-survive- quake-tsunami-but-not-twitter

Nations, Companies advising people to evacuate Tokyo. (2011, March 16). 11 Alive. Retrieved from http://www.11alive.com/news/local/story.aspx?storyid=182963

Nishimura, K. (2011, July 25). Pray for Japan. Collabo311. Retrieved from http://www.collabo311.com/2011/07/297/

Parashar, S., Uy, N.. Nguyen, H., Fernandez, G., Mulyasari, F., Joerin, J. & Shaw, R. (2011, 25 March). Mega Disaster in a Resilient Society: The Great East Japan (Tohoku Kanto) Earthquake and Tsunami of 11th March 2011.Kyoto University. International Environment and Disaster Management.

Personal Messages from Japan. (2011, June 29). Twitter [via Twitter’s Flickr photo sharing account]. Retrieved from http://www.flickr.com/photos/twitteroffice/5885172082/

Peters, M (2011, March 17). Japan’s Prime Minister Launches English-Language Twitter Account for Quake Updates, Mashable. Retrieved from http://mashable.com/2011/03/16/japan-prime-minister-english-twitter/

’Pokemon’ creator Satoshi Tajiri rumored to have died in Japan earthquake. (2011, March 12). International Business Times. Retrieved from: http://www.ibtimes.com/articles/121988/20110312/pokemon-creator-santhoshi-tijari- death-hoax-rumored-death-japan-earthquake-tsunami-facebook-twitter.htm

Saigaiji no cheenmeeru hasshinsha tsuiseki ni tsuite. Yahoo Japan Chiebukuro. Retrieved October 30, 2011 from http://detail.chiebukuro.yahoo.co.jp/qa/question_detail/q1257859667

Saigai you dengenban. (2011, March 24). Softbank [via the iTunes App Store]. Retrieved from http://itunes.apple.com/jp/app/id425650996?mt=8

Shea, A. (2011, March 14). Japanese Students In Boston Connect With Ravaged Homeland. WBUR. Retrieved from http://www.wbur.org/2011/03/14/local-japanese-reax

Shepard, W. (2011, March 18). How to Use Social Media and the Internet in Disaster Response. Vagabond Journey. http://www.vagabondjourney.com/travelogue/how-to- use-social-media-and-the-Internet-in-disaster-response/

Sherman, D. (2011, March 14). Facebook scammers take advantage of Japan disaster. KTVQ.com. Retrieved from http://www.ktvq.com/news/facebook-scammers-take- advantage-of-japan-disaster/

Shreffler, A. (2011, June 30). Crowdsourcing a crisis: mapping radiation levels in Japan. The Knight Foundation Blog. Retrieved from http://www.knightfoundation.org/blogs/knightblog/2011/6/30/crowdsourcing-crisis- mapping-radiation-levels-japan/

Sick messages on Google people finder leave Japan earthquake families distraught. (2011, March 14). Metro.co.uk. Retrieved from http://www.metro.co.uk/news/858045-sick- messages-on-google-people-finder-leave-japan-earthquake-families-distraught

Smith, C. (2011, March 23). Twitter User Statistics Show Stunning Growth. The Huffington Post. Retrieved from http://www.huffingtonpost.com/2011/03/14/twitter-user- statistics_n_835581.html

SMS (Short Message Service). NTT Docomo. Retrieved October 29th, 2011 from http://www.nttdocomo.co.jp/english/service/world/roaming/outline/content/sms/index .html

Spread of Information. (2011, June 29). Twitter. [via Twitter’s Flickr photo sharing account]. Retrieved from http://www.flickr.com/photos/twitteroffice/5884626815/

Stephens, K. (2011, March 12). Social Media and the Japan earthquake: What we can learn. idisaster 2.0. Retrieved from http://idisaster.wordpress.com/2011/03/12/social-media- and-the-japan-earthquake-and-tsunami-what-we-can-learn/

Tabuchi, H. (2010, May 18). Softbank to Help Twitter Get Even Bigger in Japan. The New York Times. Retrieved from http://www.nytimes.com/2010/05/19/technology/19twitter.html

Tabuchi, H. (2011, January 9). Facebook Wins Relatively Few Friends in Japan. The New York Times. Retrieved from

http://www.nytimes.com/2011/01/10/technology/10facebook.html?_r=1&scp=1&sq= mixi&st=cse

Tabuchi, H. (2011, July 10). Quick Action Helps Google Win Friends in Japan. The New York Times. http://www.nytimes.com/2011/07/11/technology/quick-action-helps-google- win-friends-in-japan.html?_r=1

Timeline: Japan power plant crisis. (2011, March 13). BBC News. Retrieved from http://www.bbc.co.uk/news/science-environment-12722719

The 2011 off the Pacific coast of Tohoku Earthquake –Portal-. The Japan Meteorological Agency. Retrieved October 23, 2011 from http://www.jma.go.jp/jma/en/2011_Earthquake.html

Touhoku chihou taheiyouoki jishin ni kanshite. (2011, March 12). Twitter Blog. Retrieved from http://blog.jp.twitter.com/2011/03/blog-post_12.html

Veraeck, A., & Daniell, J. (2011, October 2). Japan Tohoku earthquake and tsunami: CATDAT 41 report. Earthquake-Report. Retrieved from http://earthquake- report.com/2011/10/02/japan-tohoku-earthquake-and-tsunami-catdat-41-report- october-2-2011/

Chapter 4 Assessment of the Role of Social Media in the 2011 East Japan Earthquake and Tsunami

4.1 Methodology

This research was conducted through a two-part process consisting of information collection of examples of use of social media in the 2011 East Japan Earthquake and Tsunami and an online questionnaire survey (See Table 4-1). The process of information collection was conducted immediately following the disaster on March, 11th 2011 until the start of the survey on September 13th, 2011 through identifying groups and individuals that used social media during the disaster and examples of social media use in the disaster. Over 200 instances of individual or group use of social media in the disaster were collected in the form of articles, group webpages, first hand accounts in social media, television programs, etc. These examples were later analyzed to discover elements related to social media use in the disaster, such as types of software, organization methods, types of information exchanged, etc., and the results were used to design the online questionnaire survey. An overview of the collected examples, presented by each element, are located in the section titled “Overview of social media use in the disaster” in Chapter 2.

Following this, an online questionnaire survey targeted at social media user was held to collect opinions of individuals who utilized social media during the disaster. The online questionnaire survey was distributed through social networks in Japanese and English to individuals who may or may not have used social media during the disaster. The survey was available for response over a one month period from September 13th, 2011, to October 13th, 2011. The survey was made available to all regardless of level of use, location or level of affectedness. In total, there were 1,538 views of the survey, 156 respondents in Japanese, and 50 respondents in English.

Table 4-1: Table outlining research conducted Types of research Time period Description of the research conducted Information March 11th, 2011 – Identification of examples of social media collection of September 13th 2011 use in the disaster, groups, tools, methods. examples of social Used to design the framework for the online media use in the questionnaire survey. Over 200 collected disaster examples, such as articles, group web pages, personal accounts, etc. Online September 13th, 2011 – Aims to understand information, questionnaire October 13th 2011 motivations, problems, evaluation, survey with social expectation related to social media use in media users the disaster. Distributed to groups related to the disaster by email and made open to interested parties through social networks. Open to all. Has pathways for all types of users. 1,538 views and 206 respondents.

The reason for this two-part process is that, in order to create a questionnaire survey that is well informed, the ways in which social media was utilized during the disaster needed to be clearly understood. Given that there are many examples of social media use in the disaster and there are many tools and methods related to its use, it was necessary to create a method for organizing all of these examples to give a well-rounded view of social media use in the disaster. This was accomplished by identifying each of the elements that each example is made up of or utilized, such as tools, methods, etc., and then grouping these in a logical structure (See Table 4-2).

After organizing and identifying examples of social media use in the disaster, it became clear that there were clear differences in the use of social media in the disaster by individuals who were directly affected and those who were not. There were also clear differences between casual users and users who cooperated in some manner to accomplish a some sort of goal. Therefore, it followed that if the online questionnaire survey was to be able to handle each of these types of possible respondents, there needed to be a method of organization for types of uses. From this, a framework of user roles for possible respondents, which reflects all user roles in the disaster, was created (See Figure 4-1 and Figure 4-2).

The objective of the research is to understand the role of social media in the disaster, and therefore it was clear that the experiences and opinions of each of these types of users would need to be measured, so the structure of the online questionnaire survey was based on and divided by these user roles. Since each of these groups seemed to have very different experiences and motivation, a method for capturing these experiences and the opinions of each group was created. This was done by designing a series of questions aimed at understanding types of information exchanged through social media, motivation in using social media during disaster, challenges and problems associated with using social media during the disaster, evaluation of the usefulness of the use of social media in the disaster, and future potential for social media in disasters. These are in simple terms: information, motivation, problems, evaluation, and expectation (See Table 4-3). Each of the steps in this process are described in the following sections.

4.1.1 Collected Examples of Social Media Use in the Disaster

Examples and other information related to social media use during and after the disaster were collected and organized to find patterns of usage and identify elements of social media use in the disaster. The identified elements are related to services and software used (tools), organizational methods, group use, and goals of utilization. These identified elements were used to design a framework for creating the online questionnaire survey’s questions and answer sets. The table below is a list of the aspects and their identified elements (See Table 4-2). An detailed explanation of each type of the collected examples and one or a few representative examples of each type, presented by each element, are located in the section titled “Overview of social media use in the disaster” in Chapter 2.

Table 4-2: Identified elements in relation social media in the disaster Elements related to social media in the Groups disaster Social networks • Twitter • Facebook Services and software used • Mixi SMS Wikis Webpages or blogs Smartphone applications

Maps Crowdsourcing • Crisis-mapping Social-network-based group type Organizational methods of the software or Info-list type services used Aggregator type Innovation type Design type Raw-social-media-data-mining type Large organization type • Mass-media type Group or organization oriented social media • Direct-relief type use Government-social-media type

General disaster information Safety confirmation Fundraising Infrastructure status notification/regional facility status Housing provision Goals of utilization (information types) Goods provision Moral support Resource saving Volunteer recruitment Special needs support Safety conformation

4.1.2 Analysis Framework

To analyze the results of the online questionnaire survey, a framework for analysis was created. The framework consists of user roles, factors (general areas to be understood, such as motivation), parameters (specific questions that fall under the factors), and indicators (specific aggregated answers of question response sets).

4.1.2.1 Factors, Parameters, and Indicators

A framework of factors, parameters, and indicators related to social media use in the disaster was developed. The factors identified are information exchanged through social media, motivation of users of social media to receive and send information through social media, problems encountered when receiving or sharing information, evaluation of social media as a tool for sharing disaster information, and future expectations in terms of improvements and potential, in simple terms: information, motivation, problems, evaluation, expectations (See Table 4-3).

For each of the factors to be understood, a set of parameters were developed. The parameters are made up of individual questions or sets of questions which seek to understand or measure that particular issue related to their factor. Indicators are aggregated answer sets of response options for the questions that were created out of information from the collected examples (See Table 4-4). Question sets from which the indicators are made from have been repeated throughout the survey for simplicity in the analysis process. Since the indicators are response options, by measuring the rate at which a particular option is chosen and comparing these with the other options, it is possible to get a understanding of respondents experiences or opinions about particular parameter. With the combined information provided by the parameters it is possible to get a holistic view of the major identified factors. The groupings of factors parameters and indicators are listed below under

their corresponding factors (See Table 4-3). For parameters which have “N/A” as an indicator, this means that the parameter does not have an repeating aggregated answer set. These questions have “yes” or “no” answers, scale answers, or specific, non-grouped answers. These parameters simply state their specific answers in their corresponding graphs. In the analysis process, the parameters are measured for each of the user roles.

Table 4-3: Factors making up the framework for analysis and their corresponding parameters Factors Parameters Indicators (aggregated repeating answer sets) Information Most relied on method of information Sources of information acquisition Perceived most viable method of information Sources of information acquisition Types of information received Types of information Tools utilized Types of tools Sources of information Types of information Motivation Reasons for utilization for receiving Reasons for utilizing social media Reasons for sending information Reasons for utilizing social media Problems Results of utilizing and sending information Problems in sending information through social media Types of information not able to receive Types of information Problems with aims N/A Evaluation Level of perceived benefit during the disaster N/A Overall impression of benefit of social media N/A use in disasters Level of success of social media utilization N/A Level of desire to use again N/A Expectations Areas to improve Sources of information

Expectation of government utilization N/A

Table 4-4: Table of indicators Indicators Aggregated Actual options in question sets groups for analysis Information Sources of info Word of mouth Radio Internet Loudspeaker TV Social media Newspaper Email Types of General News on the disaster situation information Official announcements Information on how to donate money Moral support from others Information in order to volunteer Safety Evacuation information Threat information Conformation of the safety of a friend Conformation of the safety of a family member Special Information for individuals with special needs Necessities Information on how to receive goods such as food and water Information on shelters Information on temporary housing as well as other forms of housing Information on supermarkets or locations to buy goods Information to receive support from volunteers Infrastructure Information on utility outages Information on infrastructure damage like bridges Information on public transportation Information on regional communal facilities

Types of Tools Social networks Twitter Facebook Mixi Video sites Ustream YouTube Mobile devices Mobile phones Smart phones (i.e. iPhone and Android) Crowdsourced sites Any Wiki A map service (like Google Maps or Ushahidi/sinsai.info) Web pages Email Original Government Local government sources of Prefectural government information Federal government Traditional mass media Newspapers TV First hand accounts via social media or the Internet Motivation Reason for Technological barrier Other media such as radio and television was not accessible because of location utilizing social Other media was not accessible because of lack of electricity media Ability barrier Other media was not accessible because of language problems Other media was not accessible because of a physical disability such as deafness Convenience Social media is mobile Social media is more convenient Social media allows for direct connection friends and family Social media is quicker than other forms of information Quality of info Social media has more accurate information Social media allows for direct connection to trusted sources of information Prior experience User experience with social networking User has experience with volunteering for groups that provide relief services User has experience with volunteering for groups that provide relief services that use social networking Mass sending ability Used social media because of ability to send information on one's own Used social media because of ability to receive replies to information one has sent out Desire to help Felt "just needed to do something" after the disaster Recognized a need and thought could help in that area Encouraged by someone Problems Problems Trustworthiness of The information was not accurate related to information I could not trust the source of the information information in Utilization related to I had trouble understanding how to use software or websites or I didn't know how social media skill to use it There was so much information that I did not know how to use it Failure in the goals of I could not contact the people I wanted to utilization I could not get the information I was looking for There was no response to the information/requests I put out Lack of electricity Lack of electricity made it difficult to use Problems in Information put out was used sending Information put out was not used information Information put out had little impact through social No problems media

4.1.3 Online Questionnaire Survey

As previously stated, the online questionnaire survey was distributed through social networks in Japanese and English to collect opinions of individuals who utilized social media during the disaster. The survey was available for response over a one month period from September 13th, 2011, to October 13th, 2011. The survey was made available to all regardless of level of use, location or level of affectedness. In total, there were 156 respondents in Japanese and 50 respondents in English. The survey targeted social media users to find individuals who utilized social media during the disaster. The general public, individuals specifically in affected areas, or the general populace of the affected areas, which consist of a high rate of older individuals, were not specifically targeted by this questionnaire.

Table 4-5: Table outlining the online questionnaire survey conducted Details Topic Use of social media in the 2011 East Japan Earthquake and Tsunami Method of Distributed to relevant groups through email and made available by distribution open-invitation through social networks Twitter and Facebook Target audience Anyone, users of social media during the 2011 East Japan Earthquake and Tsunami in particular Time period September 13th, 2011 – October 13th 2011 Language Japanese and English Views Over 1500 Responses 206 (156 in Japanese. 50 in English) Aims Aims to understand information, motivations, problems, evaluation, expectation related to social media use in the disaster.

Further sections explain the organization of the survey. The full questionnaire of the survey is attached in the appendix.

The online questionnaire survey was made available in English and Japanese as two separate surveys that were later combined. All links made available to possible respondents lead to a page with English and Japanese, so respondents were able to respond in whichever language they preferred. The assumption is made that respondents that replied in Japanese are from Japan or have a extremely high level of Japanese language ability. It is also assumed that if a respondent was in Japan at the time of the disaster and was directly or indirectly affected that he or she may have difficulty receiving disaster information in Japanese, therefore responding language is given greater weight than nationality, which was not measured by this survey.

4.1.3.1 Types of Respondents to the Questionnaire Survey

User roles

Through collection of examples of social media use in the disaster general types of users roles were identified. Broadly, sender and receivers were identified. More specifically, the roles non-sender, basic sender, volunteer sender, and manager sender were identified.

The following is an explanation of the logic behind the identification of these user roles. When considering the motivations of individuals related to each of the collected examples of social media use in the disaster, it was necessary to understand what that user had done during the disaster. Through comparing and categorizing types of users it became clear that there were basic roles of receiving or sending information. In other words, all the observed actions of users in the collected examples could be simplified into either receiving information or sending information or both. Therefore, firstly, the roles of receivers and senders were identified.

Receivers are individuals that have joined social networks and used social media to receive information during the disaster. Senders are individuals who send out information through social networks and social media, and many times have a higher level of social media use. The assumption was made that if an individual was acting as a sender by sharing information, that they must be a receiver also, simply because the individual is using a social networking service and because much of the information that is sent out by the individual is information that was originally sent through social media. In other words, it is hard to imagine that an individual would use social media to send out first-hand information from non-social media sources without at the same time seeing what information was being shared through social

media. Therefore, senders also have a role as revivers. Following this logic, any individual that used social media is first and foremost a receiver and then if so inclined takes on the role of sender.

After identifying these two roles, it also became clear that there were different levels of participation for senders. There were individuals who shared information with family or friends, others who worked within groups or organizations, and yet others who designed or created or managed these groups. After comparing and categorizing all the collected examples, three types of senders were also identified: basic sender (sending on an individual basis), volunteer sender (sending through cooperation with others), manager sender (sending while leading a group volunteer senders). The term non-sender is used to differentiate receivers who did not become senders from other senders.

Therefore, every individual that utilized social media has a receiver-sender relationship, such as receiver-basic sender, receiver-manager sender, or receiver-non-sender. Throughout this thesis, the term receiver includes all senders types, as senders also first receive information. The following concept diagram is describes these relationships (See Figure 4-1). All roles are in reference to the use of social media during the disaster. (i.e. the term non-user refers to a person who did not utilize social media during the disaster not a person who does not use social media in general.)

Figure 4-1: Concept diagram of relationship of types of users of social media use in the 2011 East Japan Earthquake and Tsunami

To more clearly understand the relationships and how they are utilized in the analysis of the data from the online questionnaire survey, a further diagram showing the breakdown of user roles and the numbers of the respondents for those roles is included below (See Figure 4-2). The diagram shows the relationship between the user roles and the aggregated groups which they fall under. These user roles and aggregated groups are referred to throughout the analysis of results.

In the diagram, opaque boxes represent the final user role respondents can fall under: non- sender, basic sender, volunteer sender, and manager sender. The box with the solid line represents non-users, who are given less priority due to the fact that they did not utilize social media during the disaster. Boxes with dotted lines represent the aggregated group, combined amounts of user roles, for example group senders are the combined amounts of volunteer senders and manager senders. A further example is senders, which is a combination of the group senders and basic senders.

Figure 4-2: Concept diagram of types of users of social media use in the 2011 East Japan Earthquake and Tsunami, their larger groups and numbers of respondents for those groups

Reflecting the framework for analysis, the questionnaire was structured to measure the factors of motivation, problems, evaluation and expectation by user role. The following figure is a graphical representation of the theoretical and functional structure of the online questionnaire survey that combines these factors and the user roles mentioned here (See Figure 4-3).

Figure 4-3: Theoretical structure of the online questionnaire survey

4.1.3.2 Layout of the Questionnaire

In order to handle individuals with multiple experiences and roles while using social media in the disaster, a pathway structure was created (See Figure 4-4) that, through questions posed to determine user type, directed respondents to pages for their specific user type. This multipage system has the advantage of allowing respondents to see only questions that apply specifically to their situation and by making the survey able to handle any type of respondent, even ones that did not use social media at all.

Figure 4-4: Pathways layout in the online questionnaire survey

4.1.3.3 Elements of the Questionnaire

In addition to the framework of factors, parameters, indicators, and user roles other elements were incorporated. These elements aid in the analysis of the data obtained from the online questionnaire survey. The elements are level of affectedness and group information.

Level of affectedness

For identifying level of affectedness, the following categories were created:

• Directly affected (Individual personally felt the earthquake, to a strong degree or had to evacuate due to the tsunami)

• Indirectly affected to a high degree (Personally not affected, but family or friends were affected AND the individual was worried for his or her safety because of aftershocks or radiation

• Indirectly affected to a lower degree (Personally, not affected, but family or friends were affected AND the individual was NOT worried for his or her safety because of aftershocks or radiation)

• Not affected (Individual kept informed via media or social media and or helped out in someway via the Internet)

It is imagined that within the user types, users’ level of affectedness will determine what types of information are exchanged and what the needs of that individual are.

In addition, Individuals who were identified as directly affected were also asked a follow up question asking which of the following events they were affected by:

• Affected by the earthquake

• Affected by the tsunami

• Affected by the nuclear plant disaster

Group information

Questions related to groups were also created to understand the goals of groups, when in relation to the disaster were the groups created, the number of individuals in the groups, number of people in the groups, in addition to questions posed to volunteer senders and managers senders on the factors of information, motivation, problems, evaluation, and expectation, in relation to the groups.

4.1.3.4 Method of Distribution

The online questionnaire survey was performed by creating a Japanese and English version of the questionnaire using the Form function of Google Documents. A link to the surveys with a description of the research was published on an online blog that represented the survey (See Figure 4-5) and acted as an aggregation point for individuals interested in taking the survey. Web links were then distributed using email and social networks, Twitter (See Figure 4-6) and Facebook (See Figure 4-7). Initially, groups and individuals identified as having utilized social media, totaling over 30 individuals and organizations, were contacted and asked to distribute the survey to their members and followers on the social networks. Following this, general appeals were made through Twitter (See Figure 4-8) and Facebook for respondents to the survey. Finally, throughout the one month period that the survey was active, Twitter was used to perform searches using keywords related to the disaster, and user accounts recognized as having a relation to the disaster were targeted with @replies (public messages using a user’s name) asking for responses and help spreading the questionnaire survey. The blog, which acted as an aggregate location for the survey, was visited 1,538 times over the one month period.

Figure 4-5: Image of the blog used to distribute the online questionnaire survey

Figure 4-6: An Image of the Twitter account for the online questionnaire survey

Figure 4-7: Image of the Facebook Page for the online questionnaire survey

Figure 4-8: Image of tweets sent out through Twitter soliciting responses to the online questionnaire survey

4.1.3.5 Limitations of the Study

The study has several limitations related to design of the online questionnaire survey, issues related to understanding of concepts associated with social media, technical limitations of the tools utilized in distribution of the survey, and scope of the survey.

In terms of limitations related to the design of the study, the survey was distributed through social media to ensure that individuals who utilized social media in the disaster responded, but the result is that a low level of individuals who experienced the disaster and did not use social media answered. The survey could be improved by receiving equal amounts of individuals who experienced the disaster and used social media and those who did not. In addition, since the online questionnaire survey asked individuals to respond to questions on the use of social media in the disaster, individuals who did not use social media in the disaster were less likely to reply. If the survey was presented as a survey on communications or media in the disaster and then asked respondents about social media, there might be more responses from individuals who have less experience or less interest with social media.

In terms of limitations related to understanding in the study, questions in the survey ask users if they used social media during the disaster, but the time period of “during the disaster” was not clearly defined. This may have lead to confusion for respondents and therefore this is a limitation of this survey. In addition, due to the ever-changing and interpretational understanding of what social media is and the lack of distinction between the Internet itself and social media, questions that ask users to choose between “the Internet” and “social media.”

In terms of technical limitations of the study, the service utilized for distribution of the online questionnaire survey, Google Documents, does prevent multiple replies from the same IP address, therefore it is possible for an individual to answer the survey multiple times from the same computer, so this is a limitation. In addition, there is no system in place in Google Documents to check if respondents are actually human and so hypothetically responses could be generated by a computer program, although this is unlikely.

In addition, due to the fact that the survey was answered by interested parties, the survey is not representative of affected individuals as a whole or of as social media users as a whole. Further research to show the opinions of these groups, not just interested parties, is necessary.

Finally, the decision was made to target individuals who utilized social media during the disaster in order to collect information on these individuals experiences and opinions. Due to this, the opinions of individuals who did not use social media, the general public at large, individuals in the affected areas during the time of the disaster, and older individuals, who

reside in large numbers in the affected area, were not targeted in this research. The reason for this decision was that it was thought that it would be difficult to distribute and receive a viable about of responses from individuals who actually used social media in the disaster to produce valid findings. Further research that includes the groups not included in this survey would help to put this research in better context.

4.2 Survey Results

4.2.1 Results by user role

4.2.1.1 Attributes of Respondents Based on the framework for analysis in which responses are divided by user role, analysis has been conducted. Therefore, the following results are divided in this manner. Due to the nature of user roles, amounts naturally vary between user roles: there are likely to be fewer volunteer senders than there are basic senders because being a volunteer sender requires time, skill and awareness. Such is also the case for manager senders: there are likely to be less manager senders than there are volunteer senders because being a manager sender requires more skill with social media-based technologies. Along with this difference in number in user roles, there is also a difference in the weight of the responses of different user roles. The weight of manager senders’ responses is high because of their experience and insight. Even though the number of respondents that are manager senders is only 5, this low amount was expected due to the fact that this role has a higher level of responsibly and takes a higher level of motivation. Therefore, when considering the low number of manger sender respondents, the weight of their opinions due to their experience and responsibility should be taken into account. The number of manager senders is 5 out of a total of 205 respondents, to a questionnaire survey that was distributed through social media, which is likely to have a higher number of manager senders as respondents than another method of distribution, so it is likely that, in order to increase the number of manager sender respondents, a much greater number of respondents would be necessary. Such is also the case for volunteer senders, at 32 out of 206 respondents. Due to the nature of these roles, the amount of responses by these user groups should be looked at in this context.

Of the respondents, 156 individuals out of 206 responded in Japanese and the remaining 50 responded in English (See Figure 4-9). These percentages make it likely that there were a variety of respondents to so that the data is well rounded in terms of affected Japanese and foreign residents as well as individuals abroad that might have given some type of support.

Respondent language by user role follows a similar patter with the exceptions of Manager senders, which all responded in Japanese, and Non-senders, which had a higher percentage of Japanese language respondents than the average, at around 84% (Figure 4-9).

Figure 4-9: Language by user role

Respondents were primarily between the ages of 26 to 45. There were no respondents below the age of 18. Individuals over the age of 45 made up 19% of respondents. In terms of age by user role, the largest group, Basic senders, has equivalent amounts of 18 to 35 year-olds and 36 to 55 year olds, with a very small amount of individuals 65 and up (Figure 4-10). Volunteer senders have a higher amount of mid aged users (36 to 55) than young users (18 to 35). Manager senders are all mid aged (35 to 55) but one young user (18 to 35). Non-users have a large portion of mid aged users (35 to 55) as well as a fairly large amount of young users (18 to 35). A significant portion of the non-users is also older (56 and up). Non-senders are the only group where young users (18 to 35) is the majority. Of these results what is surprising is that there were no respondents below the age of 18 and young users (18 to 35) were not the majority given social media’s image of primarily being used by young people. The majority of respondents are mid aged (35 to 55) (Figure 4-10).This could be interpreted as mid aged individuals have more of a drive to try to help out through social media than younger people, which is inline with the non-sender’s age distribution. Overall, older individuals use social media at lower rates than other groups according to the data.

Figure 4-10: Age by user role

Data on gender by user role shows that the division between the sexes is nearly equal. Respondents were 50% female and 50% male, a evenly balanced sample, with the exception of manager senders, which are all male except for one female (Figure 4-11).

Figure 4-11: Gender by user role

Respondents were asked for their location during the disaster (See Figure 4-12). For analysis, areas were divided into affected areas (Tohoku region and Kanto region), other areas in Japan and outside of Japan. A large majority of respondents, 88%, were in Japan at the time of the disaster. Around 55% of respondents were in areas affected by one of the three disasters, Kanto and Tohoku, during the disaster.

Non-users have near equal amounts of individuals in other parts of Japan and in the Kanto region followed by users in Tohoku region, with individuals outside of Japan making up the smallest group. The majority of non-senders were in the Kanto area with the second largest portion being from other parts of Japan. The largest group of the basic senders was in the Kanto region, followed by other areas in Japan, and then with similar pre portions in the Tohoku region and outside of Japan. Volunteer senders are primarily from the Kanto area, with the remainders from other areas in Japan, the Tohoku region, and outside of that Japan, in that order. The manager senders are made up of individuals only from the two affected areas. All the user role groups have more than half of their respective respondents from affected areas.

Figure 4-12: Location by user role

A vast majority of respondents described themselves as being an intermediate Internet or social media users (See Figure 4-13). 15% of respondents said they were beginners and 2% said that they did not use social media at all. These results are not surprising as it was expected that a large portion of respondents would be experienced at using social media due to the fact that the survey was distributed to organizations that were identified as having used social media in the disaster and social media was used as the main pathway for distribution.

Level of experience with internet use is self reported by respondents. All user roles have a majority of users that have an intermediate level of use of the Internet and social media (See Figure 4-13). This is consistent with expectations of the author due to the fact that the survey was distributed through social media. Non-users have a large amount of beginner users, as well as some high level users and a few individuals who do not use social media, which none of the other groups have. Non-senders also have a substantial portion of beginner users. Being beginner users might be the reason why these individuals became non-senders i.e. they might have not known how to send information. Manager senders are made up of only intermediate level users and above, which is consistent with expectations of the author due to the skill required to perform the task. Conversely, the largest portion of manager senders are an intermediate level, which was unexpected. This shows that individuals can help to manage systems that use social media to support people affected by disaster with a level of experience that is lower than that of high level.

Figure 4-13: Skill level with the Internet and social media by user role

Respondents were asked as to identity their level of affectedness (See Figure 4-14). 30% of respondents were not affected. Of the remaining 70%, 18% were indirectly affected to a low level, 23% were in directly affected to a high level, and 29% were directly affected.

For non-users, the largest group is of directly affected individuals (See Figure 4-14). This may be due to the fact that many of them were not able to access social media during the disaster due to some technological barrier such as lack of mobile phone reception or they did not use social media in general when the disaster hit and began to use it later. Non-senders have nearly identical portions from all levels of affectedness. Basic senders have similar distributions. Volunteer senders have individuals with all levels of affectedness. Manager senders have equal portions of non-affected and indirectly affected to a high degree,

Figure 4-14: Level of affectedness by user role

In summary, 156 individuals out of 206 responded in Japanese and the remaining 50 responded in English. Users of social media were generally older than imagined. All the user role groups have more than half of their respective respondents from affected areas. 70% of respondents were affected to some degree. Users groups are spread out by level of affected ness at similar rates, excluding non-users who have a high level of directly affected individuals. There is an equal balance between genders except for manager senders. The majority of each user group is from the affected areas of Tohoku and Kanto regions. Most respondents are intermediate users.

4.2.1.2 Source and Type of Information

Respondents were asked which method of communication was the most reliable in their own experience during the disaster. The Internet and social media are divided into separate categories for the reason that many people use the Internet to view webpages such as online news articles but do not participate in the sharing driven sites and services that make up social media. This being so, the distinction between the Internet and social media is not clear in many people’s minds and the lines between these two things are extremely blurry due to the fact that social media exists within the Internet. Therefore, this fact must be taken into account when understanding which sources of information were depended on during the disaster, and, to some extent, these two options can be interpreted as being connected. The difference between the Internet and social media was described for respondents in the survey to try to reduce confusion.

Manager senders found social media to be the most reliable source of information during the disaster, as well as volunteer senders (See Figure 4-15). For these two groups, the Internet also makes up a very large portion of the most relied on source of information during the disaster. For basic senders, the Internet, followed by social media were the most reliable, with the two combined making up the vast majority, which is same for the other groups that share information. Basic senders also rely on TV to a large degree. Groups that do not share information have a larger variety of sources that are viewed as the most reliable. For non- senders, social media, the Internet, and TV have nearly equal amounts, although social media and the Internet combined make up over 50%. Non-users relied on TV the most, followed by the Internet, radio and then word of mouth. A small percentage of non-users found social media to be the most reliable source of information during the disaster. This might be that the users did not personally use it themselves, but perceived it as such through the observed use of others. The results show that the user groups that have a higher level of participation , such as manager senders, the more likely individuals are to find social media or the Internet to be the most reliable source of information during the disaster. Interestingly, email, which many people use with mobile phones, as opposed to SMS, was not chosen at all.

Manager"senders"(n=5)"" 1" 1" 3" Word"of"mouth"

Loudspeaker" Volunteer"senders"(n=32)" 1"1"1" 11" 18" Newspaper"

Basic"senders"(n=100)" 4" 18" 41" 37" Radio" TV""

Non.senders"(n=25)" 1"1" 2" 6" 7" 8" Email""

Internet" Non.users"(n=44)"" 7" 1" 8" 14" 12" 2" Social"media"

0%" 20%" 40%" 60%" 80%" 100%"

Figure 4-15: Most reliable source of information during the disaster

Respondents were also asked, outside of their own personal experience, what do they perceive as the most viable source of information during the disaster (reliable is used to measure their personal experience, whereas viable is used to measure what respondents thought was the most useful source of information, even if they personally did not use it during the disaster) (See Figure 4-15). Manager senders, interestingly nearly all said the Internet, which differed from their personal experience in which social media use was high. Volunteer senders followed a similar pattern to a lesser degree, so that Internet made up over 50%, with social media following a close second. Basic senders, answers were similar to that of the previous question, with the Internet increasing slightly. Non-senders also chose the Internet as the most viable source of information, with social media also making up a large portion of responses. In addition, users who personally relied on the radio the most in Figure 4-15 switched to other sources of information as being most viable in their perception. Non- users answered that the Internet was their perceived most reliable source of information, differing from TV in the previous question. They also chose social media at a higher rate. Word of mouth stayed the same. Overall, the Internet is perceived to be a highly reliable form of communication as well as social media, with the two making up the vast majority for each user group.

Manager"senders"(n=5)"" 4" 1" Word"of"mouth"

Loudspeaker" Volunteer"senders"(n=32)" 1" 18" 13" Newspaper"

Basic"senders"(n=100)" 1"3" 15" 48" 33" Radio" TV""

Non0senders"(n=25)" 1" 6" 11" 7" Email""

Internet" Non0users"(n=44)"" 7" 2" 2" 6" 1" 18" 8" Social"media"

0%" 10%" 20%" 30%" 40%" 50%" 60%" 70%" 80%" 90%" 100%"

Figure 4-16: Perceived most viable source of information for others in the 2011 East Japan Earthquake and Tsunami

Users were asked what type of information they received during the disaster (See Figure 4-17). Answers were multiple choice and the specific options have been aggregated into the groups, general (general information related to the disaster), safety (information related to

making ones self safe during the disaster), special (information for individuals with special needs in the disaster such as deaf individuals or individuals who are not proficient in Japanese), necessities (information on food, water, shelter, etc.), and infrastructure (information on power outages, public transportation, etc.). The graph represents the percentage of individuals in each group that received that specific type of information.

Basic senders and volunteer senders have similar patterns in receiving information. General information comprises the largest group followed by infrastructure and then safety information. Non-senders mostly received information related to safety followed by general information and then infrastructure. Interestingly, they received almost no information on special needs or necessities. Non-senders do not have the need for this information. Since they chose safety information first, but almost no information on necessities, it might follow that they after confirming that they are not threatened did not feel the need to receive more information or to share information. Manager senders received special needs information the most followed by general information, infrastructure information, and necessities. Due to the fact that manager senders operated a group that shared information during the disaster, it is likely that their awareness about the needs of others, the individuals they are attempting to support, is high, and why they have such high rates of received information about special needs, infrastructure, and necessities. Individuals who share information, as opposed to those who do not, seek out information that may not be applicable to themselves at higher rates with the intention of sharing that information.

Figure 4-17: Types of information received during the disaster (multiple choice)

Senders were also given a similar question asking what types of information they shared through social media during the disaster (See Figure 4-18). While patterns are similar overall rates are lower in comparison to received information. In particular, manager senders although receiving a high level of information related to special needs shared that information at a much lower level, second lowest compared with the other types of information. Besides general information, necessities and infrastructure information was the highest, but still the percentages are lower when compared with receiving. Besides general information, volunteer senders shared necessities information at a high rate, showing that they are working to support the needs of directly affected individuals. Basic senders shared special needs information at a relatively lower rate, showing that while motivated to share information, they are not seeking out information to share with individuals that might be at risk, this shows a difference between organized sharing and individual informal sharing.

Figure 4-18: Types of information sent during the disaster (multiple choice)

Respondents were asked what tools they utilized in receiving information (See Figure 4-19). Non-senders rely on primarily social networks and web pages as well as mobile devices to some extent. Basic senders also rely on social networks and webpages primarily, with video sites and email to a lower extent. While volunteer senders also rely on social networks and webpages to higher, and nearly similar rates, other tools seem to be all around the same levels. Manager senders rely on web pages, social networks, and email to the same degree, followed by mobile devices, and then video and crowdsourced sites. Senders share using a greater variety of tools when compared with non-senders. Group senders use crowdsourced sites more than basic senders.

Figure 4-19: Types of tools related to social media utilized during the disaster (multiple choice)

Senders were also asked what were the original sources of the information they received (See Figure 4-20). Manager senders received information from first government sources followed by traditional mass media and first hand accounts at equal rates. Volunteer senders received first hand accounts at the highest rate followed by traditional mass media and then government sources. Basic senders follow this same pattern as well as non-senders. First hand accounts are at similar levels for all groups, but government and traditional mass media are lower in groups that have a lower level of participation. The act of sharing seems to make users aware of the variety of information they and get information from sources that have a high level of authority or validity. Such a high dependence on first hand accounts via social media or the Internet leaves for a high likelihood of untrue or inaccurate information being shared.

Figure 4-20: Original sources of received information (multiple choice)

A following related question was asked of senders about what were the original sources of information that they shared (See Figure 4-21). Basic senders shared information from first hand accounts the most followed by traditional mass media and then government. Volunteer senders also followed this pattern, but manager senders put a high priority on information from government sources and treated information from traditional mass media and second hand accounts about the same. In comparison with the previous graph, non-senders shared more information of all types in a more balanced fashion than what they received. Basic senders shared less information than they received but with a lower amount of difference between the types of information, i.e. the sharing of information was more balanced than receiving. Manager senders having balanced levels of types of information in receiving, shared government information at much higher rates. These patterns show that in the act of sharing information groups increase the level of information from authoritative sources, which have information with a higher level of validity, when compared with receiving.

Figure 4-21: Original sources of sent information (multiple choice)

Respondents were also asked what sources of information they used besides social media (See Figure 4-22). All user role groups had very high levels of TV use. In general, radio and newspaper were much less utilized, except for manager senders who used newspapers in high rates.

100

Figure 4-22: Sources of received information besides social media (multiple choice)

In summary, Social media and the Internet combined make up the vast majority of reliable information sources in the disaster as well as perceived reliable information source for disasters in general for all groups that used social media. Groups that did not share information have a larger variety of sources that are viewed as the most reliable. Non-users relied on TV the most. Email, which many people use with mobile phones, was not mainly relied on by anyone. For less experienced people TV and radio are main forms of receiving information. Non-senders also chose the Internet as the most valuable source of information. Users who said that they radio was the most reliable in their own experience chose other forms as being more reliable for disasters in general. Individuals who share information receive information that may not be applicable to themselves at higher rates. Organized group sharing has a balance of information sought and shared and includes special needs and infrastructure information. Non-senders rely on primarily social networks and web pages. All user groups receive information from first hand accounts are at similar levels, but receiving government and traditional mass media are lower in groups that have a lower level of participation. Manager senders put a high priority on information from government sources and treated information from traditional mass media and second hand accounts about the same. Besides social media, All user role groups had very high levels of TV use. In general, radio and newspaper were much less utilized.

4.2.1.3 Motivation in Receiving and Sending Information

A series of questions on users’ motivations for using social media during the disaster were asked of each user role group. Reasons for using were aggregated into technological barrier (to using some other form of communication), ability barrier (some sort of disability drove users to utilize social media, such as non-Japanese speakers not being able to understand information being shared on TV), convenience, quality of information, prior experience, mass sending ability, desire to help, and encouraged by someone (See Figure 4-23). Non-senders cited convenience most often followed by quality of information. Basic senders cited convenience the most often followed by mass sending ability, quality of information, and then desire to help. Major reasons for volunteer senders followed the same order. Manager senders cited prior experience and mass sending ability followed by desire to help, with convenience being relatively high also. Senders have a relatively high level of prior experience which might explain why they are more likely to share. Manager senders have a high level of prior experience also, so their prior experience would explain why they have a higher level of participation. Desire to help is high among senders and low among non-senders.

101

Figure 4-23: Reasons for utilizing social media to receive information (multiple choice)

Senders were also asked a related question on reasons for using social media to send information (See Figure 4-24). All groups cite convenience most often. Importance of mass sending ability goes up by increased level of participation and as well as desire to help and prior experience.

Figure 4-24: Reasons for utilizing social media to send information (multiple choice)

In summary, convenience and mass sending ability are commonly cited reasons for utilizing social media. Senders have a relatively high level of prior experience. Desire to help is high among senders and low among non-senders.

4.2.1.4 Problems Encountered during Receiving and Sending of Information

Respondents were asked a series of questions related to problems experienced in receiving and sending information through social media during the disaster. Respondents were asked to identify specific problems they experienced during the disaster and responses were aggregated into problems related to trustworthiness of information, utilization related to skill, failure in the goals of utilization (i.e., users were not able to accomplish what they were trying to do with the technology), and lack of electricity (See Figure 4-25). The most common problems amongst all the groups were related to trustworthiness of information.

102

Figure 4-25: Problems in receiving information through social media during the disaster (multiple choice)

As a related question, senders were asked what the out come of their sharing was, whether information was used, not used or had little impact (See Figure 4-26). All groups state that the information that they shared was used in very high rates.

Figure 4-26: Problems in sending information through social media during the disaster (multiple choice)

In summary, the most common problems amongst all the groups were related to trustworthiness of information. Skill issues are a problem but secondary compared with problems with trustworthiness of information. All user groups state that the information that they shared was used in very high rates.

4.2.1.5 Evaluation of the Usefulness of Social Media in this Disaster and in Disasters in General

Respondents were asked a series of questions to measure their perception of the helpfulness of social media during the disaster and disasters in general. Respondents were asked to evaluate the helpfulness of social media to them in the disaster on a level of 1 to 4 with 1 being “not helpful at all” and 4 being “extremely helpful” (See Figure 4-27). Manager senders, although limited in number, all replied extremely helpful. Volunteer senders replied “extremely helpful” for about 65% of people, with the rest choosing “helpful.” Basic senders were over 50% “extremely helpful,” with another 30% of them saying “helpful.” Over half of non-senders replied that it was “helpful” with a large amount also relying “extremely helpful,” Overall, respondents have an overwhelming positive view of the helpfulness of social media

103

in the disaster for themselves. The higher the level of participation, the larger the likelihood that the respondents will answer “extremely helpful.”

Manager"senders"(n=5)"" 5"

Volunteer"senders"(n=32)" 11" 21" Not"helpful"at"all"" Not"helpful"

Helpful" Basic"senders"(n=100)" 1"6" 38" 55" Extrememly"helpful"

Non0senders"(n=25)" 1" 1" 14" 9"

0%" 10%" 20%" 30%" 40%" 50%" 60%" 70%" 80%" 90%" 100%"

Figure 4-27: Level of helpfulness of social media during the disaster

When asked if they thought social media was beneficial in disasters in general, similar responses were given (See Figure 4-28).

Manager"senders"(n=5)"" 1" 4"

Volunteer"senders"(n=32)" 1"1" 13" 17" Not"beneﬁcial"at"all"" Not"beneﬁcial"

Beneﬁcial" Basic"senders"(n=100)" 6" 40" 54" Extrememly"beneﬁcial"

Non/senders"(n=25)" 2" 17" 6"

0%" 20%" 40%" 60%" 80%" 100%"

Figure 4-28: Overall impression of usefulness of social media during disaster in general

In summary, overall, respondents have an overwhelming positive view of the helpfulness of social media in the disaster for themselves and in disasters in general. Respondents with higher levels of participation answer with “extremely helpful” at higher rates.

4.2.1.6 Areas of Information for Improvement and Support for Governmental Use of Social Media in Disasters

Respondents were asked a series of questions on the potential of social media in disasters and their expectations regarding its future use. Respondents were asked what areas need to be improved for sending information in disasters (See Figure 4-29). Manager senders answered special needs information followed by infrastructure and safety. Non-senders also chose special needs information and infrastructure information, incidentally. Basic senders and volunteer senders chose all areas for improvement about equally.

104

Figure 4-29: Areas to be improved for receiving information in disasters (multiple choice)

Users were also asked if they think that governments should use social media in disasters (See Figure 4-30). The graph shows from 75% to 100% with each group having high, over 80%, support for governments using social media in disaster. Support for its use in government goes up with increased participation.

Manager#senders# 4#

Volunteer#senders# 30# 2# Yes#

No# Basic#senders# 90# 10#

Non/senders# 21# 4#

75%# 80%# 85%# 90%# 95%# 100%#

Figure 4-30: Support for government use of social media in disaster

In summary, all groups feel, in general that all areas of information need to be improved in disasters, with special needs information and infrastructure information at slightly higher rates. Also, each group has high, over 80%, support for governments using social media in disaster.

4.2.2 Analysis Comparison by Location and Level of Affectedness

In order to understand the how experience and relationship with social media during the disaster varies based on location during the disaster and level of affectedness, responses were analyzed based on these aspects and the framework which seeks to measure the factors of information, motivation, problems, evaluation and expectation or potential.

4.2.2.1 Location

Responses following the framework measuring the factors was used to analyze the results based on location. Location is specified as the areas affected by the disaster, the Tohoku and Kanto regions, other areas in Japan besides these two areas, and outside of Japan. Responses

105

to the most relied on source of information during the disaster (See Figure 4-31). Although the number of respondents was small (n=31), respondents in the Tohoku region, affected by the earthquake and tsunami, relied upon social media primarily, followed by the Internet, and then word of mouth and TV. Respondents in the Kanto region also relied upon social media to the largest degree, but also on the Internet to a high degree, followed by TV. Respondents in other areas of Japan relied on the Internet for information to nearly 50%, with TV and social media being used to the same degree. Respondents outside of Japan relied upon the Internet the most, followed by social media, with radio and TV taking up equal portions. All areas except for the Tohoku region relied on social media as well as the Internet, followed by TV. Respondents in Tohoku used a greater variety of sources of information, although social media was the most relied upon. This might be due to problems such as lack of electricity and need for immediacy.

Figure 4-31: Relationship between location and most relied on source of information during the disaster

Responses based on motivation were analyzed by location (See Figure 4-32). Although the number of respondents was small (n=31), respondents outside of Japan listed selected mass sending ability and convenience most often for their motivation for using social media during the disaster. Respondents in Japan in other areas than the Tohoku and Kanto regions, chose quality of information most often followed by convenience. Mass sending ability was also chosen to a high degree and then prior experience. Respondents in the Kanto region chose convenience and mass sending ability the most, with quality of info and prior experience coming in at high levels also. Users in the Tohoku region overwhelmingly chose mass sending ability, followed by convenience for obvious reasons of needing to share information with those affected by the disaster. Quality of information, technological barriers and desire to help were also commonly chosen reasons for utilization.

106

Figure 4-32: Relationship between location and motivation

Responses on problems related to receiving information during the disaster were, for all regions except for outside of Japan, was lack in trust of information (See Figure 4-33). This would be expected as incorrect information could cause greater problems than any other issue during the disaster and clearly areas outside of Japan, which are not at risk from inaccurate information would be at less risk. All areas had similar levels of problems with skill related to utilization. Tohoku region and Kanto region had a significant amount of problems related to lack of electricity. Another reason for lack of trust in information could be due to the complexity of the nuclear issue and the unclearness of the reliability of information related to it.

Figure 4-33: Relationship between location and problems in receiving information

In regards to evaluation of the level of helpfulness of social media in the disaster, affected areas, the Tohoku and Kanto region regard social media as “extremely helpful” (See Figure 4-34). Respondents in the Kanto region chose “extremely helpful” more often than users in the Tohoku region. Respondents in the Tohoku region chose “extremely helpful” at over 50% with nearly all the remainder being made up of “helpful.” Other areas in Japan chose “extremely helpful” first, with “helpful” at a similar rate. All regions in Japan had very low amounts of “not helpful” or “not helpful at all.” Only did users outside of Japan not choose “not helpful” or “not helpful at all,” conversely these users overwhelming chose “helpful” as opposed to “extremely helpful.” This is most likely due to the probability of users closer to the disaster having a larger likelihood of receiving some sort of information that could be used in a useful way and therefore being more likely to be grateful for that information, in contrast to users outside of Japan, who have little relation to the disaster but view social media as helpful for learning about the disaster. Actionable information has greater value and therefore is perceived as more helpful.

107

Figure 4-34: Relationship between location and level of helpfulness of social media during the disaster

When asked about the overall benefit of social media in disasters results are similar except there is a slight shift toward a less positive view of social media in disasters (See Figure 4-35).

Figure 4-35: Relationship between location and overall impression of the usefulness of social media during disasters in general

In general, all areas of information that are shared through social media during the disaster need to be improved (See Figure 4-36). For individuals outside of Japan, safety information such as ability to contact ones family and friends is most necessary to improve. For individuals in Japan but outside of the affected areas, special needs information is the area that most needs improvement followed by infrastructure. For individuals in the Kanto region, infrastructure information followed by special needs information were the area that most need improvement. For individuals in the Tohoku region, infrastructure information needs improvement to a larger degree than other types of information, although all areas are high in general. Interestingly, special needs information was chosen at a lower rate. Overall, infrastructure information and special needs information are areas which are seen to need the greatest improvement.

108

Figure 4-36: Relationship between location and areas to be improved for receiving information in disasters

Individuals in the affected areas have a slightly higher instances of believing that government should use social media in disasters than individuals outside of the affected areas, although overall support is high for governments using social media in disasters. Tohoku and Kanto regions chose “yes,” governments should use social media in disasters, at just over 90% and individuals in other areas of Japan and outside of Japan chose “yes” at just over 80%.

Figure 4-37: Relationship between location and support for government use of social media in disasters

In summary, respondents in the Tohoku region relied upon social media primarily, followed by the Internet, and then word of mouth and TV and they used a greater variety of sources of information, although social media was the most relied upon. Respondents in the Kanto region relied upon social media to the highest degree, but also on the Internet to a high degree, followed by TV.

Users in the Tohoku region overwhelmingly state mass sending ability followed by convenience as the reasons for using social media in the disaster. Respondents in Japan in other areas than the Tohoku and Kanto regions, chose quality of information most often followed by convenience. Quality of information, technological barriers and desire to help were also commonly chosen reasons for utilization. Lack in trust of information was the largest problem related to receiving information during the disaster for all regions except for areas outside of Japan.

Users in the Tohoku and Kanto region primarily regard social media as “extremely helpful” in the disaster with nearly all of the remainder regarding it as “helpful.” Other areas in Japan found social media to be “extremely helpful” and “helpful” in the disaster at a similar rate,

109

making up the vast majority of responses. All areas found social media in the disaster to be helpful in extremely high rates, with users in the Tohoku and Kanto regions more likely to find it “extremely helpful.” All areas of information that are shared through social media during the disaster need to be improved according to respondents in all regions. Infrastructure information and special needs information are areas which are seen to need the greatest improvement. Tohoku and Kanto regions chose “yes,” governments should use social media in disasters, at just over 90% and individuals in other areas of Japan and outside of Japan chose “yes” at just over 80%.

4.2.2.2 Level of Affectedness

Responses following the framework measuring the factors was used to analyze the results based on level of affectedness. Factors such as information, motivation, problems, evaluation, and expectation or potential were evaluated from the point of view of level of affectedness. Level of affectedness was divided into directly affected, indirectly affected, and non-affected. Directly affected are individuals that felt the earthquake to a large degree or had to evacuate due to tsunami. Indirectly affected are individuals who had friends or relatives that were or might have been affected by the disaster after it happened. Non-affected had no relation to the disaster.

Responses on the most relied on type of information during the disaster were examined from the aspect of level of affectedness (See Figure 4-38). Non-affected individuals relied on the Internet, TV, and social media in that order, but in nearly equal parts. Indirectly affected individuals relied mainly on the Internet followed by social media, and TV to a lesser degree. Directly affected individuals used a broad range of sources of information, with social media taking up the largest portion, nearly 35%, followed by the Internet, then radio, then TV, and finally word of mouth. Social media combined with the Internet make up the vast majority for all the levels of affectedness, the highest level for indirectly affected individuals, which may be due to their need to contact friends and family to confirm their safety.

Figure 4-38: Relationship between level of affectedness and most relied on source of information

The motivation for directly affected individual in using social media is most commonly mass sending ability and convenience (See Figure 4-39), suggesting their need for fast and widespread information. Indirectly affected individuals also prize these two qualities but also rate quality of information high. Non-affected individuals cite convenience and quality of information as of importance. Desire to help and mass sending ability is also high, showing their desire to contribute to support people affected by the situation. All levels of affectedness have a similar level of experience with social media.

110

Figure 4-39: Relationship between level of affectedness and motivation

Each level of affectedness has similar trends in problems associated with receiving information with lack of trust in information being the most prominent for each group. The groups also had similar levels of skill related utilization problems. Directly affected individuals also experienced problems with lack of electricity to a fairly high degree. Failure in the goals of utilization was similar in all the groups.

Figure 4-40: Relationship between level of affectedness and problems in receiving information

Directly affected individuals overwhelmingly cite social media as “extremely helpful” during the disaster (See Figure 4-41). Indirectly individuals have near equal amounts of “helpful” and “extremely helpful.” All groups have near universal positive experience with social media during the disaster, with all having “helpful” and “extremely helpful” equaling over 90% of responses.

111

Figure 4-41: Relationship between level of affectedness and level of helpfulness of social media during the disaster

Respondents of each of the levels of affectedness give similar responses on their overall impression of the usefulness of social media in disaster in general as the previous responses regarding social media’s helpfulness in the in the disaster (See Figure 4-42).

Figure 4-42: Relationship between level of affectedness and overall impression of the usefulness of social media during disasters in general

When asked what areas of information need to be improved when receiving information through social media during the disaster, directly affected individuals stated all areas need to be improved more or less an equal amount, excluding infrastructure information which was given a higher priority. Indirectly affected individuals cited infrastructure information and special needs information as the areas that most needed improvement. Non-affected individuals chose infrastructure information, general information and special needs information as the areas most need for improvement. Infrastructure information was high amongst all the groups.

112

Figure 4-43: Relationship between level of affectedness and areas to be improved for receiving information in disasters

When asked if governments should use social media during disaster, all groups answered “yes” at an extremely high rate. Directly affected individuals chose “yes” at a rate just over 95%, non-affected individuals at just under 95%, and indirectly affected individuals at just under 85%. Support for government use of social media is overall very high.

Figure 4-44: Relationship between level of affectedness and support for government use of social media in disasters

In summary, directly affected individuals used a broad range of sources of information, with social media taking up the largest portion, nearly 35%, followed by the Internet, then radio, then TV, and finally word of mouth. Social media combined with the Internet make up the vast majority for all the levels of affectedness, the highest level for indirectly affected individuals TV makes up a large portion of most depended on type of media for the non- affected.

The motivation for directly affected individual in using social media is most commonly mass sending ability and convenience. For directly affected speed is more important than quality of information. Indirectly affected individuals also prize mass sending ability and convenience but also rate quality of information high. Non-affected individuals cite convenience and quality of information as of importance.

All levels of affectedness have a similar level of experience with social media. Each level of affectedness has similar trends in problems associated with receiving information with lack of trust in information being the most prominent for each group. Directly affected individuals experienced problems with lack of electricity to a fairly high degree. All levels of

113

affectedness think that all areas of information need to be improved for receiving in disasters with infrastructure information being particularly high amongst all groups.

All groups have near universal positive experience with social media during the disaster, with all having “helpful” and “extremely helpful” equaling over 90% of responses and the higher the level of affectedness the more positive view they have. All levels of affectedness think that government should use social media in disasters at an extremely high rate, with support being stronger with the higher amongst individuals with a higher level of affectedness.

4.3 Synthesis of Summary of Key Findings

4.3.1 Main Results

Main results of the performed analysis from the responses to the online questionnaire survey using the framework for analysis are shown in summarized form below. In addition, in order to clearly convey the results of the analysis, summary tables of the factors identified in the framework of analysis by user role, location, and level of affectedness have been produced.

The main findings and other and some details are presented showing the intersection of user roles, location, and level of affectedness by each factor chosen for analysis. The summary tables also present an “overall” category, in an attempt to show the user roles, individuals by location, and individuals by level of affectedness grouped together with their collective attributes of each factor.

Each summary table is intended to give an overview of the main results of the analysis and to create a profile of user roles, individuals by location, and individuals by level of affectedness for the reference of parties interested in the utilization of social media in disasters to better inform the tasks undertaken by them.

The following are the main results as analyzed by user role:

Main results of Attributes of respondents

• Users of social media were generally older than imagined. • The majority of each user group is from the affected areas of Tohoku and Kanto regions. • There is an equal balance between genders except for manager senders. • The majority of each user group is from the affected areas of Tohoku and Kanto regions. • Most respondents are intermediate users. • Only 3 respondents do not use social media in general. • For non-users, the largest portion are directly affected.

Main results of Source and type of information

• Social media and the Internet combined make up the vast majority of reliable information sources in the disaster as well as perceived reliable information source for disasters in general for all groups that used social media. • Groups that did not share information have a larger variety of sources that are viewed as the most reliable. • Non-users relied on TV the most.

114

• Email, which many people use with mobile phones, was not mainly relied on by anyone. • For less experienced people TV and radio are main forms of receiving information. • Non-senders also chose the Internet as the most valuable source of information. • Users who said that they radio was the most reliable in their own experience chose other forms as being more reliable for disasters in general. • Individuals who share information receive information that may not be applicable to themselves at higher rates. • Basic senders shared special needs information at a low rate compared with volunteer senders and manager senders. • Organized group sharing has a balance of information sought and shared and includes special needs and infrastructure information. • Non-senders rely on primarily social networks and web pages. • Senders share using a greater variety of tools when compared with non-senders. • All sender groups rely on social media and webpages highly. • All user groups receive information from first hand accounts are at similar levels, but receiving government and traditional mass media are lower in groups that have a lower level of participation. • Manager senders put a high priority on information from government sources and treated information from traditional mass media and second hand accounts about the same. • Manager senders having balanced levels of types of information in receiving, shared government information at much higher rates. • Besides social media, All user role groups had very high levels of TV use. • In general, radio and newspaper were much less utilized by respondents.

Main results of Motivation in receiving and sending information

• Convenience and mass sending ability are commonly cited reasons for utilizing social media. • Senders have a relatively high level of prior experience. • Desire to help is high among senders and low among non-senders.

Main results of Problems encountered during receiving and sending of information

• The most common problems amongst all the groups were related to trustworthiness of information. • Skill issues are a problem but secondary compared with problems with trustworthiness of information. • All groups state that the information that they shared was used at very high rates.

Main results of Evaluation of the usefulness of social media in this disaster and in disasters in general

• Overall, respondents have an overwhelming positive view of the helpfulness of social media in the disaster for themselves. • Respondents with higher levels of participation answer with “extremely helpful” at higher rates. • Each group has high, over 80%, support for governments using social media in disaster.

115

In addition to the above listed main results, a summary table based on user roles was produced (See Table 4-6). The table show the factors of information, motivation, evaluation, and expectation of the identified user roles.

Table 4-6: Summary table of main results of analysis of user roles and factors

The main results of the correlation analysis are presented in below.

116

The following are the main results as analyzed by location:

• Respondents in the Tohoku region relied upon social media primarily, followed by the Internet, and then word of mouth and TV. • Respondents in the Tohoku used a greater variety of sources of information, although social media was the most relied upon. • Respondents in the Kanto region relied upon social media to the highest degree, but also on the Internet to a high degree, followed by TV. • Users in the Tohoku region overwhelmingly state mass sending ability followed by convenience as the reasons for using social media in the disaster. • Respondents in Japan in other areas than the Tohoku and Kanto regions, chose quality of information most often followed by convenience. • Quality of information, technological barriers and desire to help were also commonly chosen reasons for utilization. • Lack in trust of information was the largest problem related to receiving information during the disaster for all regions except for areas outside of Japan. • Tohoku region and Kanto region had a significant amount of problems related to lack of electricity. • Users in the Tohoku and Kanto region primarily regard social media as “extremely helpful” in the disaster with nearly all of the remainder regarding it as “helpful.” • Other areas in Japan found social media to be “extremely helpful” and “helpful” in the disaster at a similar rate, making up the vast majority of responses. • All areas found social media in the disaster to be helpful in extremely high rates, with users in the Tohoku and Kanto regions more likely to find it “extremely helpful.” • All areas of information that are shared through social media during the disaster need to be improved. • Infrastructure information and special needs information are areas which are seen to need the greatest improvement. • Tohoku and Kanto regions chose “yes,” governments should use social media in disasters, at just over 90% and individuals in other areas of Japan and outside of Japan chose “yes” at just over 80%.

In addition to the main results listed above, a summary table based on location was produced (See Table 4-7). The table show the factors of information, motivation, evaluation, and expectation of the of individuals by their location during the disaster.

117

Table 4-7: Summary of main results of analysis of location and factors

The following are the main results as analyzed by level of affectedness:

• Directly affected individuals used a broad range of sources of information, with social media taking up the largest portion, nearly 35%, followed by the Internet, then radio, then TV, and finally word of mouth. • Social media combined with the Internet make up the vast majority for all the levels of affectedness, the highest level for indirectly affected individuals • TV makes up a large portion of most depended on type of media for the non-affected. • The motivation for directly affected individual in using social media is most commonly mass sending ability and convenience. • For directly affected speed is more important than quality of information.

118

• Indirectly affected individuals also prize mass sending ability and convenience but also rate quality of information high. • Non-affected individuals cite convenience and quality of information as of importance. • All levels of affectedness have a similar level of experience with social media • Each level of affectedness has similar trends in problems associated with receiving information with lack of trust in information being the most prominent for each group. • The groups also had similar levels of skill related utilization problems. • Directly affected individuals experienced problems with lack of electricity to a fairly high degree. • All groups have near universal positive experience with social media during the disaster, with all having “helpful” and “extremely helpful” equaling over 90% of responses and the higher the level of affectedness the more positive view they have. • All levels of affectedness think that all areas of information need to be improved for receiving in disasters with infrastructure information being particularly high amongst all groups. • All levels of affectedness think that government should use social media in disasters at an extremely high rate, with support being stronger with the higher amongst individuals with a higher level of affectedness.

In addition to the main results listed above, a summary table based on level of affectedness was produced (See Table 4-8). The table show the factors of information, motivation, evaluation, and expectation of individuals by level of affectedness.

119

Table 4-8: Summary of main results of analysis of level of affectedness and factors

4.4 Key Findings

Key findings show that over 60% of non-affected individuals, nearly 80% of indirectly affected individuals, and over 55% of directly affected individuals responded that social media or the

120

Internet was their most relied on source of information. While areas outside of the Tohoku region cited social media and the Internet combined at a rate near 70% (Kanto 71%, other areas in Japan 69%, areas outside of Japan 80%), individuals in the Tohoku region, the area affected by the tsunami, cited social media and the Internet combined at a much lower rate of 45%. Respondents in all areas in Japan evaluate “lack of trust in information” as the greatest problem associated with social media use in the disaster, with respondents in all locations rating it over 33%, whereas respondents outside of Japan rate it much lower at 14%. Users regardless of location or level of affectedness both overwhelmingly evaluated social media to be “helpful” or “extremely helpful” during the disaster at a combined rate of 94%, Support for government use of social media for respondents in the disaster affected areas (Tohoku and Kanto regions) is 95%, higher than those of other areas, which is over 84%. Support for government use of social media in disasters is extremely high, above 80% for individuals in all locations, all levels of affectedness, and all user roles.

4.5 Key Results

Using the main findings of the analysis, the summary tables of the factors related to the role of social media in the disaster, and the key findings, some key results were produced. The key results are as stated below:

• Social media and the Internet were found to be reliable in the disaster to very high levels regardless of user role, location or level of affectedness and perceived reliability is even higher.

• For directly affected individuals and individuals in the affected areas convenience and mass sending ability were the strongest reasons for utilizing social media during the disaster.

• Problems with trustworthiness of information were experienced by all to a high degree.

• Even though problems were experienced, users found social media to be extremely helpful in the disaster and beneficial in disasters in general to an overwhelmingly degree.

• Users feel that all areas of information sharing through social media in the disaster need to be improved, infrastructure information to a particularly high degree.

• Support for government use of social media in disasters is extremely high and is higher among directly affected individuals, individuals in disaster areas, and group senders.

• The higher the level of participation in sharing of information through social media during the disaster, the more likely an individual is to received and share a broader amount of information, with a higher likelihood that that information is from credible sources. References

Acar, A. and Murakami, A. (2011). Twitter for crisis communication: lessons learned from Japan's tsunami disaster. International Journal of Web Based Communities. 7 (3). 392-402.

121

122

Chapter 5 Discussion and Conclusion

5.1 Discussion

Results from data collection of examples and the online questionnaire survey show there is an argument to be made for the utilization of social media in disasters. The following is a discussion of the results of this survey and how it relates to the future use of social media for disaster preparedness and disaster relief, as well as how it relates to social media’s possible use for these purposes by organizations and governmental bodies.

The objectives of this research are to understand the role of social media in the disaster, measure the experience and opinions of users of social media during the disaster, identify problems related to its use in order to improve social media use in disaster for more effective disaster preparedness and disaster response, and to identify approaches of social media use for disaster preparedness and response for NPOs and local and national government bodies. The findings of this research demonstrate that social media was convenient and helpful as a method of communication during the disaster. It also shows that users experienced problems related to trustworthiness of information to a high degree. It was also demonstrated that users largely support government use of social media and that users that have a higher level of participation in sharing through social media receive and share information from creditable sources at higher rates.

These results show that social media is a powerful tool for disaster situations, given its flexibility and usability by individuals to meet their needs in many different situations, and that users evaluate it highly. The results also shows that there are problems associated with social media use that need to be considered, but that these problems do not outweigh the benefits of the tool. For these reasons, organizations and governments should consider utilizing social media for disaster preparedness and disaster response as one tool to exchange information related to disasters.

5.1.1 Reasons Organizations and Government Bodies Should Utilize Social Media for Disaster Preparedness, Relief, and Recovery

As Tseng, Chen and Chi (2011) demonstrate, social media can be used for information dissemination and resource mobilization that allow the public to participate in disaster relief and play vital roles as collaborators, and that social media is a more efficient method of disaster backchannel communication, so government and public participation in disaster is possible. In addition to this fact, results of this survey show that participation by organizations and government bodies in social media use for disaster preparedness and disaster relief can be beneficial. Using some of the key findings of this research explanations will be given for reasons why organizations and government bodies should utilize social media for disaster preparedness, relief, and recovery.

• Social media and the Internet were found to be reliable in the disaster to very high levels regardless of user role, location or level of affectedness and perceived reliability is even higher.

In one example of the findings, key findings show that over 60% of non-affected individuals, nearly 80% of indirectly affected individuals, and over 55% of directly affected individuals

123

responded that social media or the Internet was their most relied on source of information. This shows that social media is found to be a highly reliable source of information during disasters, higher than other sources of information, therefore it can be assumed that it is an effective way to share information to those who need it during disasters and benefit the organizations and government bodies that utilize it by getting their message out to those they want it conveyed to.

• For directly affected individuals and individuals in the affected areas convenience and mass sending ability were the strongest reasons for utilizing social media during the disaster.

Directly affected individuals chose convenience at 47% and mass sending ability at 48%. Individuals in the Tohoku region chose convenience at 50% and mass sending ability at 57%. Individuals with lower levels of affectedness and in other regions chose these options at lower rates. This shows that these qualities are important to user who were at risk. Therefore, there would be benefit for local governments to use social media to inform residents in the case of disaster. In order to do this effectively, it would be beneficial to create accounts and educate residents of the existence of accounts and on how to use social media, before a disaster occurs as a type of disaster preparedness. In addition, these groups overwhelmingly said social media was “helpful” or “extremely helpful” during the disaster, so social media is clearly a beneficial addition to other sources of information inside disasters directly affected persons and persons in affected areas. This is another reason why organizations and local governments should utilize social media as a part of disaster preparedness plans. Doing so could help improve the reliability of information through social media in disasters because local residents could receive trusted information directly from a source that is reliable and trustworthy, local government and the shared sources of information that that government has deemed reliable and shared through social media.

• Even though problems were experienced, users found social media to be extremely helpful in the disaster and beneficial in disasters in general to an overwhelmingly degree.

Respondents in all areas in Japan evaluate “lack of trust in information” as the greatest problem associated with social media use in the disaster, with respondents in all locations rating it over 33%, whereas respondents outside of Japan rate it much lower at 14%. Although, users from all user groups, areas, and levels of affectedness found social media to be “helpful” or “extremely helpful” at a combined rate of over 90%, so these problems did not affect the vast majority of users so much that they thought that social media was not helpful, thereby further making the argument for its increased use in disasters. Furthermore, these users believe that social media use in future disaster will be beneficial at similar rates, above 85%, demonstrating its future potential.

• Support for government use of social media in disasters is extremely high and is higher among directly affected individuals, individuals in disaster areas, and group senders.

Support for government use of social media for respondents in the disaster affected areas (Tohoku and Kanto regions) is 95%, higher than those of other areas, which is over 84%. Support for government use of social media in disasters is extremely high, above 80% for individuals in all locations, all levels of affectedness, and all user roles. This information shows a clear mandate for government bodies to utilize social media during disaster. Given the rise in the use of the Internet and social media as demonstrated in the literature review, it

124

is likely that a rising number of individuals will use social media in the future and therefore depend on it in disasters when they occur. Social media users in this disaster, especially those who were affected, want governments to use social media in disasters, so doing so could

Users of social media in disasters think that governments should use social media in disasters and directly affected individuals have the highest rates of support for doing so.

This research show that individuals who share, and those who share in groups even more so, receive balanced information and share information from creditable sources, so by organizations and governments encouraging users to get involved in sharing and connecting users with ways to share information in disasters, organizations and governments can increase the sharing of more creditable information during disasters.

5.1.2 The Flow of Information in Disasters

During the course of data collection of examples of the use of social media in the disaster, a diagram demonstrating the theoretical flow of information during disasters was created to aid in the visualization of how information from traditional sources and social media is combined in disaster (See Figure 5-1). This show how information in social media can come from traditional sources, or senders, and individuals using social media and that information can go out to both groups. Starbird, Palen, Huges and Vieweg (2010) showed that official information remains important and is complimented, not usurped, by information coming from social media that is shared by the public. The results of the authors’ research show that individuals with higher levels of participation in organized sharing shared information through social media received information from reliable sources such as traditional mass media and government sources at higher rates. Therefore, an important point to grasp is that by encouraging individuals to share information, it may be possible to increase the reliability of information shared through social media. Organizations and government bodies could do this by promoting social media use amongst the public and by the groups themselves using social media before and during disasters.

125

Figure 5-1: Diagram of the flow of information in disasters

5.1.3 Social Media Use from Inside and Outside of the Disaster

Another point that can aid in understanding how social media functions in disasters and can be of help to an organization or government body interested in implementing social media use for disaster preparedness and disaster relief is the differences in social media use from outside and inside of disaster.

An important aspect of social media use in disaster is its ability to be used inside and outside of disasters. When speaking of social media use in disasters, often people imagine only directly affected individuals utilizing social media. This research shows that directly affected individuals and individuals in the affected areas believe that social media and the Internet combined were the most viable forms of communication so social media is valuable to these groups, as the collected examples show there are many individuals who use social media to contact their family or friends and there are also many individuals who use social media to provide support such as the case of Anpi Report or to receive vital information such as the case of the “homemade” sign language news group. There are many examples of ways individuals can offer support from outside of the disaster, such as gathering information on the safety of individuals or gathering information for crisis-mapping, etc.

After the 2011 East Japan Earthquake and Tsunami, there was a great outpouring of good will and desire to help, but it was made clear by the government and organizations that the only way to help was to donate funds. One possible way to direct this good will and mass amounts of human resources into a positive form of support for affected individuals may be through encouraging individuals to participate in social-media-driven support projects.

126

Since support for government use of social media in disasters is high amongst respondents, it is likely that many social media users would be willing to participate in projects run by organizations or governments or projects introduced to them by organizations or governments. Organizations and government bodies can use the goodwill of these individuals to provide support during disaster by communicating and coordinating prior to and during disasters.

In addition, it should also be pointed out that affected individuals can also provide support while receiving information through social media. Social media use in disaster does not only mean use by directly affected individuals and the relationship of senders and receivers is not only one way from outside of the disaster to inside of the disaster.

5.1.4 Social Media use throughout the Disaster Cycle

Similar to how social media use in disasters does not only concern directly affected individuals, use of social media in the disaster is not only restricted to the immediate moment when a disaster strikes. There are many examples of how social media was used when the earthquake and tsunami struck, but there are also examples of social media use in the recovery and preparedness phases also, such as Peace Boat’s use of social media to recruit volunteers and solicit donations after the disaster. A tool such as the Ushahidi crisis-mapping system is developed before a disaster occurs and implemented in the relief phase. Using social media to give updates on the status of infrastructure information is a way that social media could be utilized in the recovery phase. In fact, infrastructure information was cited as the area of information that needs most improvement in social media. Users in the Tohoku region cited infrastructure information as an area to be improved at 36%, nearly 10% higher than any other type of information. This shows that users expect and want information through social media in the disaster phase but are not receiving it. Organizations and governments could improve support to affected individuals by recognizing what types of support can be given over all phases of the disaster cycle.

5.1.5 Social Media Tools

Tools related to social media, as shown in the literature review, have been changing rapidly over time, and as shown in the data collection of examples of social media use in this disaster there are a multitude of tools that are utilized in various of ways. Therefore, it is important to recognize the constantly changing nature of social media tools and understand that there is no one formula for utilizing them in disasters and that the most effective use of them is to be flexible, and adapt and adopt new tools when they develop. Organizations and governments that utilize social media in disasters should pay attention to this fact and not stick to one style of social media use. When a disaster happens, observing others and what new ways they utilize social media is beneficial.

Another important fact to keep in mind when utilizing social media in disasters, as this research shows, is that social media is one type of tool that receivers of disaster information utilize in conjunction with other traditional methods of receiving information in disasters. Over 72% of all users also received information from TV during the disaster, showing that most users were not specifically reliant on only social media. Each method has its benefits and its drawbacks. Social media should not be seen as a “magic bullet” for all. It has problems with reliability of information and usability as the results showed. Social media should be used when its use is appropriate and when users can receive information from trusted sources within it.

Keeping these facts in mind, social media can act as another layer of information to lay on top of a situation to better understand it. Users and non-users regardless of level of affectedness

127

or location or user role rate it as a highly reliable source of information. This is most likely due to its flexibility and ability for innovation in conjunction with other sources of information, working with these aspects will make use of it in disaster more successful. One of the terms strongly associated with social media use is “mash up,” meaning to combine several sources of information in different ways. This is an important attribute to keep in mind when trying to accomplish tasks with social media.

5.1.6 Recent Trends and Changes which Might Affect Adoption of Social Media Use during Disaster

As shown in the literature review, use of mobile phones is ubiquitous and usage of the Internet and Internet-capable mobile devices is growing rapidly. Given the use of social media in past disasters and the strong approval of it as a reliable source of information, as shown by the results of this survey, it is likely that social media use in disaster will rapidly grow, whether it is supported by organizations or government or not. Therefore, it is necessary to recognize that this increased use will likely create dependence on social media as a form of communication in disasters. With this knowledge, organizations and governments would benefit from following changes related to the use of social media in disaster and look into how to protect the pathways of information that social media use from breaking down in disasters.

Another possible change concerning the use of social media in disasters could be a decline in the age-based digital divide over time. Respondents to the online survey, in general, were older than expected. Mid-aged users made up 50% of the respondents. As current generations age and tools for accessing social media become simpler to use, there could be more older individuals who are capable of accessing social media in disasters. This is a possibility that should be observed.

Another possible change related to social media, which could be related to the digital divide, is a possible lowered dependence on radios in disasters. Radios are extremely valuable in disaster situations because of their mobility and relatively low use of energy, but respondents to the online questionnaire survey chose radio as the most reliable source of information in their personal experience at a low rate. Of non-users of social media during the disaster, 18% said that radio was the most reliable source of information during the disaster in their experience, yet when asked what was their perceived most reliable source of information, only 4% chose radio. This shows that these individuals think that social media is a more reliable source of information although they happened to rely on radio themselves.

Another reason for this lowered reliance on radio may be that, since many individuals carry around mobile devices at all times, they are a quicker and more convenient source of information. One way to combine social media and traditional mass media sources such is radio are new applications for smartphones that allow users to listen to radio over the Internet. These applications should be looked into as alternatives to radios and promoted by organizations and governments as methods to access reliable information sources via social media sources in disaster.

Social media use in disasters could grow as battery life of mobile devices increases and as decentralized energy such as solar power grows in popularity. Increased battery life would allow users to access social media for longer periods of time after a disaster and if non- centralized power sources are available, users may still be able to charge mobile devices or use computers for longer periods of time during disaster.

Finally, current social media tools combined with the US Defense Department’s newly developed tool called the Transnational Information Sharing Cooperation (TISC), which is a

128

centralized social network or “community” made up of organizations and individuals related to the disaster relief could be create unique methods to approach disaster relief.

5.1.7 Concerns, Barriers and Problems Related to the Use of Social Media in Disaster

Along with the benefits of social media there are also limitations, barriers, and problems related to its use. Some of these problems are related to reliability of information, misinformation, technological difficulties, problems related to skill, digital divide issues, and privacy concerns.

Although social media was evaluated very highly by respondents in the survey, social media was also recognized as having problems with reliability of information such as misinformation, which is supported by the research of Acar and Murakami (2011). Information is social media can easily be reproduced and spread quickly, so if that information is incorrect or is malicious, it could have serious consequences. Furthermore, checking the original source of information or knowing the reliability of information is very difficult, so systems need to be implemented to make information more reliable and provide and easy way to check the origin of information during disasters. Qu, Huang, Zhang, and Zhang (2011) stress the need for systems to recognize the difference between official information and other information, as well as between first and second-hand information. Also, Acar and Murakami (2011) suggest the use of official hashtag keywords for use in Twitter. This does not yet exist, but organizations and government bodies that plan on using social media in disaster preparedness and disaster prevention should look for changes in ways of functionality of social media such as this and actively promote such measures.

Social media is also dependent on electricity and data networks, so in disasters where the supply of energy has been disrupted or data networks have been damaged, it may not be useable. This is a limitation associated with its use.

Due to the many software and services that make up social media and the devices such as computers and mobile phones that it relies on, inexperienced individuals may have problems access information through it. The majority of respondents to the online questionnaire survey were intermediate social media and Internet users, so they have a large amount of experience with the software, services, and devices, This limitation is also connected to digital divide issues; due to the complexity of the tools and the their lack of experience, older individuals are often not capable of accessing information through social media. This is why social media, if supported by governments and organizations, should consider it only one of the methods to get information to individuals in disasters. Organizations and government bodies can provide training for older individuals on the use of social media in disaster. A more realistic approach for including these individuals in disaster preparedness training related to social media is to stress that these individuals can get vital information via social media by connecting to an individual who has access to it such as a family member or friend.

Finally, due to the fact that social media use in disasters is relatively new, it is likely that, just as new uses for social media in disasters are being developed, new problems will arise. Organizations and governments that utilize social media in disaster should be aware of these limitations and risks and do their best to minimize negative impacts from them.

129

5.2 Conclusion and Policy Implications

5.2.1 Conclusion

Social media use in disasters is a new phenomenon that is constantly changing, evolving, and likely to grow. The unique set of circumstances that existed when the 2011 East Japan Earthquake and Tsunami occurred gives insight into this new phenomena and provides a model for how to understand social media use in disaster and a roadmap for future developments at a depth that was not possible in earlier disasters due to immaturity of the related technologies or lack of popularity of the services.

This research has demonstrated the scale and variety of innovative responses to the disaster performed with these newly popular social networks, services and software. This research also measured the experiences and opinions of those who have first hand experience using social media in and outside of the disaster as receivers and senders of information. This has made it possible to understand the role social media played in the disaster, the type of information exchanged through it, the challenges associated with its use, the value of it to its users and its potential role in the future.

It has also identified problems associated with its use in disaster such as the spread of misinformation and digital divide issues and propose recommendations for pre- and post- disaster scenarios for the public, civil society, and government organizations, so to identified entry points to utilization of the new models of relief communication for NPOs and local and national governments for future utilization.

5.2.2 Methods of Future Application of Social Media in Disasters by Organizations and Governments

Given the overwhelming approval of social media in the disaster and the strong support for government use of social media in disasters, which has been demonstrated by this survey, as well as the strong likelihood that social media use will increase, in general and in disasters, it is recommended that organizations and governments become familiar with the associated tools and issues and take action to implement social media use in disaster preparedness and planning so in case of disaster they may be ready to use it for disaster response and recovery.

As well as the previous recommendations, the following are suggestions for organizations and government bodies intending on doing so. A concept diagram demonstrating these suggestions has also been included below (See Figure 5-2)

• Create accounts for the organization in social media services

• Keep up-to-date on changes with social media and new possible uses in disaster

• Publicize the fact that the organization is utilizing social media for the intended use in disasters

• Provide education on the use of social media for groups likely to have issues with using the technology, such as older individuals

• Connect with existing groups using social media with the intention of utilization in disasters

130

• Create or join projects that seek to provide support through social media during disasters

• Provide disaster preparedness updates through social media to educate the public about disaster issues

• Set up a method for providing information in case of disaster through social media with an emphasis on infrastructure information

And in the case of disaster:

• Provide updates about the disaster situation through social media

• Provide infrastructure information through social media

o Plug into existing crisis-mapping solutions such as Ushahidi/sinsai.info

• Keep aware the how users are utilizing social media in the disaster and of new innovations that might be useful at that time

• Encourage people to help provide information based support through social media

• Utilize the “crowdsourced” power of the crowds in social media

• Direct users who want to help in some way to organizations that they can provide support to via social media

• Provide information for individuals with special needs such as deaf individuals or foreign residents who might have language issues, to specialized sources of information that they can access

• Share your experiences

131

Figure 5-2: Concept diagram of possible participation in disaster preparedness and disaster relief by organizations and government bodies

References

132

133

134

Appendices

Appendix 1: Questionnaire survey on the use of social media in the 2011 East Japan Earthquake and Tsunami (English) Appendix 2: Questionnaire survey on the use of social media in the 2011 East Japan Earthquake and Tsunami (Japanese)

A - I

A - II

Appendix 1: Questionnaire survey on the use of social media in the 2011 East Japan Earthquake and Tsunami (English)

A - III

A - IV

Appendix 2: A survey on the use of social media in the 2011 East Japan Earthquake and Tsunami (Japanese)

A - V

A - VI