key: cord-344832-0ah4w59o authors: Sakurai, Mihoko; Adu-Gyamfi, Bismark title: Disaster-Resilient Communication Ecosystem in an Inclusive Society – A case of foreigners in Japan date: 2020-08-15 journal: Int J Disaster Risk Reduct DOI: 10.1016/j.ijdrr.2020.101804 sha: doc_id: 344832 cord_uid: 0ah4w59o The number of foreign residents and tourists in Japan has been dramatically increasing in recent years. Despite the fact that Japan is prone to natural disasters, with each climate-related event turning into an emergency such as with record rainfalls, floods and mudslides almost every year, non-Japanese communication infrastructure and everyday disaster drills for foreigners have received little attention. This study aims to understand how a resilient communication ecosystem forms in various disaster contexts involving foreigners. Within a framework of information ecology we try to get an overview of the communication ecosystem in literature and outline its structure and trends in social media use. Our empirical case study uses Twitter API and R programming software to extract and analyze tweets in English during Typhoon 19 (Hagibis) in October 2019. It reveals that many information sources transmit warnings and evacuation orders through social media but do not convey a sense of locality and precise instructions on how to act. For future disaster preparedness, we argue that the municipal government, as a responsible agent, should (1) make available instructional information in foreign languages on social media, (2) transfer such information through collaboration with transmitters, and (3) examine the use of local hashtags in social media to strengthen non-Japanese speaker’s capacity to adapt. The geographic characteristics of Japan makes the country highly vulnerable to disasters such as earthquakes, typhoons, volcanic eruptions, flash floods and landslides [1] . Therefore, disaster risk reduction and resilience measures are enshrined in the every-day activities of the Japanese people including school curriculum, building regulations and design, as well as in corporate organization setup [2] . Disaster risk reduction drills often take place in schools, work places, homes, and include the publication of detailed evacuation plans and procedures in all local communities [3] . However, these efforts of building the capacity of residents facing emergencies could be falling short in terms of coverage and results due to the influx of foreigners who may neither be enrolled in schools or engaged with Japanese establishments or are staying only for a short period of time [4] . The population of foreign residents in Japan has been increasing exponentially in recent years according to a report by Mizuho Research Institute [5] , with their number reaching a record high of 2.5 million people by January 1, 2018. The report reveals an increase by 170,000 residents from January 1 to December 31 in 2017. Additionally, the Japan National Tourism Organization reports that the number of foreign tourists visiting Japan is rapidly increasing. It recorded 31 million in 2018, which was four times as many as visiting in 2008, and six times over the number visiting in 2003 1 . This trend of increased foreign tourists may partly stem from the influence of numerous sporting and other events across the country which include the Formula 1 Grand Prix, the FIVB Volleyball Women's championship and the Japan Tennis Open [6] . The Tokyo 2020 Olympic and Paralympic Games, which has been postponed to 2021, as well as the push by government stakeholders to amend bills to relax immigration laws can also be expected to boost the number of foreign residents and tourists in the country [7, 8] . Even though some foreigners become permanent residents, the qualification for permanent residence status requires a ten-year continuous stay in Japan including five years of work, but does not necessarily require the acquisition of Japanese language skills [9] although it may matter at some points. Living in japan briefly or as a permanent resident does not necessarily mean the person speaks Japanese or is accustomed to disaster risk reduction culture and procedures in Japan. Although the composition and dynamics of the Japanese population is gradually changing in terms of the non-Japanese population, existing infrastructure and systems that support non-Japanese speakers and foreigners as a whole during disasters seem to 1 Last accessed on March 6, 2020. https://www.jnto.go.jp/jpn/statistics/visitor_trends/ J o u r n a l P r e -p r o o f be inadequate or lacking 2 . The current Japanese disaster management system is composed of both vertical and horizontal coordination mechanisms and, depending on the scale of the disaster, activities flow vertically from the national level, through the prefectural to the municipal level before getting down to the communities themselves [10] . The municipal disaster management council together with other parties are close to disaster victims and responsible for municipal disaster management plans and information [11] . When necessary, the municipality is also responsible for the issue of evacuation orders to residents during disasters [11] . Again, when it comes to emergency alerts or risk information, there is an existing mechanism where warning messages proceed from the national government (the cabinet, Japan Meteorological Agency (JMA)) to local administrations (prefecture, municipalities) and from there to the people. Messages may also be transmitted directly top-down through base stations and are received on cell phones. However, reports by news outlets suggest that language barriers, coupled with the inexperience of many foreigners with regard to Japanese disaster procedure protocols, create a huge sense of panic and confusion during disasters [12, 13] . Lack of appropriate risk information and procedural evacuation or alert actions create this confusion amongst the foreigners. Therefore, most foreigners access news outlets in their respective countries or rely on social media for disaster risk particulars, alert instructions, or evacuation information [14] . The use of social media is seen as a significant trend in accessing swift, precise and easy feedback information in critical situations [15] [16] [17] . Social media has become the most sourced avenue for information during disasters [18] . This applies to both Japanese and non-Japanese users. Irrespective of the chaotic nature or confusion that erupts due to limited risk information accessibility and delivery, a certain level of resilience is achieved through instinctive user responses. Thus, in extreme events, a spontaneous system of resilience is often formed within the context of all the actors involved in the situation [19] . This process usually develops because the participating actors share a common interest or find themselves in a situation that requires urgent solutions. The characteristics of this phenomenon create a state of ecosystem which becomes unique to the area affected and to the nature of the event; it generates interactions and communication procedures or methods to reduce vulnerabilities to the disaster [19, 20] . The interactions within this ecosystem could be formal or informal, depending on the situation and different communication structures, modes and tools employed. The high frequency of disaster occurrences in Japan in conjunction with a booming foreign population up to Covid-19, and an anticipated, fast rising population again thereafter, provides an ideal case for trying to understand how this ecosystem is established within the context of resilience and communication. Resilience in this study is defined as a system's ability to absorb disturbances [21] . To analyze this ability, we define the system in terms of the information ecology framework. We regard disaster resilience in the information ecology framework to encompass the efforts of collaboration and communication dependencies that exist amongst stakeholders engaged in the situation within a local context. We want to investigate how foreigners in Japan obtain disaster related information while facing a language barrier and inexperience with disaster management procedures. This will guide us to understand the characteristics of the communication ecosystem for a foreign population. This paper is divided into the following parts: first, we introduce the information ecology framework as the theoretical premise of a resilient communication ecosystem. A literature review should give us insights into current studies on the topic and their deployment. A key question is how a resilient communication ecosystem is formed in different disaster contexts. To better understand this question, we seek cases amongst the literature which highlight disaster resilience through collaboration, communication and stakeholder participation. Cases with such attributes are selected, reviewed and discussed from the view point of information ecology. Following the case review, we use the Twitter API (Application Programming Interface) application and the R programing language to collect and analyze English language tweets shared on the Twitter platform during Typhoon Hagibis which hit Japan in 2019. This gives us an empirical understanding of a resilient communication ecosystem. It is assumed that information shared on Twitter in English is meant for consumption by non-Japanese. Again, media coverage of the typhoon is also monitored to serve as supplementary information to our analysis. Based on insights from literature review and findings from tweet analysis, we try to describe a structure which guides our understanding of a communication system applicable to foreigners during a disaster. We conclude the paper with observed limitations and future research directions. Information ecology is composed of a framework which is defined as "a system of people, practices, values, and technologies in a particular local environment" [20]. Previous research uses the notion of ecosystem to describe the nature of resilience in a societal context [22] . Information ecology framework, which is an extended notion of the ecosystem, helps us to examine the capabilities of each element within the system. The framework contains five components, including: system, diversity, coevolution, keystone species and locality. Table 1 summarizes the elements of information ecology. Not shown but also included is the perspective of a technology role which focuses on human activities that are enabled by the given technology implementation. Table Table Table Table 1 1 1 The presence of keystone species is crucial to the survival of the ecology itself, e.g., skilled people/groups whose presence is necessary to support the effective use of technology. Locality Locality Locality Local settings or attributes that give people the meaning of the ecology. Given that resilience is defined as a system's ability to absorb disturbances [21] , information ecology believes that this ability emerges through interrelationships or dependencies between system entities [20] . These interrelationships exist between a diversity of players equipped with relevant technologies. Keystone species play a central role in creating interrelationships, which, in turn, may lead to new forms of coevolution. The notion of locality reminds us of interrelationships and coevolution emerging in a local setting [23] . The information ecology framework comes down to a set of key elements providing a systematic view of resilience. It enables us to understand who within a network of relationships is a key for realizing resilience, and what relations can be observed within a system and under which local contexts. Coevolution is a driving force for forming the resilient ecosystem in times of disaster and is produced by collaboration of systems with J o u r n a l P r e -p r o o f diverse species that provide local context or knowledge [19] . To this we need to add the role of technology in helping human activities and support the formation of coevolution. We notice a similar framework exists as communication ecology [24] . It refers to individual or socio-demographic group connections that strengthen neighborhood communication infrastructure. It helps identify communication patterns of local communities or groups of people. In a disaster context, communication ecology includes community organizations, local media and disaster-specific communication sources [25] . Such communication fosters community resilience [26] , which can be described as the ability of a neighborhood or geographically defined area to collectively optimize their connected interactions for their benefit [27] . Collective ability is required to deal with stressors and resume daily life through cooperation, following unforeseen shocks. Hence, community resilience and its collective ability aim to empower individuals in a given community [28] . Resilience is enhanced through economic development, social capital, information and communication, and community competence [29] . Provision of or ensuring access to vital information through proper communication systems is essential to strengthen community capacity towards the unexpected. In this regard, a resilient communication ecosystem that this paper tries to get an understanding of, encompasses the dynamically evolving processes that empower the collective ability of information gathering and provision, as well as interactions and collective communication structures among individuals, communities, and local organizations. A literature review in the following section will help us extract the essential elements of a resilient communication ecosystem based on information ecology. This study follows a systematic mapping approach [30] as it has the flexibility, robustness and ability to chronologically categorize and classify related materials in research contexts. The process was based on papers which applied the same approach, i.e., [30] [31] [32] . We set the following research question (RQ); RQ: How is a resilient communication ecosystem formed in time of a disaster? We want to understand how resilience mechanisms are able to evolve spontaneously during disasters through communication, collaboration and the roles of different J o u r n a l P r e -p r o o f stakeholders affected by them. To accomplish this, this study categorized the process into three main activities [30] . They are (1) search for relevant publications, (2) definition of a classification scheme, and (3) review of relevant journal papers. The search for relevant publications was undertaken in two steps. First, an online database search was conducted with the keywords "Collaboration in Disaster", "Communication in Disaster" and "Institutional Role in Disaster" across a number of journal databases such as ScienceDirect (Elsevier), SpringerLink, and Emerald This procedure resulted in the selection of 51 papers from all the sampled papers. The classification used in this research was to read, clearly identify and arrange the contents of the 51 potential papers to extract the following content; 1. Research titles, 2. Source (type of journal database), 3. Research question of the article 4. Purpose (which keyword is dominant), 5. Type of disaster, 6. Country or region of disaster, 7. Identified stakeholders, 8. Communication tools used, and 9. Communication structure of the case. This stage identified the contents relevant to this study. The communication ecosystem found to be prevalent within the reviewed cases highlights the presence of institutions, individuals and other actors who become involved in disaster events and perform various tasks or activities aimed at reducing either their own risk or that of the stakeholders. In most cases, a system evolved consisting of actors receiving and sharing information when a disaster or crisis took place [33, 34] . Information ecology promotes sharing and learning; particularly about the use of new technologies, and to reduce given levels of confusion, frustration and ambiguities [35] . In this review, social media emerges as a new trend in technology and rather becomes the medium for sharing information with the aim to reduce anxiety about a disaster situation that could negatively affect the people involved [36] . Actors include government agencies, non-government agencies, and other actors who evolve to become part of the resilient structure of the framework, matching the nature of each situation. Actors can be clearly distinguished based on the roles they play within the system. They include (1) remote actors, (2) responsible agents and (3) transmitters, and (4) Austria [37] . The World Health Organization also supports local communities by giving constant guidance and recommendations as a remote actor during Ebola outbreak in West Africa [38] . Responsible agents have the highest authority and are responsible for issuing needed guidance for people to stay out of risk zones, advising rescue efforts, and providing information on how to reduce risk [39] . They are the national agencies such as at the level of ministry or governmental organization [40] , including security agencies and fire brigades [41] . and information flow between them and other parties [45] . Keystone species identified in the described disasters represent the group or actors whose presence is crucial for all resilience activities. Although they are usually described as highly skilled actors, evidence from the reviewed cases also proves that their activities would not be successful without the cooperation and support from other entities [46] . For instance, staff of the Oslo University Hospital can be described as "keystone species" in the case of terrorist attacks in Norway [44] . However, actors such as blood donors to the hospital and patients, news agencies who covered all events, individual volunteers who help spread "blood request" alerts by the hospital on social J o u r n a l P r e -p r o o f media, as well as others who sent time situational information on social media to update others can be said to be the auxiliary actors who complemented the "keystone species" effort. They give rise to the notion that resilience efforts are often created by a system of actors or groups who are connected through information sharing and play specific roles in a cohesive and coordinated manner during events or disasters [47] . Furthermore, the key to all these systems are the space or context in which their actions take place [48, 49] . In information ecology, it is the "locality" or the setting that initiates all processes. This gives the specification of the context and guides the actions that follow. In some cases, factors such as the geographical location, messaging or information content, characteristics of actors, the medium of information delivery, duration of the event, and the actions taken give the "local identity" of the system [17, 41, 50] . During the Ebola crises, this meant the names of the affected countries [43], while for earthquakes and disasters such as fires it refers to their exact locations. For the purpose of this study, locality refers to a system of individuals, agencies and local communities that refer to a local information and actions to be taken in order to reduce risk within a given area. Table 2 summarizes insights from our case review which illustrate the structure of a resilient communication ecosystem. Table Table Table Table 2 As stated previously, the Japanese governance system is composed of three tiers, national, prefectural, and municipal. Among these, municipal government is closest to people at large and therefore is in charge of issuing evacuation orders, opening and operating evacuation centers, and managing relief efforts [58] . When an evacuation order is issued by a local municipality, residents who live in specific zones are supposed Data collection was done using Twitter API in R package to scout tweets in the English language. To extract the tweets, the following steps were used: a. Registration with Twitter API to secure API credentials b. Integrating Twitter CRAN (programming language for R package) into R statistical package c. Searching by tweets. This was conducted by searching the hashtags "#Hagibis" and "#Typhoon_19". To restrict tweets to only English language, the code "en" was added. Typhoon Hagibis made a landfall on October 12, 2019. Therefore, the tweet search was restricted to search for information of tweets made on that day. The entire code use for the search was (Hagibis <-searchTwitter ("#Hagibis","#Typhoon_19", n="5000", lang = "en", since = "2019-10-12 00:00", until = "2019-10-13 00:00"). Five thousand tweets were collected from this search. d. Export and further analysis: the results were exported to Microsoft Excel for further analysis. Data analysis was based on the framework of a resilient communication ecosystem. The first aim was to find keystone species supporting the communication ecosystem. Therefore, we focused on the number of tweets and Twitter ID accounts with a high number of retweets. As a result of retweet number analysis, we established the formation of coevolution within the ecosystem. We also aimed to explore the kind of information content exchanged in the ecosystem. Hence we conducted thematic analysis [59] which allowed us to apply key elements of a communication ecosystem. It shows how the communication ecosystem for English language was formed when the typhoon hit Japan. In order to extract a sense of locality, we picked up the name of a region or city, and searched for local information such as evacuation orders or emergency alerts within a certain tweet. Diversity in this case reflects the number of tweets and different account ID information that posted or retweeted with #Hagibis or #Typhoon_19 during the sampled time frame. frame, but could be the n th retweet for that particular tweet. From our findings, the highest number of original tweets from a single account was 50 out of the original 145 tweets. They were sent by @NHKWORLD_News, a public broadcaster in Japan (Figure 1 ). This was followed by 25 tweets created through a Twitter handle name @EarthUncutTV; which is leading online portal, documenting extreme natural events or disasters in the Asia Pacific region and around the world. Around half the tweets within the 145 original tweets were made by individuals. This was determined based on their handle IDs. The total aggregated number of retweets we found from our time frame was 19,582. Amongst them, 71% were retweets which originated from @NHKWORLD_News ( Figure 2 ). @EarthUncutTV covered 16% of total retweets while 13% came from other origins. The results show that within our sample framework, @NHKWORLD_News is the most dominant tweeter account. The ranking of retweets exceeding one hundred per source is shown in Figure 3 . The most retweeted tweet recorded 1,870 retweets. The original tweet was posted by @NHKWORLD_News saying "Hagibis is expected to bring violent winds to some areas. A maximum wind speed of 144 kilometers for the Kanto area, and 126 kilometers for the Tohoku region." The second most retweeted one was posted by an individual account saying "Ghost town //Shibuya at 6pm before a typhoon" with a photo taken in Shibuya City. The top ten retweeted tweets are shown in Table 3 . Names in italics stand for region or city name. Table Table Table Table 3 The 5 th and 10 th tweets mentioned evacuation orders issued by local municipalities while other tweets shared the event as it happened in different places around Japan, or they reported on its effects. We found only three among 145 original tweets that captured the emergency alert issued by JMA (Figure 4 ). Those three were tweeted by individuals and generated only a few retweets. The emergency alert was described in Japanese which is translated as "Special alert at level 5 for heavy rain has been issued to Tokyo area. Level 5 alert means the highest level of warning. Take action for saving your life. Some areas have been newly added under the special alert. Check TV, radio, or local municipalities' information," as of October 12, 21:27 pm. The alert requested people to refer to information from local municipalities but all of it was in Japanese. Our constant monitoring of the event on the internet also reveals that JMA was giving constant updates of the real-time weather status and alerts on its website. This information was provided in English. However, most of it did not appear in our sample frame as it did not utilize Twitter in English. As the Japanese disaster management system also grants direct access of emergency information from JMA to the population, an emergency alert was sent to people's cell phones from JMA, but this was written in Japanese ( Figure 4) . Hence, people who do not speak Japanese may not have understood this, and the real time information could only be solicited by a few who originally knew about such a website or the information provided. On the other hand, much evacuation information is issued by municipal government. We found five tweets mentioning evacuation orders. Two of those were the J o u r n a l P r e -p r o o f @NHKWORLD_News (italics refers to location), as follows. As to take [60] . For example, a Dutch man, who had been living in Japan for a half year, received an emergency message but he could not find information where to evacuate. Another American woman was supported by her friend in translating disaster information. Emergency situations generate exceptional communication features which contravene the norm [61] . Therefore, a communication ecosystem forms and breaks up every time a disaster occurs. We observed dynamic characteristics of crisis communication structure with the involvement of various organizations [62] and tools [52] . The tools greatly hinge on social media which have become essential for digital crisis communication [55] . However, more research is required to better capture the nature of social media use in crisis communication [63] . Several studies reported that J o u r n a l P r e -p r o o f local governments especially in Thailand and the U.S. are reluctant to utilize social media in disaster situations [64, 65] . They prefer to use traditional media such as TV and radio though those mass media tend to allow only one-way communication during a crisis [66] . As our study shows, information through mass media becomes limited or challenged to meet specific needs due to its broader audience base [67] compared to the multiple-interaction communication style that characterizes social media [17] . Government, as a responsible agency, needs to understand the characteristics of mass media and social media communication and create an appropriate communication strategy or collaboration scheme when preparing for future catastrophes. In the recent Covid-19 crisis, the need for more targeted health information within a community and the importance of strong partnerships across authorities and trusted organizations are being discussed [68] . We hold that more consideration should be paid to the provision of targeted information to those who do not share a common context or background of disaster preparedness. On daily-basis practices in Japan, almost all local municipal governments prepare a so-called "hazard map" and distribute it to residents. Ordinarily, municipalities only issue evacuation orders and do not provide residents with specific information where to evacuate to. From a regular everyday disaster drill, trained residents are expected to know where the evacuation centers are in their area. As people interpret new information based on previously acquired knowledge [69] , a responsible agent should also consider how to provide information to people who haven't had any disaster training. Beside the feature of interactive communication characterizing social media, the pull of multiple information sources and agents creates a sense of assurance and security in coping with or adjusting to events. This was experienced in the 2010 Haiti earthquake, when social media contributed swiftly to creating a common operating picture of the situation through the collection of information from individuals rather than from a hierarchical communication structure [70] . In addition, what is important in crisis communication is who distributes a message and how it is mediated in a population [62] . Our study shows that news sources who already seem to have gained social trust played a role as a keystone species in foreign language communication minorities prioritize personal information sources over the media [69] . These cases suggest that social media promote effective resilience in communication, and that the delivery of information to foreigners in Japan from different language backgrounds and cultures further creates traits where personal connection contributes to information accessibility choices. It is also true that social media-based communication raises situation awareness [72] by receiving warning alerts as well as emotionally-oriented messages such as from friends and relatives [62] . Language barrier is likely to have been an obstacle that actually helped to shape the ecosystem identified in this study. During Typhoon Hagibis, responsible agents sent warning messages only in Japanese language as shown in Figure 4 , so this could have been a major factor in harnessing the full potential of social media for disaster communication. Results from this study reveal a lack of local knowledge or reference points across Twitter communication, which shapes discussion within the ecosystem. For instance, in the case of the 2015 Chennai flooding in India, tweets covered subjects such as road updates, shelters, family searches and emergency contact information [73] , none of which were much referred to in our sampled tweets. While in the 2012 and 2013 Philippine flooding, the most tweeted topic was prayer and rescue, followed by traffic updates and weather reports with more than half the tweets written in English [74] . Moreover, around 40% of those tweets were secondary information, i.e., retweets and information from other sources [75] . Our study recognizes the importance of secondary information. Information exchange between agencies and affected residents and tourists assists in reducing risks such as, for example, during Hurricane Sandy in 2012, where the top five Twitter accounts with a high number of followers were storm related organizations posting relevant news and the status of the hurricane [72] . As a remote actor becomes a source of secondary information, however, such information should guide people in what they should do next, rather than just point to the disaster situation. In this context, we note the importance of instructional messages whose content guides risk reduction behaviors [76] . An example of such a message is "do not drive a car" which was tweeted by police during a snowstorm in Lexington [65] . It tells what specific life-saving action should be taken [77] . Generally, however, there appears to be little investigation of instructional information in crisis communication [65] . The findings of this study imply that there could be two types of disaster information: a) risk information that refers to the potential effect of the disaster, i.e., emergency alert or warning, and b) action-oriented information that carries instructions for reducing the risk, i.e., an evacuation order and itinerary to be followed. Risk information is published by remote actors while action-oriented information is provided by responsible agents. Both types of information must contain local knowledge and allow easy access when people search for that information [78] . In order to make sure people can reach instructional messages, it is not enough to just point to the information on organizational websites [65] , as JMA did during Typhoon Hagibis. Instead, "localized hashtags" [64, 65, 79] can support people to find life-saving information as quickly as possible. A previously cited study argues that social media can be useful when sending a request for help [67] . "Typhoon 19 damage of Nagano" is an example of a localized hashtag, assisting a fire brigade in Nagano prefecture utilizing Twitter to respond to calls for help from residents during Typhoon Hagibis 9 . A localized hashtag can also be used as a point of reference where people can find relevant information [65] . 9 last accessed on July 20, 2020, http://www.nhk.or.jp/politics/articles/lastweek/25652.html. In future disaster communication, local information should be distributed using geographical information with instructions as to what to do next [62] . Responsible agencies should take this into account, while also analyzing which information was shared across social media in previous disaster cases [64] . Furthermore, cultural differences can also be considered, as Western culture emphasizes individual action [68] while Asian cultures prioritize collective actions or community commitment. As this may generate behavioral differences among Japanese and foreign residents or tourists, further investigation will be beneficial. Based on information ecology framework and literature review, the structure of resilient communication ecosystem is proposed and verified through empirical data analysis. The resilient communication ecosystem is structured with heterogeneous actors who could be a driving force for collaboration or coevolution. The empirical case revealed that a media source might transmit warnings and evacuation orders through social media but that such information does not contain points of reference. Limited delivery of such information particularly in the English language results in confusion to non-Japanese communities who need it most. Based on study results and discussions, we suggest that in any disaster a form of ecosystem is spontaneously generated. Municipalities, which are often responsible agents should (1) produce instructional information in foreign languages on social media, (2) transfer such information through collaboration with transmitters who may have a strong base on social media to assist translating it to reach the wider audience, and (3) examine, in association with the mass media and weather forecast agencies, the use of local hashtags in social media communication for future disaster preparation. It brings various actors together, creates a sense of locality, and supplements individual efforts in risk reduction. Our empirical data is limited to some extent as we only observed Twitter for one day, October 12, 2019. Again, we intended to extract tweets in English language just before and after the Typhoon Hagibis hit Japan. Hence, we chose hashtag #Hagibis and "#Tyhoon_19" when we were crawling Tweets. We are also aware that our data is based on only a single disaster case and requires further data sets from other disaster events for corroboration. We may consider different characteristics of information and actors in the resilient communication ecosystem in future research. Nevertheless, our findings provide interesting insights into disaster communication which may guide the direction of future research. As instructional information sometimes contains ethics and privacy issues [62] , rules for dealing with specific local or personal information should be taken J o u r n a l P r e -p r o o f into account. Behavioral differences are also worth investigating because of increasing human diversity in Japanese society. The case of Japan reveals the importance and potential of communication as a main mechanism for offering information and response activities to promote resilience and reduce the risk of disasters. Social media space serves as a major platform that brings numerous stakeholders together. This platform creates the main avenue for information sharing on events and offers feedback mechanisms for assessment and improvement. For future disaster preparedness, then, we may benefit from a closer understanding of the nature of a resilient communication ecosystem, its structure and the knowledge that it allows us to share. Resiliency in tourism transportation: Case studies of Japanese railway companies preparing for the 2020 Tokyo Olympics A study of the disaster management framework of Japan Effect of tsunami drill experience on evacuation behavior after the onset of the Great East Japan Earthquake A framework for regional disaster risk reduction for foreign residents (written in Japanese) Japan's foreign population hitting a record high The Tourism Nation Promotion Basic Plan Japan's Immigration Policies Put to the Test, in nippon Japan's immigration chief optimistic asylum and visa woes will improve in 2020 Immigration Services Agency of Japan, Guidelines for Permission for Permanent Residence Sustaining life during the early stages of disaster relief with a frugal information system: learning from the great east Japan earthquake. Communications Magazine, IEEE Who would be willing to lend their public servants to disaster-impacted local governments? An empirical investigation into public attitudes in Disaster Message Not Getting Through to Foreign Residents Hokkaido quake reveals Japan is woefully unprepared to help foreign tourists in times of disaster Media Preference, Information Needs, and the Language Proficiency of Foreigners in Japan after the 2011 Great East Japan Earthquake Social Media Communication During Disease Outbreaks: Findings and Recommendations, in Social Media Use in Crisis and Risk Communication, H. Harald and B. Klas, Editors Social Media Use in Crises and Risks: An Introduction to the Collection, in Social Media Use in Crisis and Risk Communication, H. Harald and B. Klas, Editors Social media, trust, and disaster: Does trust in public and nonprofit organizations explain social media use during a disaster? Social Media Use during Disasters: A Review of the Knowledge Base and Gaps. 2012, National Consortium for the Study of Terrorism and Responses to Terrorism Building resilience through effective disaster management: An information ecology perspective Information Ecologies Using Technology with Heart Resilience and stability of ecological systems Social-ecological resilience to coastal disasters Understanding Communication Ecologies to Bridge Communication Research and Community Action Disaster communication ecology and community resilience perceptions following the 2013 central Illinois tornadoes The Centrality of Communication and Media in Fostering Community Resilience:A Framework for Assessment and Intervention Building resilience: Social capital in post-disaster recovery Community Disaster Resilience and the Rural Resilience Index Community Resilience as a Metaphor, Theory, Set of Capacities, and Strategy for Disaster Readiness Guidelines for conducting systematic mapping studies in software engineering: An update. Information and Software Technology Software product line testing-a systematic mapping study. Information and Software Technology Systematic literature reviews in software engineering -A systematic literature review. Information and Software Technology Microblogging during two natural hazards events: what twitter may contribute to situational awareness An Analysis of the Norwegian Twitter-sphere during and in the Aftermath of the 22 Information Ecology of a University Department Understanding social media data for disaster management. Natural Hazards Flows of Water and Information: Reconstructing Online Communication During the 2013 European Floods in Austria Lessons from Ebola affected communities: being prepared for future health crises Information and Communication Technology for Disaster Risk Management in Japan: How Digital Solutions are Leveraged to Increase Resilience through Improving Early Warnings and Disaster Information Sharing Government-communities collaboration in disaster management activity: Investigation in the current flood disaster management policy in Thailand Twitter in the cross fire-the use of social media in the Westgate Mall terror attack in Kenya Municipal Government Communications: The Case of Local Government Communications. Strategic Communications Management Media coverage of the Ebola Virus Disease: A content analytical study of The Guardian and Daily Trust newspapers, in The Power of the Media in Health Communication Blood and Security during the Norway Attacks: Authorities' Twitter Activity and Silence, in Social Media Use in Crisis and Risk Communication, H. Harald and B. Klas, Editors Social media and disaster communication: A case study of Cyclone Winston Challenges and obstacles in sharing and coordinating information during multi-agency disaster response: Propositions from field exercises. Information Systems Frontiers Indigenous Institutions and Their Role in Disaster Risk Reduction and Resilience: Evidence from the 2009 Tsunami in American Samoa Emergent use of social media: a new age of opportunity for disaster resilience The role of data and information exchanges in transport system disaster recovery: A New Zealand case study Institutional vs. non-institutional use of social media during emergency response: A case of twitter in 2014 Australian bush fire Social Movements as Information Ecologies: Exploring the Coevolution of Multiple Internet Technologies for Activism Crowdsourced mapping in crisis zones: collaboration, organisation and impact Role of women as risk communicators to enhance disaster resilience of Bandung, Indonesia. Natural Hazards Providing real-time assistance in disaster relief by leveraging crowdsourcing power. Personal and Ubiquitous Computing Digitally enabled disaster response: the emergence of social media as boundary objects in a flooding disaster Why we twitter: understanding microblogging usage and communities Social media usage patterns during natural hazards Securing communication channels in severe disaster situations -Lessons from a Japanese Earthquake. in Information Systems for Crisis Response and Management Using thematic analysis in psychology Emergency Warnings and Expat Confusion in Typhoon Hagibis, in NHK The design of a Dynamic Emergency Response Management Journal of Information Technology Theory and Application Social Media in Disaster Risk Reduction and Crisis Management Social media for knowledge-sharing: A systematic literature review A community-based approach to sharing knowledge before, during, and after crisis events: A case study from Thailand Thor visits Lexington: Exploration of the knowledge-sharing gap and risk management learning in social media during multiple winter storms The role of media in crisis management: A case study of Azarbayejan earthquake Social media and disasters: a functional framework for social media use in disaster planning, response, and research Using social and behavioural science to support COVID-19 pandemic response Crisis communication, race, and natural disasters Emergency knowledge management and social media technologies: A case study of the 2010 Haitian earthquake Exploring the Use of Social Media During the 2014 Flood in Malaysia Understanding the efficiency of social media based crisis communication during hurricane Sandy The Role of Social Media for Collective Behavior Development in Response to Natural Disasters Understanding the Behavior of Filipino Twitter Users during Disaster Communicating on Twitter during a J o u r n a l P r e -p r o o f disaster: An analysis of tweets during Typhoon Haiyan in the Philippines The Instructional Dynamic of Risk and Crisis Communication: Distinguishing Instructional Messages from Dialogue. Review of Communication Conceptualizing Crisis Communication, in Handbook of Risk and Crisis Communication Social media and disaster management: Case of the north and south Kivu regions in the Democratic Republic of the Congo Social media and crisis management: CERC, search strategies, and Twitter content This work was supported by JSPS KAKENHI Grant Number JP19K20635.