Integrating Crisis-Mapping into Coast Guard Incident Response

P a g e 1 Integrating Crisis-Mapping into Coast Guard Incident Response Using Crisis-Mapping to Incorporate Information From Social Media During Disaster Response Evan Twarog US Coast Guard Academy April 2017 Contents: Final Deliverable for 8470

P a g e 2 Table of Contents Abstract... 3 Purpose... 3 Background... 4 Filling the Information Gap... 4 The Rise of Crisis-Mapping and Digital Humanitarianism... 5 Crisis-Mapping Applications... 6 Coast Guard and Crisis-Mapping... 7 Approaches to Crisis-Mapping... 9 Developing Coast Guard Crisis-Mapping Software... 10 Forming Partnerships... 11 Partnerships... 12 Characteristics of DHOs... 12 Dangers of Cherry-Picking DHOs... 13 Noteworthy DHOs... 14 Challenges When Working With DHOs... 16 Agencies Involved in Crisis-Mapping... 17 Incorporating Information Into the Incident Management Software System... 18 Information Flow in IMSS... 19 Theoretical Case Study... 19 Findings... 20 Key Recommendations... 21 Author s Note... 24 Works Cited... 25 About the Author Evan Twarog is a cadet at the United States Coast Guard Academy in New London, CT where he is majoring in Electrical Engineering. Cadet Twarog is focusing on electrical systems while taking courses in Geospatial Information Systems among other topics. His piece Data-Driven Disaster Response: How Crisis-Mapping can Improve Coast Guard Incident Response is being published as a Professional Note in the US Naval Institute s Proceedings magazine. Beyond the Academy, he is a cyclist for the US Military Endurance Sports Cycling Team and a volunteer Firefighter/EMT. He can be reached at Evan.J.Twarog@uscga.edu

P a g e 3 Abstract: With the emergence of the Digital Humanitarian community following the 2010 Haitian Earthquake, social media is becoming an increasingly important component of disaster response. Survivors are turning to social media to request aid and assistance more and more with each passing year. During the rise of crisis-mapping and social media, the Coast Guard has struggled to keep pace with an ever changing digital world. An information gap within the Coast Guard has opened up: The service doesn t have the ability to listen to disaster survivors through social media. Crisismapping, a tool developed by the Digital Humanitarian community, offers the potential to visualize information from social media on an easy-to-read, real-time platform. It can close the Coast Guard s information gap. Up to this point, crisis-mapping has not even been in the Coast Guard s vocabulary. While there are a number of solutions to how crisis-mapping can be incorporated into its incident response framework, the most realistic solution is to activate digital humanitarian organizations and utilize the information they gather. Platforms like the IMSS have the ability to directly feed information into the Coast Guard s Common Operational Picture. Ultimately, the most important point to highlight to leaders within the service is that crisis-mapping is a tool that the Coast Guard has available to it. While it is easy to only look to time-tested software platforms during disasters, crisis-maps are an innovative solution the service can use to close its information gap. Purpose: In the days following Hurricane Katrina making landfall in the Gulf of Mexico, the Coast Guard mobilized over 5,600 members to assist in search and rescue, marine pollution response and maritime commerce management (Price 2). Ten percent of the service, and forty percent of its helicopters deployed in response. As one of the largest natural disasters to strike in US history, over 60,000 people needed to be rescued from rooftops, hospitals and nursing homes. In many ways, the response to Hurricane Katrina is viewed as one of the Coast Guard s finest hours: The service is credited with saving more than 33,500 people, 24,135 of which were lifted off rooftops (Price 2). Rear Admiral Robert F. Duncan wrote in the months following: This is a defining moment, from my view, for the United States Coast Guard. I ve said for a number of years now that the Marines are making combat veterans in Fallujah and Ramadi and Anbar Providence, and it s painful, it s awful, it s a national treasure that we re losing, but they re a better fighting force, they re a better Marine Corps for having combat veterans in the 21st century that will be with them for a long time. That s what I think this did for the Coast Guard. Ten years from now second class petty officers, lieutenants, you know they re going to be the chiefs and the commanders that are in positions of significant responsibility around our service. They re Katrina veterans. They know how it was done. They know the kind of centralized command, decentralized execution that worked very well here that gave us international acclaim and we ll be a better Coast Guard for that (Price 2). Undoubtedly, the Coast Guard came out of the response to Katrina a stronger service. That being said, the way in which the world communicates has dramatically changed since 2004. The first IPhone wasn t marketed for another two years after Katrina, and Facebook was only 6 months

P a g e 4 old. It wasn t until 2008 when it broke the 100 million user mark (Carlson). Today, billions of people use social media as their primary form of communication. The way the world uses information and how it spreads is fundamentally different than it was during Katrina. Despite this shift in communication, the Coast Guard has largely remained unchanged in the way we would respond to another Hurricane Katrina. This paper is aimed at addressing a fundamental issue facing Coast Guard incident management: The service doesn t have the ability to listen to information posted on social media. This needs to change before the Coast Guard finds itself responding to another Katrina. Instead of being one of the service s finest hours, another Katrina tomorrow could turn out much differently. Background: With the world changing at an increasingly rapid pace, the Coast Guard has and will continue to struggle to keep pace with emerging technologies. As social media and smartphones have become a part of everyday life, the way in which the world communicates has fundamentally changed. It s become so engrained in daily life that populations are increasingly turning to it following major disasters. This was highlighted by Hurricane Sandy when over 20 million Hurricane Sandy - related tweets were posted in the days surrounding its landfall (Lessons 10). When traditional landlines and cellphone coverage got knocked out by the storm, many of these areas retained 3G access, forcing those affected to communicate using social media. This is a very real scenario that could pan out in the future. In coming years, this trend of increasingly relying on social media is only going to become more engrained throughout the world. While the Coast Guard isn t alone in being slow to adopt new technologies, it needs to be aware that this is one of its weaknesses. How many Admirals know that most teenagers communicate using SnapChat? The Coast Guard has an information problem. Today, the service has no means through which it can listen to what s happening on Facebook or Twitter during a disaster. When another Sandy strikes again, the Coast Guard will be left that much less aware of what s happening on the ground. The service is not going to have an excuse for why it didn t know there were people left stranded on roofs, and frankly, people could die because of this. Filling the Information-Gap: Faced with the challenge of closing the social media information-gap, the Coast Guard needs to develop a system or process that will: 1. Gather information from major social media sites (Facebook, Twitter, etc.) 2. Visualize this information in a meaningful way 3. Integrate it into the Coast Guard s current Incident Management framework Mining information from social media involves searching through posts for specific keywords or Hashtags that are relevant to the incident. These keywords might include Help, #HurricaneSandy, Trapped, Rescue among dozens of others. By looking for posts with these

P a g e 5 specific words in them, the system would identify which posts are potentially relevant to response efforts. This becomes a complicated issue during major disasters like Hurricane Sandy because of the sheer volume of posts. It s extremely difficult to identify requests for help when millions of people are posting to social media. After collecting this information, the next step will be visualizing it in a way that is meaningful to responders. While there are a number of different formats that this visualization could come in (word map, graphs, etc.), the most valuable is mapping. A large percentage of social media posts today have a geotag of some sort attached to them. This would allow responders to create a map of how the crisis is unfolding. The final link in the process is incorporating this visualization into the Coast Guard s existing incident management framework. The service already has enough obsolete or poorly designed software suites. Unless this information is smoothly incorporated with the Coast Guard s current systems, it s going to be pushed aside during the chaos of a disaster. The Rise of Crisis-Mapping and Digital Humanitarianism: With these three requirements in mind, crisis-mapping is perhaps the best tool the Coast Guard can use to fill its information gap. It mines information from social media, displays it in a meaningful way and can fit in with existing Coast Guard geospatial software packages. To understand the potential of crisis-mapping, it would help to understand the rise of digital humanitarianism and crisis-mapping. Digital humanitarianism and crisis-mapping emerged out of the rubble of the 2010 Haitian Earthquake. In the aftermath of the 7.0 magnitude earthquake, a group of computer scientists and political scientists based out of Tufts University in Boston, MA launched an effort to gather information through SMS and Facebook to help response efforts (Meier Crisis). Using the Ushahidi mapping platform (which means witness in Swahili), the group began to collect and map text messages and social media posts. Some of these posts included: My name is [...] I m not dead. I am under the rubbles in University Caraibes, which is in [...]. Please come and get me? My name is [...] my brother is working in Unicef and I live in [...] I have 2 people that are still alive under the building still! Send help! (Meier Humanitarian) The Marine Corps, international aid organizations, and urban search and rescue teams all used the map to help guide their efforts. By the time response efforts ended, the team plotted and translated over 2000 individual reports (Meier Crisis).

P a g e 6 In the weeks following the response, the Marine Corps and FEMA were both vocal in their support of the work being done at Tufts. Lt. Gen. H. Steven Blum, the second in command at NORTHCOM visited the group, noting We all need to learn from you. Additionally, the Tuft s leader, Patrick Meier received an email in the weeks following noting: I cannot overemphasize to you the work of the Ushahidi/Haiti [Crisis Map] has provided. It is saving lives every day. I wish I had time to document to you every example, but there are too many and our operation is moving too fast. Here is one from the 22 MEU (Marine Expeditionary Unit): We had data on an area outside of Grand Goave needing help. Today, we sent an assessment team out there to validate their news and everything checked out. While the team was out there, they found two old women and a young girl Image 1: The Ushahidi Haitian Earthquake crisis map collected thousands of requests for help over the course of several weeks, launching the Digital Humanitarian movement. with serious injuries from the earthquake. All were evacuated. Your site saved these people s lives (Meier Humanitarian). Additionally, Craig Fugate, the FEMA administrator tweeted: Keep up the good work!! You are making the biggest difference of anything I have seen out there in the open source world (Meier Humanitarian). Out of the response in Haiti emerged a Digital Humanitarian Revolution. In the months and years ensuing, dozens of crisis-mapping organizations and networks began making an impact during a variety of incidents. Crisis-Mapping Applications: Crisis-mapping has emerged as a powerful tool for first responders since it first emerged in 2010. To date, crisis mapping has been used during hurricanes, earthquakes, the response to the disappearance of flight MH370, the humanitarian crisis in Syria, wildfires in Russia, among dozens of other settings (Meier Humanitarian). The value that crisis-mapping brings to response efforts has been proved time after time. FEMA, the Red Cross and the United Nations have all launched their own initiatives to incorporate crisis-mapping into their response framework. Part of what makes crisis-mapping such a valuable tool is the flexibility it provides. Each disaster is unique in some

P a g e 7 way, and with each activation (when an organization begins combing social media for information), new applications emerge. Some noteworthy disasters where crisis-mapping efforts have proved valuable include: Hurricane Sandy (2012): When Hurricane Sandy struck the eastern seaboard, dozens of organizations turned to social media to help in their response efforts. Most notably, Hurricane Sandy took about landline and cellular reception, but survivors were still able to access social media. This is the largest crisis-mapping activation in US history, and it was used to assess damage and help survivors get aid. Uttarakhand Floods (2013): In 2013, major flooding in Northern India killed over 5700 people and caused widespread devastation. The Humanitarian OpenStreetMap Team (HOT) mapped over 4000 villages affected by the flooding using satellite imagery (Uttarakhand). This has been used to help determine how to access remote communities since large sections of road Image 2: Flooding in Northern India in 2013 have been entirely wiped out. West African Ebola Outbreak (2014-2015): Over the course of nearly a year, crisismapping was used to help track the progression of the Ebola outbreak throughout Africa (West Africa). Nepal Earthquake (2015): The HOT responded within 72 hours of the earthquake with over 4000 volunteers working to assess damage. Interestingly, the GISCorps developed the Nepal Vulnerability Index which helped to identify which areas are most prone to malnutrition and disease outbreaks (Nepal). Additionally, the Interactive Winterization Dashboard helped to identify communities prone to being heavily impacted by the lack of warm shelter (Nepal). Most noteworthy, crisis-mapping was used similarly to the Haitian Earthquake to help save lives of those trapped by the quake. These responses only represent a small percentage of the activations that have happened over the last 7 years. However, these four in particular highlight how crisis-mapping can be used in a number of ways, both immediately to help coordinate Search and Rescue efforts and over a much longer timeframe to track the progression of a disease. The Coast Guard and Crisis Mapping: Crisis-mapping offers a massive set of opportunities for the Coast Guard. As a hugely diverse service, the Coast Guard is exposed to a variety of missions ranging from ice-breaking to search and rescue to environmental response to maritime security. This ultimately means the service could be thrown into a wide variety of crises at any given point. During Katrina, the Coast Guard was responsible not only for search and rescue, but also environmental response and maritime commerce management. Crisis mapping offers the potential to improve the way the service conducts each of these missions. Social media could be used to locate where survivors are trapped on rooftops, identify where spills are extending to, and assess coastal damage.

P a g e 8 As of right now, crisis-mapping isn t even in the Coast Guard s vocabulary. Despite the value it can provide, the Coast Guard has yet to seek out how it can incorporate information from social media and crisis-mapping into its response efforts. In the last seven years, the Coast Guard has responded to several national-level crises, notably the Deepwater Horizon Oil Spill and Hurricane Sandy, in which digital humanitarians were also activated. Frustratingly, the service didn t take advantage of the value these organizations offered. Deepwater Horizon and Crowdsourced Imagery: After Deepwater Horizon, the Coast Guard missed an opportunity to work with the Public Laboratory for Open Technology and Science (PLOTS). Over the course of 7 weeks, the lab organized 36 sorties of volunteers that launched weather balloons collecting 3cm resolution photographs. During this time, these volunteers collected 11,000 images that could have been used to identify impacted wildlife and track the extent of the oil spill (Waldman Formal). As one responder (now Lieutenant, USCG) explained: The Coast Guard got a lot right during [Deepwater Horizon]. It also got a lot wrong. We largely failed at tracking where the oil was going There were times when I d have to look at photos from the news of the oil spill, estimate where it was extending and then direct the asset there. Satellite imagery, the images from PLOTS volunteers and the Coast Guard s established response framework could have been used in unison to provide a more accurate common operational picture. Hurricane Sandy: Beyond Deepwater Horizon, the response to Hurricane Sandy proved to be another major case where the digital humanitarian community activated. One of the unique components of the response to Sandy was that landlines and cellphone towers failed while data-coverage remained active in many regions. People could use social media to reach out for help, but couldn t call 911 (Lessons). Because traditional methods of communicating with the public failed in places, power companies, FEMA, the Red Cross, local fire departments, and other organizations all Image 3: Google s Hurricane Sandy crisis map helped to track the progression turned to social media of the storm in real time to work with the public. The New York Fire Department (FDNY), for example, initially instructed citizens not to

P a g e 9 tweet emergency calls. As the storm progressed, the FDNY began responding to and directing the information to dispatchers (Lessons). DHOs that activated include the Standby Volunteer Task Force, the Digital Humanitarian Network, Facebook Safety Check among dozens of others. Multiple grassroots organizations popped up specifically to respond to Sandy (Meier Humanitarians). Hurricane Sandy represented the largest digital humanitarian response effort to date. During all of this, the Coast Guard failed to capitalize on the potential of the digital humanitarian community and crisis-mapping. These are just two instances where the Coast Guard could have benefitted from crisis-mapping efforts. Information Needs Current Collection Methods Crisis-Mapping Impact Ground Damage *Aircraft Fly-Overs High Assessment *Working With Other agencies Search and Rescue *Satellite Imagery *"Boots on the Ground" *Working with Other Agencies *Aircraft Fly-Overs *911/VHM-FM Radio Calls Very High HAZMAT Management *National Response Center Minimal Terrorism Threats *Working With Other agencies Minimal *Internal Intelligence Food/Water/Aid Needs *Working With Other agencies Very High Oil Spill Modelling *National Response Center Minimal Table 1: The service has a variety of different information needs which can be addressed with varying levels of impact by crisis-mapping Ultimately, the Coast Guard is a highly diverse service. Table 1 illustrates the wide variety of information needs and collection methods the service regularly has during a major incident, as well as the impact potential that crisis-mapping has. With its 11 different missions, the service is forced to deal a vast variety of challenges ranging from terrorism to climate change to international organized crime to oil spill response. While crisis-mapping will not impact every mission of the service, Table 1 highlights some areas where it can be valuable. Approaches to Crisis-Mapping: Crisis-mapping has the potential to be a valuable tool to Coast Guard incident management. With this being noted, there are two main ways that the service can begin utilizing it: It can design its own software and train personnel to do crisis-mapping or it can partner with other crisis-mapping organizations and agencies and utilize the information they extract. The biggest difference between the two paths is who gathers the information.

P a g e 10 Avenue Advantages Disadvantages *Limits risk of accidental release of PII to the public *Can tailor the software to fit service's needs Internal Software Development Partnering with DHOs/government agencies *Innovation prioritized, keep pace with changes *Draws on experince of world experts *Greater map flexibility *Frees IM personnel to focus on IM vs. Crisis-mapping *Little service knowledge about crisismapping *Software not used enough to justify costs *Constant training needed, specialized expertise required *Coast Guard already slow to adapt technologies *Information security/accuracy more difficult to guarantee *Reliant on efforts of volunteers Table 2: The two paths along with their advantages and disadvantages Table 2 shows the advantages and disadvantages associated with each option. This paper will argue that the Coast Guard should seek to work with digital humanitarians and other agencies that are involved with crisis-mapping. Developing Coast Guard Crisis-Mapping Software: The first proposed option is to develop a service-specific crisis-mapping software package. Ultimately, the collection of information, incorporating it into the Common Operational Picture and acting on it would happen within Coast Guard operations. Specific personnel from each sector s Incident Management Division (IMD) would be trained in how to use the software and during a National-Level incident, an individual would be assigned to run it as a collateral duty. The main advantage of this option is that the Coast Guard doesn t have to rely on other organizations and agencies for information. With an entirely in-house operation, the service would fully understand the capabilities of the system. There is Image 4: Purdue s The Social Media Analytics and Reporting little guess-work, and there are no Toolkit (SMART) software is a platform the Coast Guard could use volunteers to rely on for potentially to conduct its own social media monitoring and crisis-mapping critical information. Additionally, the service can tailor the software platform to fit its needs. If the Coast Guard intends to use the platform for only extracting Search and Rescue information, it can specifically design the software for that. The issue with this is that there is no way for the service to predict what disaster could

P a g e 11 strike next. It could be anything from a dirty bomb being detonated in Times Square to an arctic oil spill to a devastating hurricane. Specialization could be a hindrance. There are a number of disadvantages of this proposal that are important to note. The reason the Coast Guard has an Information Gap is because the service can t keep pace with the changes happening in technology. The Coast Guard is inherently slow to innovate because it is a federal agency with restraints placed on it. If the Coast Guard is already behind the eight ball, it s going to be very difficult, if not impossible for the service to keep pace with the rapid changes happening within the crisis-mapping community. The other issue with pursuing this avenue is that the Coast Guard already has too many software packages that are a) very expensive b) poorly designed and c) rarely used. The Coast Guard doesn t see enough cases on an annual basis where the software would be used to keep the personnel who use it well-trained. Additionally, the Coast Guard hasn t ever been exposed to crisis-mapping. Unlike with the development of SAROPS (the service s Search and Rescue planning software), the Coast Guard doesn t have decades of crisis-mapping experience to outline what a good software would look like. It would be difficult to develop a good package in a realm the service knows very little about. While the service might think the information the software is gathering, it might not be as accurate as that gathered by more innovative grassroots DHOs. While this could become an option that the Coast Guard pursues over the long-term, it doesn t make sense in near-term and until crisis-mapping becomes something the Coast Guard starts using regularly. Forming Partnerships: The second proposal is fundamentally different than the first because it draws on the expertise and experience of outside organizations and agencies to gather the information. During a major disaster, an individual within Coast Guard Incident Management would activate a digital response where multiple digital humanitarian organizations and agencies would begin crisismapping. The service would then use the information to help its response efforts. There are a few noteworthy advantages of this approach. The first is that it would draw on the expertise of a wide body of computer scientists, humanitarians and GIS experts to gather the information. Someone who works with GIS or Python coding is going to get better information than a Lieutenant who is juggling a half-dozen collaterals. While the service might feel more comfortable with the information generated by a customized software, experts are more likely to get reliable results. Beyond this, the crisis-mapping world is constantly changing as new softwares emerge that validate a post s accuracy, improve filtering results, etc. The digital humanitarian community doesn t have the constraints that the service does, so it can innovate much quicker. With this innovation comes greater flexibility to generate new products. During the Nepalese Earthquake in 2015, crisis-mappers helped to identify which communities were vulnerable to the dropping temperatures of the winter, potential outbreaks of disease and malnutrition (Nepal). Each disaster is unique. The need for innovative solutions cannot be stressed enough. Finally, this approach allows incident management experts to do incident management, not troubleshoot a new set of code. It frees Coast Guard personnel to focus on what they do best. The main disadvantage of this approach is that it relies on a community that is rapidly evolving. There is no way to predict the future of digital humanitarianism. It s possible that in 10 years, crisis-mapping has evolved to the point where this approach simply doesn t work anymore. Not only this, but the organizations discussed later in this paper could fizzle and

P a g e 12 disappear. The Coast Guard has traditionally focused on risk-management, increasing certainty. The biggest challenge with this approach is that the service would have to deal with uncertainty. Partnerships: Of the two paths discussed above, the best solution is to draw upon the expertise of Digital Humanitarian Organizations and other government agencies to gather information. The single greatest obstacle facing this path is understanding the organizations the service would be working with. It s critical incident responders understand how these organizations function, what drives their work, what their limits and capabilities are and what the challenges of working with them are. This section will discuss Digital Humanitarian Organizations in depth as well as other government agencies involved with crisis-mapping. Characteristics of Digital Humanitarian Organizations Up to this point, the vast majority of crisis-maps have been created by coalitions of online volunteers working for a grassroots organization. These grassroots organizations have been termed digital humanitarian organizations. According to the Digital Humanitarian Network, a Digital Humanitarian Organization (DHO) is: A grassroots organization that mobilizes a large number of individuals that share a set of open tools, practices, and ethical standards to create a collective intelligence for providing information as aid. While grassroots organizations have some negative connotations associated with them within the military, it s important to understand the value that they offer the Coast Guard. In fact, this report can t stress enough the value they offer. DHOs have four components that enable them to effectively conduct their work: 1. Technologies: They ve developed a host of hardware and software they can use to gather and process information into a usable product. Most notably, these communities use open source software like OpenStreetMap to power their response efforts. 2. Communities: They have a network of GIS experts, traditional humanitarian experts, computer scientists, linguistic experts (the information often has to be translated) and other concerned Samaritans that they can mobilize in response to a specific event. 3. Practices: They apply their tools using a set of best practices for maximizing information timeliness and accuracy. 4. Ethical Standards: With the sensitivity of the information these organizations deal with, it s important to note that they uphold a defined set of ethical values. These values govern which events they re willing to mobilize for (they won t mobilize on behalf of a particular side of an armed conflict), how they deal with information privacy, and how they validate information accuracy (Waldman Volunteer). In general, these organizations are nimble and composed of hugely talented and motivated individuals. Reflecting back on its origins, the Digital Humanitarian community began as a volunteer effort made of political and computer scientists in a basement apartment. In many ways, this has changed dramatically and it hasn t changed at all. It is still a coalition of volunteers, but in the seven years since Haiti, the community has built up a new software infrastructure, and

P a g e 13 established strict ethical conduct standards that it holds itself to. While the community has made mistakes in the past, the value it s brought to the table has far outweighed its shortcomings. With each year DHOs become more established and learn from their previous response efforts. There is no reason to believe this community is going to disappear in the future. Instead of ignoring their presence, it s in the Coast Guard s interest to work as closely as possible with these organizations. Beyond these four operational facets, similar cultural values drive DHOs work: Open-source ideology: DHOs widely operate with the belief that information and data should be available to the public. While open-source standards might be widely accepted, this isn t to say that DHOs don t work with restricted information. Each organization has different policies that govern how information is handled, particularly PII. Collaborative Workflow: A huge component of what makes DHOs powerful is that they rely on crowd-sourced efforts. Experts with different backgrounds from around the world collaborate together during response efforts. Collaboration rests at the center of every DHO response. Common means for collaboration include Google Drive, Skype, wikis, and other microtasking platforms (Volunteer and Technical 8).. Altruistic Nature: Digital responders are volunteers and they believe that their collective efforts can make a meaningful impact. Flexible Heirarchy: Because disasters can strike anywhere in the world and can take any form, DHOs value flexibility. Responders need to be able to contribute from anywhere in the world (Volunteer and Technical 8).. Drive to Improve Technical Skills: The work that DHOs do is primarily technical. There s an ongoing need to train future responders in the technical skills needed to make an impact during response efforts. Eagerness to Form Partnerships: Partnerships are mutually beneficial for both DHOs and traditional humanitarian organizations. The information that DHOs mine aids directly in relief efforts that humanitarians are involved in. Additionally, traditional organizations often struggle to move with the speed that DHOs do, which helps them generate an environment of innovation and growth. DHOs benefit from these partnerships because they become more credible and professionalized (Volunteer and Technical 8). These cultural values are comparable to the Coast Guard s Honor, Respect and Devotion to Duty. Altruism ultimately drives volunteers to work on responses into the early hours of the morning. Collaboration and an open-source ideology allow these volunteers to work together effectively. An eagerness to form partnerships multiplies the impact of their work. The Danger of Cherry-Picking DHOs If the Coast Guard decides to pursue the proposed path to using crisis-mapping, it will be easy to try to say The Coast Guard is only going to activate and collaborate with these three organizations during a digital response. While saying this might make it easier to activate a response, it can put incident managers into a mindset that will close off valuable partnerships. As of February, 2017, there are dozens of active digital humanitarian organizations that could provide value to the Coast Guard. As of February, 2017, there are dozens of active digital humanitarian organizations that could provide value to the Coast Guard. Image 5 shows a small number of the dozens of DHOs who have responded in the past.

P a g e 14 Image 5: A small fraction of the number of active Digital Humanitarian Organizations There are a couple of important points to highlight by observing this. More than anything else, the Coast Guard shouldn t attempt to cherry-pick which organizations it should try to work with. With the fluidity of the digital humanitarian community, organizations will inevitably fold, and new ones will pop up. By cherry-picking organizations to work with, the Coast Guard risks picking ones that will fold, or ones that won t be active during the crisis it s responding to. Another problem that arises is that with a national crisis, there may be too many crisis maps to effectively use. It does no one any good if there are 20 different maps that arise out of a hurricane. This is simply too many to manage effectively. The ideal outcome would be to find an organization that networks different DHOs together. Each partner would have a niche, and during an activation, each would contribute with its own specialized expertise. The advantage of this is that the Coast Guard wouldn t have to reach out and organize its own response. It would essentially press the Go button and the network would organize the response internally. Noteworthy DHOs: While there are dozens of DHOs currently active, there are a few organization that are particularly valuable to reach out to. The Digital Humanitarian Network The Digital Humanitarian Network is unique from other organizations in that it is a coalition of DHOs, each with their own niche. Instead of organizing the volunteers itself, DHN works by acting as a coordinating body for a digital response (About). Depending on what information is being posted and what first responders need, DHN will activate different partner organizations to perform specific tasks. There are a few advantages with this model:

P a g e 15 First responders interface with a single organization, making their job easier. As a result, responders are more likely to utilize it because it simplifies their efforts. There is a standard, simple process for activating the Digital Humanitarian Network. Activating is as simple as sending an email or filling out a form, as illustrated by Image 6. DHOs can specialize in a specific task and become much more proficient at it. GISCorps can begin to focus in on a much more specialized niche as compared to the Standby Volunteer Task Force. The DHN can begin to enforce a standard ethical model across a large number of partners. Image 6: The Digital Humanitarian s created a systematic process for activating a Digital Response. Most significantly, a response can be launched by anyone with a single email. Beyond these advantages, DHN is composed of the leading DHOs. Table 3 shows that the vast majority of leading DHOs are partner organizations with DHN. These organizations include The Humanitarian OpenStreetMap Team, Standby Task Force, and Translators Without Borders.

P a g e 16 Easily Activated Table 3: The leading DHOs are partners with the Digital Humanitarian Network DHN Partner Organization/Platform Niche Digital Humanitarian Network of 3+ dozen DHOs, coordinates Network (DHN) Y digital response among these partners n/a Always Team responds with create of Person Finder, Google Crisis Response Available Crisis Map, or Resource Page Y Humanitarian OpenStreetMap Team Crisis-mapping, damage assessments, (HOT) Y information management Y International Network of Crisis Mappers n/a Community of Digital Humanitarians Y Standby Task Force Crisis-mapping, damage assessments, (SBTF) Y information management Y Translators Without Borders (TWB) Y Professional crowdsourced translations Y Y, but fees Data collection and crisis-mapping platform, Ushahidi apply widely used Y With all of this, in mind, the Digital Humanitarian Network would be a natural starting point for anyone within the Coast Guard looking to activate a crisis-mapping response. Ushahidi Ushahidi is a data collection and visualization platform that has been used by Crisis-Mappers in multiple noteworthy deployments. The software can take information from a number of sources including text/sms, surveys, social media and email. More specifically, it s been used to: Collect reports from victims and field staff via SMS, email and social media Organize response efforts Track changes in the field with mapping (Ushahidi)) Ushahidi is noteworthy because it is a tool that s been used by crisis-mappers across organizations. Beyond crisis-mapping, it has been used with election monitoring The software has been deployed 90,000 times across a variety of applications (Ushahidi). While it s unlikely the Coast Guard will directly be using it, the service should be aware of Ushahidi because it is such a widely used platform. Standby Task Force and the Humanitarian OpenStreetMap Team SBTF and HOT are both DHOs that specialize in creating crisis-maps. Both organizations are DHN partners that has been frequently called upon during major responses. They are two of the most influential DHOs and can easily be activated through their websites (Meier Humanitarian)) Challenges When Working With DHOs Perhaps the biggest challenge in incorporating open-source information into the Coast Guard s incident response framework is understanding the cultural differences between DHOs and the Coast Guard. In general, DHOs are small, nimble grassroots organizations that value transparency and innovation (Waldman Volunteer). The Coast Guard, like other government agencies tends to be more conservative in adopting new technologies and favors accuracy and reliability over

P a g e 17 innovation. DHOs wishing to work with the Coast Guard need to understand that it tends to be a slow-moving organization. The Coast Guard needs to understand that DHOs won t always be able to guarantee information quality or accuracy. It is possible to bridge this cultural gap though. Within the government, one of the organizational fears is being in error. Making a mistake can be at a minimum be costly, and at worst, it can end a life. In comparison, the fear with the open-source community is burying an error (Crowley 10). This is important to note, and explains why the Coast Guard in the past has been hesitant to utilize open-source information. Federal agencies like the Coast Guard are legally obligated to provide data that s accurate and reliable. Beyond this, they have to take steps to prevent the release of data that breaches the privacy of citizens, or compromises operational security. The Coast Guard expresses these values in Coast Guard Publication 3-28 where it lays out that the four necessary qualities to its communication system s; flexible and reliable communication systems, an established dissemination Process, timeliness, and accuracy (Incident). The challenge is creating an environment where the values of each organization aren t compromised. The Coast Guard can t compromise its need to reliable, accurate information and DHOs can t compromise their drive for innovation. While this might sound difficult to do, government agencies have partnered with grassroots organizations on at least three occasions the last 7 years. In 2010, the US State Department s Humanitarian Information Unit and the National Geospatial-Intelligence Agency partnered with the Humanitarian OpenStreetMap Team (HOT) to develop remote mapping activities (like those that happened in Haiti). HOT has the ability to crowdsource satellite image processing and NGA and the HIU provide the imagery (Crowley 49). Beyond this, USAID partnered with the Standby Volunteer Task Force and GISCorps to merge geographic data so that entrepreneurial loans in emerging economies could be tracked (Crowley 50). Most notably for the Coast Guard, FEMA partnered with HOT and the Civil Air Patrol to process aerial imagery taken by the after a disaster. The Civil Air Patrol has a fleet of 550 aircraft in every state and as a result, has the potential to be one of the first organizations activated after a disaster (Crowley 51). Neither FEMA nor CAP has the ability to process the images, which is where the Humanitarian OpenStreetMap Team comes into play. This relates directly back to the response mission the Coast Guard has in a disaster-scenario. Agencies Involved with Crisis-Mapping Within the last several years, major agencies have begun to recognize the impact that crisis-mapping can have on their response efforts. These agencies have included the United Nations, FEMA and the Red Cross. The Red Cross Digital Operations Center The Red Cross believes in the need to trust citizens to report on what is happening right in front of them, and that these reports are every bit as reliable as reports by experts, and a whole lot quicker because people have mobile phones now (Department 10). Building off of this, it decided to launch its Digital Operations Center and Digital Volunteers program in 2014. Compared with other agencies, the American Red Cross has taken a unique approach with how it interacts with social media. Starting in 2014, the Red Cross partnered with Dell to open up the Digital Operation Center (DOC) in Washington DC (Department 19). The DOC is a center that is

P a g e 18 designed to manage information being posted to social media during disasters. It utilizes tools like heatmaps to visualize spikes in social media traffic and crisis-mapping to help manage public information and relay information to first responders. The DOC is unique in that it is the first social-media based operation devoted to humanitarian relief. In conjunction with this, the Red Cross announced a Digital Volunteer program to help respond to questions and provide information to the public during disasters (Department 20). What the Red Cross has done is unique in that the DOC represents a holistic social media strategy. The Red Cross recognizes that social media is a two-way road. The public posts information about disasters that responders have an interest in, and responders should be providing accurate and timely information throughout the course of the disaster. Traditionally, first responders have posted to social media, but haven t been on the receiving end of public posts. Digital Humanitarian Organizations have listened to social media, but because of their code of ethics, they don t have the ability to post to social media to influence a response They are silent observers. The DOC merges the best of these two worlds by both listening and posting to social media sites during disasters. To date, the DOC s largest response has been to Hurricane Sandy. Thirty-one volunteers reviewed more than 2.5 million posts, sent 2,386 responses to posts and posted 229 times about needs in affected areas to the Red Cross Mass Care team (whose job is to provide basic needs to disaster victim). Ultimately, 88 of those 229 posts resulted in action on the ground (Department 20). The DOC could be a valuable partner for the Coast Guard to work with in the future, but the service needs to know that this is a partnership it needs to established before an incident happens. Whereas a DHO can be activated by anyone, it will take time and effort to build a relationship with the Red Cross. That being said, the DOC brings a lot to the table that grassroot DHOs can t: The name recognition of the Red Cross, the funding and influence, and the more holistic social media strategy. FEMA GeoPlatform: The Federal Emergency Management Agency (FEMA) uses it GeoPlatform to help provide geospatial data analytics to support its missions. The platform is used mainly in the planning and damage assessment stages, and up to this point, the agency hasn t engaged in crisis-mapping (Vaughan). FEMA has encouraged private companies like Google to post its information publically so that it can be integrated into the platform. Because the information on the GeoPlatform is publically available, survivors can use it to check damage to their homes from a distance. The main difference between the GeoPlatform and the crisis-maps created by DHOs is that the information typically doesn t come from social media. It is valuable information, however this platform doesn t close the Coast Guard s information gap. In other words, it doesn t address the fundamental reason driving the Coast Guard s need to engage in crisis-mapping. Incorporating Information With the Incident Management Software System The Incident Management Software System (IMSS) is one of the Coast Guard s incident management platforms. IMSS allows the service to coordinate its resources, maintain a common operational picture (COP), and fill out the necessary paperwork, among hundreds of other smaller

P a g e 19 tasks. Because the IMSS is widely used throughout the service, it offers a path through which crisismapping can begin to be incorporated into incident management. Image 7: The Common Operational Picture (COP) through the Incident Management Software System Information Flow Within IMSS: The information from crisis-mapping efforts is going to come in a variety of formats. These might include JPEG/PNG images, ShapeFiles, Excel documents, etc. IMSS s power lies in that it can handle each of these formats. The Layer Management tool allows the user to upload Shapefile, KML, WMS, and ArcGIS Server Service files into the platform (IMSS). These file formats are Geospatial Information System (GIS) files. In other words, these files allow different information to be displayed on a map. Anything from requests for help, to the extent of oil spills, to estimated damage, to damaged roads can be mapped using these files. Additionally, these file formats are widely held as the standard for displaying GIS information. Open-source communities and governmental agencies alike are familiar with these file formats. Theoretical Case Study : Image 8: Various types of data can be imported into the IMSS Below is a theoretical scenario for how an activation would take place Hurricane Carter, a category 4 hurricane made landfall 20 miles away from Mobile, Alabama in July 2019. Similarly to Hurricane Sandy, the storm damaged traditional landlines and cell towers, but online mobile access remained active. Thousands of people turned to social media looking for rescue or aid. In the aftermath of the storm, the Coast Guard activated a handful of Crisis-

P a g e 20 Mapping organizations, including the Digital Humanitarian Network and the Standby Volunteer Task Force. The Coast Guard specifically requested information relating to people needing rescue or aid as well as information that helped assess the extent of damage on the ground. Within 24 hours, these online communities began gathering information from Facebook and Twitter. Thousands of volunteers poured through millions of posts and tweets. Posts where there was critical information (Ex. person in immediate danger) was passed on directly to the Coast Guard and other response agencies through a Skype Conversation. Additionally, this information was gathered in a ShapeFile (by request) that was sent every 6-12 hours to the Coast Guard where it could be uploaded into the IMSS platform. The service established a similar partnership with the Red Cross s Digital Operations Center. The information was used directly by the Coast Guard to track down people trapped by flooding and in collapsed buildings, locate small HAZMAT spills, and identify damage even before firefighters and first responders could step foot on the ground. Because the Coast Guard leveraged the open-source community and the DOC, there was no cost to the service, and the IMSS software made it easy to upload the information from partners into the Common Operational Picture. Major Findings and Recommendations: The following are the major findings from this paper: The Coast Guard has a Social Media Information Gap As of 2017, the Coast Guard has no means through which it can listen to information being posted on social media sites like Facebook and Twitter. This Social Media Information Gap could impact future incident response efforts. The Red Cross has noted that more and more disaster survivors are turning to social media following disasters seeking aid and assistance. In the future, people are only going to become more reliant on social media to communicate with others. During the next Hurricane Sandy or Hurricane Katrina, people might die because the Coast Guard never saw their Twitter post about being trapped in a collapsed building. Crisis-Mapping has the potential to close the Coast Guard s Information Gap Crisis-mapping is a visual tool that allows information from social media and other sources to be mapped on a single, easy-to-read platform. It combs through geotagged social media posts looking for relevant information, maps it, and then it can be incorporated into a range of different analyses. Part of what makes crisis-mapping so valuable to the Coast Guard in particular is that the it can be used in a wide range of incidents. The service could be responding to a major oil spill, earthquake, hurricane, even a tsunami. In each of these incidents, crisis-mapping can be used to enhance the Common Operational Picture (COP). Past digital activations have used crisismapping to look for trapped earthquake victims, track the progression of wildfires, assess hurricane damage, track the spread of Ebola, among dozens of other applications. Drawing on information from DHOs and other government agencies will yield the best results There are two options of how the Coast Guard can begin crisis-mapping: It can develop its own software or it can look to outside organizations for information. The Coast Guard has no

P a g e 21 experience with crisis-mapping and in the past, it has struggled to keep pace with the changing technology landscape. Because of these reasons and several others, it would make more sense for the Coast Guard to look to Digital Humanitarian Organizations (DHOs) and other government agencies for relevant information. There are multiple advantages associated with this path. DHOs are networks of volunteers, and these volunteers are highly trained in GIS, computer science, and humanitarian response. A network of 1000 altruistic, highly trained and motivated volunteers are going to be able to gather much more meaningful information than a Junior Officer assigned to crisis-mapping as a collateral duty. Beyond this, DHOs are much more nimble than the Coast Guard and can innovate at a much faster pace. This would minimize the service s Information Gap in the future. It s critical to understand the DHOs in particular DHOs at their core are grassroots organizations. It s important for the service to understand how they operate, where their expertise lies, what their capabilities are, and what their core values are. Working with DHOs is an entirely new concept to the Coast Guard, and it s one that might make some leaders within the service uncomfortable. Information confidentiality, reliability, response turnaround, and ethical conduct are all challenges that threaten to hold back future partnerships. By understanding the work that DHOs do, Coast Guard leaders can begin to see that these are all issues that they are working constantly to address and improve. Information can be incorporated into Incident Management Software System The IMSS can handle data in a number of different formats, including ShapeFiles, KML files, and images. This is ultimately how information from outside sources can be incorporated into the Coast Guard s COP. Activating a digital response can be done by anyone Part of the power of the Digital Humanitarian community is that anyone can activate a response. A team of cadets can activate and manage a digital response for the Coast Guard from the Academy in New London, CT. While the service should seek to establish a set of official protocols for crisis-mapping, it s important to understand that anyone within the Chain of Command has the capacity to make a meaningful impact. Key Recommendations: The following recommendations seek to guide Coast Guard leaders in successfully integrating crisismapping into the service s Incident Management framework. 1. Increase Coast Guard Awareness of Crisis-Mapping Crisis-Mapping is not even in the Coast Guard s vocabulary as of April 2017. In order to begin integrating it into incident response efforts, crisis-mapping needs to be widely understood as a tool that can benefit the service. Building this awareness begins within the communities in the Coast Guard that can benefit it the most: The IMAT and major sector IMDs (Sector San Francisco and New York City, for example) are the communities that would likely be the first in the service to use crisis-mapping in some capacity. Unfortunately, the impact it can have on the Coast Guard can t be

P a g e 22 demonstrated until it s used for the first time. When crisis-mapping is shown to be valuable in practice, it will be much easier to adopt service-wide. 2. Solidify Understanding of CG1View and IMSS Data Capabilities and Limitations One of the bottlenecks with utilizing crisis-mapping as a tool is understanding how information can flow from outside organizations like DHOs or the Red Cross into Coast Guard software platforms. CG1View and the IMSS both are widely used within the service and can handle different formats of information. However, these need to be understood as a prerequiste to crisis-mapping. 3. Clarify the Importance of Listening to Social Media During Response Efforts in Coast Guard Literature The most likely individual to be tasked with organizing a crisis-mapping activation for the service is the incident s Public Information Officer (PIO). The PIO has two major resources available to them: The PIO Job Aid and the Incident Management Handbook. Ultimately, the reason crisis-mapping isn t in the Coast Guard s vocabulary is because it doesn t exist in the How-to guides for incident management. Reading through the Incident Management Handbook, the PIO is listed as having 21 Major Responsibilities, the only one relating to social media being: Monitor and utilize social media as approved by the IC/UC (Incident Handbook). Additionally, the Public Information Officer Job Aid mentions the need for a Social Media Specialist as a sub-branch of the Assistant PIO for Community Relations, however this duty is not described (PIO 44). Current Coast Guard literature is vague at best about the need to utilize social media as an information source. A PIO juggling 21 different major Image 9: The Chain of Command under the Public Information Officer as part of the Incident Command System that the Coast Guard uses. Note that there is a designated Social Media Specialist responsibilities is not going to read into the PIO Job Aid and conclude that crisis-mapping is a valuable tool to the service. By modifying the two documents discussed above to highlight the value of crisis-mapping, the service can slowly begin to transition into the mentality that PIOs are translators.