Beginning Sunday October 21, 2007, San Diego County experienced the worst fires ever reported in a large urban area in the U.S. As seven major fires and five lesser fires raged, Ron Roberts, Chair of the San Diego County Board of Supervisors, said, “We are entering day three of what appears to be one of the worst fires, probably the worst fire in San Diego County history, and easily one of the worst fires in the history of the state of California.” The after-action report2 later said the seven fires collectively caused 10 civilian deaths, 23 civilian injuries, and 89 firefighter injuries, consumed 370,000 acres, or about 13% of the county’s total land area, destroyed 1,600 homes, forced the closure of a major highway between San Diego and Los Angeles, and cost in excess of $1.5 billion. The fires also caused the largest single fire evacuation in U.S. history, involving 45 shelters that even had to care and feed more than 5,000 large animals daily. The worst of it lasted five days, Sunday to Thursday.
This article covers how Web 2.0 technologies helped volunteers in unexpected ways and how others might use the lessons learned.
Sociological research on disaster situations repeatedly recognizes the value of volunteers, demonstrating their importance providing physical and emotional assistance, as well as their need for timely information.5,6,7,13,18,19 However, local and federal disaster-management planning and policy implementation have generally ignored the role of volunteers.19 Because volunteers sometimes add to the chaos of a disaster, disaster officials may even discourage their participation.16,18 Nevertheless, recent examples that indeed overwhelmed the resources of emergency-response organizations have caused policymakers to reconsider their role in disaster response.4 As such, emergency-response decision makers must understand how volunteers coordinate their efforts and how those efforts can be leveraged.10,18
A critical capability for leveraging volunteer participation in disasters involves information and communication technology, particularly Web 2.0-enabled.1,8,14,16 However, more research is needed on how volunteers coordinate through Web 2.0 technology and the lessons learned managers might use to leverage a volunteer work force during a disaster.
Hastily Formed Networks
Our story here concerns how volunteers in the 2007 San Diego fires organized into two hastily formed networks,3 using Web 2.0 technologies to keep the public informed of the progress of the seven fires, along with home destruction, changing evacuation routes, open and closed hospitals, and the availability of dozens of locations for sheltering people and animals. The first network involved two departments at San Diego State University: the Geography Department and the Visualization Lab, or VizLab. The second involved KPBS (www.kpbs.org), the local Public Broadcasting Station, Google, and (also) the Geography Department.
This wasn’t the first time San Diego was hit by fire, having suffered most recently in 2003, when KPBS reporters, using a Thomas Brothers commercial road map, had read route closures and open evacuation shelters over the radio and displayed the map of the fire provided each day by the San Diego County Emergency Operation Center (EOC). During that fire, the SDSU Geography Department posted on its Web site geographical information system (GIS) maps crudely combining government satellite image data with Thomas Brothers commercial road maps. None of this information could be kept up-to-date due to how quickly the various fires changed direction. The information was also not very specific, since it identified only the fires in relation to major roadways rather than specific streets or neighborhoods. Consequently, the value of the information to the public was limited.
In 2007, things were different, partly because Web 2.0 technologies enabled volunteers to engage the public in ways that would have been unimaginable only four years earlier. At the same time, use of Web 2.0 tools raised multiple unexpected questions in need of quick resolution. Our intention here is to demonstrate that Web 2.0 tools can yield unexpected emergent behaviors and that the secret to success with Web 2.0 is agility in the face of surprise.
Fires Begin
That October Sunday in Southern California began like many previous days, with high winds blowing off the desert toward the Pacific Ocean. But they were the notoriously hot Santa Ana winds that dry out any moisture on the land, along with enough strength to topple power lines. The people of Southern California awoke to a red sky not only in the east as the sun came up but in the north from a wildfire that became known as the Canyon fire starting at 4:50 A.M. near Malibu in Los Angeles County. Before noon, three more fires had ignited: in Santa Barbara, Ventura, and San Diego counties. Shortly after noon, another fire ignited in San Diego County, and by the time the sun set, three more wildfires had ignited. In all, Sunday ended with seven major fires throughout Southern California, with two in San Diego County.
Most press attention focused on the Canyon fire in Malibu due to the costly homes of movie stars being consumed; but more than 100 miles south in San Diego, several people recognized a pattern similar to what they had experienced in 2003. Ming-Hsiang Tsou (Ming), an associate professor of geography and leader of the GIS Lab in the Geography Department, began calling his current and former students about the seriousness of the fires. They agreed to gather and upload satellite images, as well as create and upload GIS maps of the fires with overlays relative to landmarks (such as highways and cities) to enable the public to see where the fires were located.
Ming also contacted a colleague in Taiwan to convince the Taiwanese government to focus one of its satellites on the Southern California region to provide higher-resolution GIS images. Ming was also scheduled to attend a conference in Washington, D.C. that week, leaving with much trepidation; yet when there, he happened to meet a colleague from NASA’s Jet Propulsion Laboratory who had created wind overlays for GIS maps. Ming’s volunteer students were able to integrate the wind overlays on the satellite images to identify wind direction, helping predict not only where the fires were coming from but where they were likely to go next. Figure 1 is an example of the kind of maps posted on the Geography Department Web site (map.sdsu.edu).
Meanwhile, also on Sunday, as the fires grew and began to threaten homes, KPBS senior online managing editor, Leng Caloh, posted a Google MyMap linked from the KPBS Web site showing evacuation shelters. Caloh said, “As I drove to the [KPBS] station, thinking about what I could do to support the wildfire coverage effort, I asked myself, What do people most want to know? They want to know where the fire is! I made up my mind at that point to create a Google MyMap. I’d used the Google MyMap feature in the past for mapping my vacations, so I knew how easy it would be to put together. Earlier in the month, at KPBS, we had embedded Google MyMaps into the KPBS Web site for two projects—the Envision San Diego “Tapped Out” water special and the Mt. Soledad landslide news story—so I knew it was easy to grab the embed code for the map.”
Caloh felt a MyMap containing the fire perimeters, evacuation routes, evacuation shelters, and first-aid locations would be especially helpful, since it would allow the public to zoom in on the map to their own streets to see if their houses were affected and/or get directions to shelters. As Caloh created the KPBS Fires MyMap at www.kpbs.org, she thought it would be just another way to display evacuation shelters; instead it evolved into something much more useful. The KPBS staff used the MyMap polygon feature to draw crude fire perimeters by looking at the map of the fire perimeters emailed by the EOC as a .pdf document estimating where the perimeter was in relation to specific streets on the MyMap; Figure 2 is screenshot of the KPBS Fires MyMap.
As the fires spread, the public, not just in San Diego but around the world, grew concerned. Family members serving in Iraq and Afghanistan, as well as others in China and Japan, wanted information to calm their fear that family members might be in danger. To learn more, many turned to Google’s search engine, querying “San Diego wildfires.” The highest-ranked link the public would get in response to the query was the SDSU Geography Department GIS Web site, largely because the department had been so involved in the 2003 fires. Clicking on the link led not only to the department’s 2003 fire data but to the ever-expanding maps and satellite images Ming and his colleagues were adding for the current fires. “We got over 3,000 separate Web site referrals. Other TV stations, including CBS Channel 8, and newspapers, including the San Diego Union Tribune, all linked to our Web address,” said Ming. The Geography Department’s Web site also included a link that drove traffic to KPBS, further increasing the number of hits on the KPBS Fires MyMaps.
On Sunday, within hours of the start of the fires, the KPBS and Geography Department Web sites were receiving hundreds of thousands of hits. By Monday, more than 236,000 unique IP-address hits were recorded on the department site server. Without being asked, the VizLab’s director, Eric Frost, offered his servers to the Geography Department to manage the unexpectedly heavy traffic. A similar thing happened for the KPBS Fires MyMap on Google’s MyMap server. On Monday, the director of Google’s MyMap Division noted a spike in one of its servers, leading him to initially suspect the server was experiencing a hacker attack. As he investigated, he realized it was the result of the heavy traffic to the Fires MyMap only recently created by KPBS. Without contacting KPBS, the MyMap Division shifted the KPBS Fires MyMap to a larger server to handle the traffic so it would continue to serve the public. In addition, as Google’s MyMaps development team became more aware of and engaged in how KPBS was using MyMap, Google added the KPBS Fires MyMap as a featured link on its search pages, as well as on Google Maps. Within two days (Monday to Wednesday), 1.5 million hits were recorded on the KPBS Fires MyMap on the Google MyMap server. Even with increased server capacity, this volume was enough to slow response time for those seeking access.
On Monday morning, with the fires raging, SDSU professor of geography André Skupin made his way to campus only to find it closed due to the fires. Heading home, he stopped for coffee at a Starbucks in La Jolla and logged into the SDSU server to continue his map work. He had trouble getting to the Geography Department Web site through the SDSU server, so on the spot signed up for a T-Mobile account to access Starbucks Wi-Fi and logged into the Geography Department Web site. He also accessed regularly updated satellite image data of fire hotspots available through a U.S. Department of Agriculture Forest Service Web site. As he worked with the maps, he overheard other Starbucks customers commenting on the fires, expressing their fear and need for information, including “Google Maps…What’s that?” and “I wish I could just print out a map.” This “market research” pushed Skupin to try to make the maps printable. “The maps I created were meant to help people see where the fires had occurred. That’s why I ended up also creating maps that focused on particular fires, instead of just showing complete USGS [U.S. Geological Survey] map sheets of the whole region.”
At the same time, Cris Giovando, a recent graduate of the SDSU Geography Department, offered to help coordinate the changes Ming and his student volunteers were making to their GIS maps. Using a Web site hosted by Pbwiki.com, Giovando created his own wiki, encouraging volunteers to post notes, maps, questions, and coordination details. Pbwiki.com employees apparently noticed the increased traffic on the wiki and offered Giovando more server and network capacity, allowing volunteers to post their mapping progress. Giovando later said, “When the wiki came on, I started looking at the maps on the wiki. I would then download from the USGS the base maps the public is familiar with. I combined them with the latest fire data—detected by satellite—and other map layers the Geography Department had available. In designing the maps, my focus was on conveying the temporal progression of the fires to give people a data-driven idea of the past and current locations of fires and thus their approximate direction. The resulting maps were converted to letter-size .pdf format and uploaded to the wiki site. Due to the temporal focus of the maps, my main challenge was how to keep them updated. A key enabling element was the availability of Web-based, geo-referenced fire data published by the Forest Service, as opposed to having just graphic images. That enabled me to create maps with a traditional cartographic look yet update them regularly, all while sitting in my neighborhood coffee shop.”
Skupin’s printable .pdf maps were an invaluable tool, with the public printing them to guide their drive home or to shelters. The Geography Department Web site became the global knowledge source that included updated satellite images of the fires, as well as updated .pdf files. Additionally, pilots in private and commercial airplanes flying over the region emailed pictures of the fires they took to the Geography Department, making them available for uploading as well.
For members of the public with computer access, KPBS Fires MyMaps was another resource. As the number of fires in San Diego County increased from two to 12, the KPBS New Media division created a Twitter account (@ kpbsnews) to provide fire updates to the public while monitoring the KPBS Radio broadcast and using the Twitter messages to update evacuations, fire developments, and shelter information, particularly which ones were open and which were full. The team organized into shifts to ensure it could update the information around the clock.
The public sent KPBS email via news@kpbs.org, letters@kpbs.org, and interact@kpbs.org on, say, road closures, shelters, and evacuation routes. Some also reported errors in posted shelter locations and fire perimeters, essentially becoming reporters on the street, while others reported via personal twitter accounts (http://www.wired.com/threatlevel/2007/10/firsthand-repor/).
The fire-perimeter lines drawn on the KPBS Fires MyMap were crude because they were drawn on the un-zoomed map, so when the public zoomed in to streets and houses, the lines were no longer necessarily accurate. The public also used the zoom feature of the KPBS Fires MyMap to determine if their own streets and houses were affected, a level of detail never intended for the crude polygonline feature of Google MyMap. Caloh from KPBS was increasingly concerned about this limitation, knowing there was a well-regarded GIS group in the Geography Department, so asked Ming for help with the fire-perimeter data. Giovando from the Geography Department also offered his services. Following discussions between Ming and the EOC, the EOC agreed to release—only to Ming—the fire-perimeter shape files, including GIS coordinates it used to generate its maps for the public. Releasing the shape files allowed Giovando to translate them into actual coordinate-based renderings that could be overlaid onto the KPBS fire MyMap to create accurate fire perimeters as the public zoomed in for a closer look. For the next 36 hours (virtually without sleep), he carried out a detailed process of translating the EOC maps into accurate renderings for the KPBS Fires MyMap, restarting it each time the EOC released a new fire-perimeter shape file.
Each shelter, route, first-aid location, and road closure on the KPBS Fires MyMap was represented as a graphical icon. As the EOC announced more shelters, evacuation routes, and closures, the KPBS Fires MyMap team had to add more and more icons. But by the third day of the fires, the Fires MyMap was running out of on-screen real estate, since the Google MyMap application was not designed for hundreds of icons. Having room for all of them, organized in a way the public could search and identify, quickly became a major interface-design problem.
Serendipitously, Eric Frost of the VizLab stopped by KPBS, offering the assistance of a member of a Google Earth mapping team visiting the VizLab when the fires began. Following a few calls to Google, the Google MyMap development staff in Northern California also offered its services, participating in several phone meetings with KPBS staff to explore the icons problem.
Google sent an emergency email to its employees in Irvine soliciting volunteers to drive to San Diego to help KPBS. One was a Google hardware engineer not previously involved in MyMap development, GIS, or software development but who simply wanted to help. Later, his on-site discussions with KPBS staff and Google MyMap developers in Northern California led him to try a number of software solutions described by remote MyMap developers. This led to the realization that the problems KPBS was having with icons could be solved only with a pre-release of a newer version of the MyMap application. Within hours, Google had drafted a non-disclosure agreement and had it signed by all parties so pre-release features could be used on the KPBS Fires MyMap.
On Tuesday, fire destroyed a KPBS radio-transmission tower, knocking KPBS off the air. Until an alternative tower was identified, the only way the KPBS audience could stay informed was through the KPBS Web site, with online audio stream, @kpbsnews Twitter feed, and the KPBS Fires MyMap on the Google server.
Also on Tuesday, a KPBS staff member questioned the accuracy of the Web site’s information, saying, “What happens if a house on our Fires MyMap isn’t in a burn area, but the house owners report fire damage; should the information be pulled off the Web site?” Caloh, the New Media manager, said, “Put the information out there now; we’ll put a disclaimer if necessary. The public needs to know, even if it’s inaccurate. The public will help us correct the problem.”
Despite Caloh’s concern over inaccuracies in the fire-perimeter data, by the time Giovando’s two-hour process integrating the KPBS fire perimeter data into the icons on the KPBS Fires MyMap, the fires had moved on, making obsolete at least some of the data. However, the public kept emailing and calling KPBS regarding inaccuracies in the fire perimeters that were difficult to fix quickly. The Google on-site hardware-expert volunteer worked with Google software engineers in Northern California and other locations within the company to develop software to automate Giovando’s multi-step translation process, requiring the hardware engineer learn GIS, the MyMap application, and software development—all in two days—while Google software engineers created new software in 36 hours. The software was not installed on the KPBS Fires MyMap, as it would have required Caloh to agree to a disruption of the Web site. Moreover, by Thursday when the software solution was proposed to Caloh, the wildfires were no longer centered in urban areas, and Caloh and her staff were too exhausted to consider new software. Nevertheless, the speed the problem was defined and the software designed and written by remote software engineers was impressive.
As the wildfires moved away from urban areas on Thursday, more than 1.5 million people had viewed the KPBS Fires MyMap, as well as the Web site of the California Department of Forestry and Fire Protection (http://www.fire.ca.gov/) linked to the KPBS Web site. Google’s MyMap project manager Jessica Lee said, “There have been several other instances of MyMap being used for sharing news on natural disasters and current events but never before on this scale. It’s really amazing that they managed to do this. I know they’re a small, shoestring operation, but if you give people the tools and the technology to do this, they’ll do it.” About 1,000 people also used the KPBS Twitter feed to receive KPBS text updates.
Personal Commitment
What does all this say about Web 2.0 in volunteer emergency response? All players, except the San Diego County EOC, were individuals voluntarily engaging in a process requiring their personal time, commitment, and new ways of thinking. KPBS staff could simply have relayed to the public the briefings provided by the EOC but chose instead to create, populate, and update the Fires MyMap on their own. The SDSU Geography Department could have populated its own Web site rather than split its resources with KPBS. The public could have passively accepted or complained about the KPBS MyMap data rather than inform KPBS staff of changes and inaccuracies as shelters closed and fires shifted direction.
Many studies have documented the emergence of volunteer effort in response to emergencies. Members of the public frequently want to help, but emergency-response organizations traditionally have difficulty coordinating them,4,18 since they stretch already thin resources, given the food, water, and coordination they require and can overwhelm on-site emergency commanders. How best to incorporate volunteers into emergency response is an open research and management issue.
Lessons Learned
How volunteers responded to the 2007 San Diego wildfires provides lessons learned for incorporating volunteers into emergency response directly tied to Web 2.0 technologies. In addressing them, we recognize San Diego in 2007 is but one case, though an extreme one. However, extreme cases often help private and public agencies prepare for less-extreme situations.
An infrastructure intended to support volunteer response to emergencies does not focus on control and limited use but on facilitating broad-based serendipity.
The most important lesson involved the match between Web 2.0 and the way volunteers coordinate their activity. Volunteers cannot be preplanned5; who chooses to volunteer in what way is emergent behavior. During the 2007 San Diego wildfires, a small, local public-broadcasting station—KPBS—became the public’s voice and ears, emerging as a key player not because it intended to be. The Geography Department’s GIS Web site emerged as a key player for integrating overlays of satellite images from as far away as a Taiwanese satellite not because it intended to be. Ming did not go to his conference expecting to get wind overlays, though that is what he found. André Skupin wasn’t told to create downloadable .pdf maps or what exactly the public might be interested in seeing in them, yet his time in Starbucks gave him a firsthand view of public fear and need for information. Google’s MyMap application was not created as a resource for use by the public deciding whether to leave or remain in their homes when fire was near or inform far-flung family members as to whether they should worry about their loved ones. When KPBS launched its Fires MyMap, its staff never anticipated there would be so many fires and so many shelters it would run out of screen real estate. The Google hardware engineer with no prior GIS experience never imagined he would someday help improve Google software for GIS translations. EOC staff never thought its maps would end up as an overlay on a Google MyMap. These aspects of this particular volunteer response should inform companies and public officials alike that, while emergency-incident commanders continue to practice their emergency-response roles and use secure collaborative-knowledge systems to exchange information among vetted disaster-response professionals,20 volunteers respond to emergencies in ways never fully anticipated.
Web 2.0 applications are and were ideally suited for such emergent behavior. Self-contained, highly decentralized components encourage individuals to not passively accept information posted on public Web sites but contribute to them in ways that leverage the information. Because Web 2.0 tools make actions and actors visible, allowing quick “traces” of information to be shared,9 viewers readily share a localized context that facilitates communication. The KPBS Fires MyMap garnered attention due to its ease of use, extensibility with icons, mashup with fire-perimeter overlays, and the ability to zoom in—all characteristics of Web 2.0. The short character string of tweets was just long enough to allow quick (wireless) sharing of updates. Mobile access to the SDSU Web site allowed an expert cartographer to download and enhance, then upload, .pdf maps as he had coffee. Web 2.0 applications may simply be ideally suited for volunteerism in emergency response, beyond the kind of highly centralized information systems in traditional emergency-command centers. Needed now is further development of Web 2.0 components that are deployable in disasters in ways neither anticipated nor envisioned.
Another lesson involved what we call “planned serendipity.” Web 2.0 is ultimately about encouraging serendipity, or connections, new tools and applications, and meaning in data. Establishing a technical and organizational infrastructure that fosters planned serendipity is essential if volunteers are to help respond to disasters. Such an infrastructure involves at least four requirements:
A culture of experimentation and innovation. At KPBS, such a culture fostered experimenting with MyMap applications two weeks prior to the fires on a report about water reservoirs. It proved critical for the KPBS staff’s understanding of how to create value from the MyMap application when the fires began;
The ability to distinguish between robust Web 2.0 tools and fast-response uses. The Google MyMap application was robust, thus serving a valuable function during the fires. The fire-perimeter translation process was equally critical but created quickly and thus lacked the robustness of the MyMap application. Infrastructure must support both robust tools and less-robust but fast-response tools to serve the particular need;
The capacity for people to connect with one another, before an event, as happened between the SDSU Geography Department and the VizLab, and during an event, as happened between KPBS and Google engineers. An infrastructure for planned serendipity should be able to facilitate connections in multiple ways, ranging from low-tech (such as by encouraging people to phone one another in an emergency) to higher-tech (such as by encouraging use of social-networking tools); and
For the disaster-response industry, cultivating relationships is already expected; the new twist is cultivating them for the purpose of leveraging search engines during moments of need.
The ability to format data in ways that are reusable by unexpected others. Addressing the fact that EOC map files and the Google MyMap application used different formats required enormous manual effort, frustrating both KPBS and the public. Creating data standards that facilitate mashups is an important next step in the evolution of disaster response. An infrastructure intended to support volunteer response to emergencies does not focus on control and limited use but on facilitating broad-based serendipity.
A third lesson is that volunteerism means extra concern and complications for managers. Some might use them as a reason to prohibit volunteerism due to the inherent lack of control they can impose. Others, as seen here, use them to develop new alternatives; for example, concern over information liability could easily have forced a halt to publication of the KPBS Fires MyMap; instead, the KPBS manager identified ways to keep the Fires MyMap going by suggesting a disclaimer statement be added to the Web site. Google’s concern over allowing KPBS to use features of an unreleased new version of the Google MyMap application could also have halted the Fires MyMap; instead, Google and KPBS managers constructed a non-disclosure agreement to solve the problem. The lesson learned was that volunteers inevitably cause concern; managers need to be open to creatively identifying ways to handle them rather than use them as an excuse to stop volunteers organizing through Web 2.0 technology.
A fourth lesson involved leveraging search engines to increase public visibility, thus helping volunteers know where to put their time and resources to optimize their contribution. As volunteers put more time into publicly visible activity, the activity benefits from their collective wisdom17 and in turn has greater value, continuing a positive spiral. Both the Geography Department Web site and the KPBS MyMap site captured the attention of other volunteers in part because they were linked to yet other sites with an interest in the San Diego fires. Cross-linking to the SDSU site occurred largely because the University had been involved in the 2003 wildfires and continued reporting on post-fire recovery of burned areas, leading other organizations to include links to these sites.
Research into other disasters has documented the value of early public visibility of Web sites to public volunteer efforts.10,16 Linking serves an additional purpose beyond visibility by pointing people to relevant resources. The lesson learned is that visibility and volunteer motivation are achieved not through a single extraordinary effort but by being part of a network of cross-links within a community of like-minded Web-based organizations. For the disaster-response industry, cultivating relationships is already expected; the new twist is cultivating them for the purpose of leveraging search engines during moments of need.
A final lesson involved clarity of purpose. One reason for the success of the San Diego effort was that all the volunteers apparently individually arrived at a similar vision of what had to be done. All wanted to give the public as much information as possible, without overwhelming them, providing it in a way that would reduce public fear while allowing individuals to make their own decisions as to how to respond. All seemed to recognize what was characterized in previous research as “information dearth” during wildfires, despite the public seeking as much up-to-date information as possible.16
The cartographer responding to the 2007 San Diego wildfires wanted maps to be easy to print, and the geography professor wanted satellite images to provide detailed information about topography and wind and weather conditions, allowing the public to make decisions about its personal behavior based on its assessment of the possible future behavior of the fires. The concept behind the KPBS MyMap application was to help the public visualize the overwhelming amount of information it was receiving. This shared vision, coupled with a virtual workspace facilitating visualization and making sense of what was needed,9 provided a “conversation space” for volunteers,3 helping them coordinate without a central leader.
By observing the content of the KPBS Web site, each volunteer could appreciate what had to be done next. The cartographer could see which maps had to be made printable next. The geography professor could see the value of wind conditions and satellite images. And the public could compare what it was seeing while riding in cars with what it was seeing on the KPBS Web site. The lesson learned was that, while volunteers may not always be controlled or commanded, a Web 2.0-enabled Web site can help them visualize the necessary coordination. Provided the public shares a purpose, this may be all the coordination it needs.
Conclusion
The 2007 San Diego wildfires illustrate the value of Web 2.0 technology while suggesting it be matched with practices and attitudes that encourage serendipity and creativity, not simply following directions. The combination of Web 2.0 and these practices encourages volunteers to do what they do best—think and act creatively—improvising new means of addressing the challenges posed by a disaster. Harnessing this collective wisdom is a laudable goal for future emergency-response systems.
Acknowledgments
The authors are indebted to all who gave their time and commitment to keep the information flowing during the 2007 San Diego wildfires. Our research is supported by National Science Foundation grant SES: 0725088. See the YouTube video (http://www.youtube.com/watch?v=y7y4MlYNBTI) created by Google on how KPBS used Google MyMaps.
Join the Discussion (0)
Become a Member or Sign In to Post a Comment