The mapping of crisis information online is on the rise among nonprofessional cartographers. Map-based "web mashups" result from the application of social media or Web 2.0 technology to existing or developing data sets. Map mashups combine or "mash up" multiple sources of data, which are displayed in some geographic form. Though "participatory" forms of geotechnology--such as Google My Maps--makes maps and geographic information relatively accessible, obligations of accuracy and careful interpretation do fall to the neogeographers who pursue this new form of technical enterprise. The rise of the neogeographer in the hazards and crisis context is of particular interest, as the desire to mitigate crises through some sort of participation and assistance by members of the public is strong. The management of crisis information, and its spatial and temporal modeling, presents particular challenges which are specific to the new map-based forms of social media.
Sociology of Disaster: Models of Spatio-Temporal Behavior
Disaster researchers and practitioners often use spatial and temporal models (Dynes 1970; Powell 1954) to describe and anticipate macro social behavior. Typically, the codification and classification of time-and-space models are important methodological disaster research tools and heuristic devices, since the different disaster phases and zones represent different types of individual and group behavior (Stoddard 1968; Neal 1997). For example, Dynes (1970) describes the geography of disaster events based on a series of concentric zones. The center is an area with very severe impact, which is surrounded by a fringe area with significant damage and disruption. Aid from distant communities passes through the regional and adjacent filter zones to provide resources to the impacted areas.
The following four disaster phases are used in practice to describe macro-behavior: preparedness, response, recovery, and mitigation. Powell (1954) elaborates this depiction to include eight finer temporal stages. In our own earlier work, we relied on these macro social descriptions of spatial and temporal ordering to help frame a larger set of imminent changes arising from pervasive information and communications technology (ICT) diffusion (Palen and Liu 2007). Neal (1997), however reconsiders the disaster phases and suggests that the staged model limits how we might understand the many diverse behaviors arising in disaster and masks the variety of experiences relative to different populations and stakeholders.
The rise of what is known as Web 2.0 technology supports, inadvertently perhaps, an ability to tease apart actual behavior in disasters and pinpoint the multi-dimensionality of the experience and its effects on social life. In particular, map-based "mashups," through the use of frequently updated data from multiple sources, allow us to "see" micro-behavior spatio-temporally. As such, crisis map mashups are emerging as interesting artifacts in the practical work of reporting on, assisting in, and managing emergencies.
This observation is in line with ideas of neogeography. More than a decade ago in the GIS (Geographic Information System) community, Dangermond (1995), Monmonier (1998), and Krygier (1999) urged that next-generation GIS should be more interactive and accessible to citizens to foster public participation and collaboration in the development and management of geographic databases and any decisions made based on such data. Now in the geography discipline, the notion of "neogeography" has emerged to address a new set of geographic concerns with the rise of such enabling technologies as web mapping services and pervasive GPS-enabled devices. Turner (2006) describes neogeography as "a set of techniques and tools that fall outside the realm of traditional GIS" (p. 2). More specifically, it is about "people using and creating their own maps, on their own terms, by combining elements of an existing toolset" (Turner 2006, p. 3). Goodchild (2009) defines neogeography as "a blurring of the distinctions between producer, communicator and consumer of geographic information" who become involved in the "mapping process" (p. 82). Within neogeography, volunteered geographic information (VGI) describes the increasing "interest in using the Web to create, assemble, and disseminate geographic information provided voluntarily by individuals" (Goodchild 2007, p. 211).
The emergence of the Geospatial Web, particularly Web Mapping 2.0 (Hakley et al. 2008), has led to increases in geobrowsing activities (e.g., browsing through Google Maps or Google Earth). According to Kraak (2001a), web maps can "function as an interface or index to additional information" (p. 1) in a way that facilitates an up-to-date, dynamic, and interactive presentation and dissemination of geospatial data to many more users at a minimal cost. Web maps also allow users to explore and find answers to location-specific questions as opposed to mainstream media's broad reporting tailored to an "average" viewer (Blok 2001). Kraak (2001b) also points out that maps aid thinking and prompt decision-making. The key difference in the networked world is collaboration between people, with information flows that are changing from "a linear, publishing 'push' model...to an inter-networked, participatory model" (Hakley et al. 2008, p. 2033).
Social Media in the Crisis Context and the Rise of Map Mashups
Research on social media use in crisis situations is beginning to emerge in the crisis informatics field, a research area investigating the socio-technical concerns and the "changing information pathways" of computer-mediated communication (CMC) and information and communication technology (ICT) use in large-scale emergency response (Hagar and Haythornthwaite 2005; Palen and Liu 2007; Palen et al. 2009). For example, distributed problem solving and collective sense-making emerged through the use of social networking sites (including Facebook) after two school shootings (Vieweg et al. 2008; Palen and Vieweg 2008; Palen et al. 2009). The emergence of eyewitness photojournalism and other social practices around photographic content has emerged after several crises worldwide through online photo-sharing sites, including Flickr (Liu et al. 2008). Mark and Semaan (2008) show how ICT supports the repair of broken routines in war-torn areas. Shneiderman and Preece (2007) consider how "community response grids" can support rapid emergency response based on public participation. Qu et al. (2009) elaborated the ways in which a social networking site was used to support communication after the 2008 Sichaun earthquake. A recent study analyzed the information activities within Twitter, a micro-blogging service, after the 2009 Red River flooding in the U.S. and Canada (Starbird et al. 2010).
Data mashups are another form of social media. Increasingly, people are creating map mashups by aggregating two or more data feeds or functionalities from other web sites using application programming interfaces (APIs). While investigating the experiences of web mashup developers, Zang et al. (2008) found that map mashups are a common form of data mashups because they are "the most visual and adaptable of the mashup options" (p. 3175). Hakley et al. (2008)further point out that "different categories of neogeography mashups have emerged, depending on the type of data collected (e.g., scientific, commercial, or user-generated data).
The Harvard Humanitarian Initiative's Program on Crisis Mapping and Early Warning (1) was launched in 2007 with the vision of developing a "geo-referenced crisis monitoring platform for conflict prevention and disaster management." Meier (2009a) and his colleagues are starting to develop this cross-disciplinary "field of crisis mapping" which focuses on three research agendas: crisis map sourcing, crisis mapping analysis, and crisis mapping response. Furthermore, Nourbakhsh et al. (2006) mention how "the emergence of a new breed of volunteers--online data managers--highlights the potential of a web-based community approach to disaster operations" (p. 787), which may be a step toward seeing "the public's role shift from passive viewer to active contributor" (p. 788). This brings us back to the following question that Hakley et al. (2008) raise: "What kind of participatory practices are emerging with the support of these technologies and how do they influence the relationship between people and places?" (p. 2035).
The purpose of this paper hopes to elaborate on the first part of this question: the emerging neogeographic practices around crisis map mashup development. To that end, we qualitatively examine the origins of a set of crisis map mashups which have experienced success vis-a-vis their longevity and presumed viewership and describe the issues that the neogeographers have faced in the design and long-term support of the mashups. We then focus on two cases which illustrate how interaction between participatory and professional geotechnology platforms and expertise arises, and the lessons this might have on future expansion of cartographic skill or literacy.
We conducted a qualitative study which focused on nine crisis mashups to illustrate different, key approaches to mapping hazards and disasters. These mashups were originally selected as part of a complementary research project on interface design choices and the implications on use (Liu and Palen 2009). However, as a second stage of the research, the results of which are presented here, we collected and descriptively report on additional data which account for the conditions of their creation, objective, and design choices. These mashups, then, were selected because of the diversity of their information representations vis-a-vis hazards and crises and the diversity in those choices arising from the different circumstances from which they were created. These circumstances--the emergent neogeographic...