Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=raag21

Annals of the American Association of Geographers

ISSN: 2469-4452 (Print) 2469-4460 (Online) Journal homepage: http://www.tandfonline.com/loi/raag21

Implications of Volunteered GeographicInformation for Disaster Management andGIScience: A More Complex World of VolunteeredGeography

Billy Tusker Haworth

To cite this article: Billy Tusker Haworth (2018) Implications of Volunteered GeographicInformation for Disaster Management and GIScience: A More Complex World of VolunteeredGeography, Annals of the American Association of Geographers, 108:1, 226-240, DOI:10.1080/24694452.2017.1321979

To link to this article: https://doi.org/10.1080/24694452.2017.1321979

Published online: 09 Jun 2017.

Submit your article to this journal

Article views: 640

View Crossmark data

Citing articles: 3 View citing articles

Implications of Volunteered Geographic Informationfor Disaster Management and GIScience: A More

Complex World of Volunteered GeographyBilly Tusker Haworth

School of Agriculture and Environment, University of Western Australia

Volunteered geographic information (VGI) refers to changing practices in recent years associated withtechnological advancements that provide increasing opportunities for private citizens to produce geographicinformation. VGI activities range from public contributions to online crowdsourced mapping projects to loca-tion-related posts on social media sites. These changing practices have important implications for citizens, tra-ditional authoritative systems of geographic knowledge production, and the disciplines of geography andGIScience. One field affected by VGI is disaster management, with numerous studies reporting on the opportu-nities associated with increased citizen data and involvement in crisis response. There are also significant limi-tations to the application of VGI, however, notably related to scale, the digital divide, trust, uneven powerrelations, and adaptability of existing authoritative systems, such as formal emergency management. In thisarticle, these issues and more are critically discussed through examination of three discreet yet related studiesof VGI in community bushfire (wildfire) risk reduction in Australia. Although each study has its own uniquecontributions already published, the collective insight gained by analyzing the studies together provides newand deeper perspectives on critical issues of relevance to both disaster management policies and geography andGIScience. Importantly, the article advocates for greater emphasis on the social aspect of VGI, with citizensmapping and sharing knowledge together, rather than on individual observations and large volumes of data.Further, it raises questions of some of the much-promoted promises of VGI, particularly those that suggest thatVGI can allow “everyone” to contribute to geographic knowledge production. Key Words: digital divide, disastermanagement, disaster risk reduction, GIScience, volunteered geographic information (VGI).

自发地理信息 (VGI) 指的是与近年来为普通公民提供更多机会生产地理信息的科技进展相关的改变中

的实践。 VGI 的活动, 从追踪网路群体的製图计画之公共贡献, 到社交媒体网站上与地点有关的帖

子。这些改变中的实践, 对于公民、地理知识生产的传统权威系统, 以及地理学领域和地理信息科学具

有重要的意涵。灾害管理是受到 VGI 影响的领域之一, 其中诸多研究报导了增加的公民数据及参与之

于灾害回应的契机。但 VGI 的应用同时有着重大的限制, 尤其是关于尺度、数码鸿沟、信赖、不均权

力关系, 以及诸如官方应急管理之既有权威系统的可调适性。本文透过检视有关澳大利亚社区矮林地

大火 (野火) 风险降低的 VGI 之三个互异但相关之研究, 批判性地探讨这些议题及其他。尽管这些研

究出版各自有其独特的贡献, 但共同分析这些研究所获得的集体洞见, 则对于关乎灾害管理政策与地理

学及地理信息系统科学的关键议题, 提供了崭新且更为深刻的观点。重要的是, 本文提倡应更为强调

VGI 的社会面向, 其中公民共同绘製并分享知识, 而非强调个人化的观察与大量的信息。此外, 本文对

于 VGI 受到大力提倡的若干允诺提出质疑, 特别是 VGI 让 “人人” 皆可对地理知识生产作出贡献的主

张。 关键词： 数码鸿沟, 灾害管理, 灾害风险降低,地理信息科学, 自发地理信息 (VGI)。

La informaci�on geogr�afica voluntaria (IGV) se refiere a pr�acticas variables que en a~nos recientes han estado aso-ciadas con los avances tecnol�ogicos que brindan mayores oportunidades a las personas privadas para producirinformaci�on geogr�afica. Las actividades de la IGV incluyen desde contribuciones p�ublicas a proyectos de mapeode colaboraci�on abierta hasta puestos en sitios de los medios sociales relacionados con localizaci�on. Estaspr�acticas cambiantes tienen importantes implicaciones para la ciudadan�ıa, los sistemas tradicionales autorizadosen la producci�on de conocimiento geogr�afico y para las disciplinas de la geograf�ıa y la SIGciencia. Un campoafectado por la IGV es el manejo de desastres, sobre lo cual hay numerosos estudios que informan de las oportu-nidades asociadas con crecientes datos del ciudadano e involucramiento en respuesta a las crisis. Sin embargo,hay tambi�en limitaciones significativas para la aplicaci�on de la IGV, notablemente relacionadas con escala, ladivisoria digital, confianza, relaciones desiguales de poder y con la adaptabilidad de los sistemas autorizadosexistentes, tales como el manejo formal de las emergencias. En este art�ıculo, estos y otros asuntos son discutidoscr�ıticamente por medio del examen de tres estudios individuales de la IGV, pero relacionados entre s�ı, sobre la

Annals of the American Association of Geographers, 108(1) 2018, pp. 226–240 � 2018 by American Association of GeographersInitial submission, March 2017; revised submission, April 2017; final acceptance, April 2017

Published by Taylor & Francis, LLC.

reducci�on de riesgo comunitario por incendios forestales en Australia. Aunque cada uno de esos estudios tienesus propias contribuciones espec�ıficas ya publicadas, la perspicacia colectiva que se gana al analizarlos en con-junto provee nuevas y m�as profundas perspectivas sobre asuntos cr�ıticos de relevancia para las pol�ıticas de man-ejo de desastres y para la geograf�ıa y la SIGciencia. Muy importante es que el art�ıculo propugna por un mayor�enfasis en el aspecto social de la IGV, esto es, los ciudadanos mapeando y compartiendo conocimiento juntos,en vez de centrarse en observaciones individuales y grandes vol�umenes de datos. Surgen, adicionalmente, inter-rogantes sobre algunas de las promesas m�as promocionadas de la IGV, en particular las que sugieren que la IGVpuede permitir que “cada quien” contribuya a la producci�on de conocimiento geogr�afico. Palabras clave: divisoriadigital, manejo de desastres, reducci�on del riesgo de desastre, SIGciencia, informaci�on geogr�afica voluntaria (IGV).

The world has recently witnessed the rapidlygrowing phenomenon of user-generated con-tent (UGC) resulting from the advent of the

Internet and the Geoweb 2.0 environment (Elwood,Goodchild, and Sui 2012). Web 2.0 (O’Reilly 2005)referred to the bidirectional online environmentwhere users are able to add their own information tothe Internet as well as read from it (Goodchild2007b), and the geoweb is defined as the geographicextension of Web 2.0, encompassing geographicallyrelated Web services, locational technologies, anddata (Sieber et al. 2016). In 2007, Goodchild coined aspecial case of UGC, volunteered geographic informa-tion (VGI). VGI refers to the widespread voluntaryengagement of large numbers of private citizens in thecreation of geographic information (Goodchild2007a). Ricker, Daniel, and Hedley (2014) arguedthat whereas Goodchild’s definition explicitly refers todata collection, much of the literature since refers toVGI as also encompassing data presentation and dis-semination (e.g., Elwood, Goodchild, and Sui 2012;Bruce et al. 2014).

VGI has been facilitated by the development of par-ticular enabling technologies, including Web 2.0 andsocial media, georeferencing, geotags, Global Position-ing Systems (GPS), graphics for visualization, broad-band communication (Goodchild 2007a), and, morerecently, cloud storage, mobile locational platforms,and smartphones (Raento, Oulasvirta, and Eagle2009). VGI is characterized by an increasing variety ofpractices (Haklay 2013) ranging from contributionsmade in response to disaster events (Zook et al. 2010),to public use of geographic information systems (GIS)and global crowdsourced mapping efforts such asOpenStreetMap (OSM; Haklay and Weber 2008) andactivities that are “fun,” like locating holiday photoson social media (Goodchild 2007a). There is a needfor critical research on the societal benefits and impli-cations of these practices.

Differentiation between active and passive sensinghas been described through a typology of VGI offered

by Craglia, Ostermann, and Spinsanti (2012). Theyposited that volunteering can be explicit or implicitand so, too, the geographic location can be eitherexplicit or implied in VGI. True VGI in the strictestsense is information that is explicitly volunteered andexplicitly geographic, such as OSM (Craglia, Oster-mann, and Spinsanti 2012). Implicit VGI, such as asocial media post containing a place name, is beingincreasingly used for geospatial research, however(Senaratne et al. 2017).

VGI represents unprecedented shifts in the content,characteristics, and modes of geographic informationcreation, sharing, and use. Despite concerns over thedigital divide, data quality, and trustworthiness, VGIhas potential application to research questions in allareas of the geography discipline (Elwood, Goodchild,and Sui 2012). This emerging field has importantimplications for what GIScience, geography, andsocial practices will look like in the future and hasbeen linked to the concept of neogeography (seeGoodchild 2008; Haklay, Singleton, and Parker 2008;Sui 2008). Neogeography denotes the proliferation ofWeb-based geographic information technologies andthe precipitated phenomenon of laypeople or nonex-perts creating their own geographic content and maps(Leszczynski 2014). What is the role of the expertgeographer when those in the general public can noweasily create their own geographic data and maps?Some have argued that VGI is both exciting for theopportunities associated with a dense network of indi-vidual, intelligent observers and worrisome as theidentity of geography as a discipline is becomingincreasingly indistinct and the implied assumptionthat geography is about describing the world ratherthan understanding and explaining it (Elwood, Good-child, and Sui 2012). The notion of software encom-passing the skills of the geographer has promptedcriticism concerning trivialization of the geographicdiscipline (Leszczynski 2014). Neogeography has beencritiqued for its instrumentalist reductionism of geog-raphy, but Leszczynski (2014) highlighted the social

Implications of VGI for Disaster Management and GIScience 227

implications of the neo component of this concept.There is a need to explain why geography is more thanmapmaking and why the methods of geographers arepowerful and far from intuitive (Goodchild 2008).Whereas neogeography is about collection and compi-lation of geographic information and facts without fur-ther analyses, expert geography is an enterprise ofknowledge production, involving differentiationbetween spatial relationships of contingency and cau-sality and the devising of explanatory spatial theories(Goodchild 2009; Leszczynski 2014). Critical geogra-phers have experience and skills absent in nonexperts,allowing for filtering of the most important trendsfrom noisy data; geographers help put context andlocal meaning back into big data. Geographers, with abroad cross section of expertise spanning technicalcomponents of GIScience as well as the critical andsocial dimensions of modern human geography, areideally positioned to examine the opportunities, limi-tations, and broader impacts of VGI (Elwood, Good-child, and Sui 2012). Although Connors, Lei, andKelly (2012) argued that there are benefits of diverseparticipants, including citizens, contributing to newknowledge discovery and multimodal interaction,there remains a need for geographers to engage inresearch and critique of the social dimensions andimplications of such diverse participation. VGI repre-sents changes to the ways in which geospatial informa-tion is created, used, and consumed, with importantimplications for numerous fields and applications. Onesuch field impacted by VGI is disaster management(Haworth and Bruce 2015), which is the focus of thisarticle.

VGI and Disaster Risk Reduction

This article examines VGI in the context of disasterand emergency management, in particular disaster riskreduction (DRR). Disaster management is a particularlyuseful case study through which to consider the impactsof VGI on traditional systems more broadly. This is notonly because disasters are important social, political, andenvironmental events with significant local and globalimpacts but because key debates associated with VGIand its benefits and limitations are intensified and dra-matized in a disaster management context. Mattersrelated to VGI such as public engagement and data reli-ability are particularly pertinent to disaster management,for instance. Moreover, the top-down arrangements ofinformation control and service delivery adopted by

authoritative emergency management parallel otherinstitutional systems impacted by VGI, such as gover-nance structures or traditional map production.

Historically, disaster management has been con-ducted for, not with, the community, following a para-military, top-down model (Pearce 2003; Palen and Liu2007). Prior to the end of the twentieth century, emer-gency management was an enterprise concerned largelywith the hazard event itself (Buckle 1999). Disasterthinking has since progressed to placing increasedemphasis on humans and their vulnerabilities as causa-tive factors (Wisner et al. 2003; Manyena et al. 2011).Alongside this, the focus of disaster management alsoshifted from emergency response to recognizing theimportance of mitigation and the possibility ofincreased public participation (Pearce 2003). Pearce(2003) reported that disaster management policiescame to recognize that public participation in combi-nation with both disaster management planning andcommunity planning results in sustainable hazard miti-gation. This recognition was evident in the HyogoFramework for Action 2005–2015, a major policy doc-ument adopted by 168 countries that emphasized com-munity resilience building for effective disastermanagement (United Nations International Strategyfor Disaster Reduction [UNISDR] 2005), and its suc-cessor, the Sendai Framework for Disaster Risk Reduc-tion 2015–2030, which focuses on communityparticipation and disaster risk management as opposedto disaster management (UNISDR 2015). Disastermanagement has gradually moved beyond a purely top-down bureaucratic model to become a more collabora-tive activity and dynamic enterprise that facilitatesmultiorganizational, intergovernmental, and intersec-toral cooperation (Waugh and Streib 2006).

In Australia, these shifts have led to a philosophy ofshared responsibility, primarily driven by the VictoriaBushfires Royal Commission (Teague, McLeod, andPascoe 2010; McLennan and Handmer 2012). As aprinciple it implies increased responsibility for all con-cerned (being the state, municipal councils, the pri-vate sector, individuals, household members, and thebroader community) and a focus on community safety.As such, disaster management now places greateremphasis on community engagement and DRR.

The notion of disaster resilience has gained cur-rency in recent decades. This is partly in response toconcerns that the language of vulnerability is disem-powering for communities but also increasing govern-mental expectations that communities will takegreater responsibility in emergencies and disasters

228 Haworth

(Cretney 2014). Resilience has been interpreted inmany ways, but a geographic perspective regards resil-ience as “the capacity of hazard-affected bodies toresist loss during disaster and to regenerate and reorga-nize after disaster in a specific area in a given period”(Zhou et al. 2010, 28). The Australian National Strat-egy for Disaster Resilience (NSDR; Council of Austra-lian Governments 2011) avoids defining resilience,describing instead “common characteristics of disasterresilient communities, individuals and organizations,”which include functioning well while under stress, suc-cessful adaptation, self-reliance, and social capacity(5). Current disaster management in Australia aims toachieve community disaster resilience through themodel of prevention, preparedness, response, andrecovery (PPRR; Prosser and Peters 2011). Preventionrepresents regulatory and physical measures aimed atpreventing emergencies; preparedness denotes arrange-ments to ensure that individuals and communities areprepared for potential disaster impacts and that allresources and services that might be needed to copecan be mobilized and deployed in a timely manner;response refers to actions taken during and immediatelyafter a disaster event to ensure impacts are minimizedand that people affected are provided immediate reliefand support; and recovery is the coordinated process ofsupporting disaster-affected communities in rebuildinginfrastructure and restoring well-being (Abrahams2001).

Research so far has emphasized the role of VGI indisaster response (Haworth and Bruce 2015; Klonneret al. 2016). The presence of both researchers and vol-unteers is concentrated in response to crises, asopposed to during mitigation or preparedness activi-ties, likely related to response being more visible andprominent, especially in the media (Klonner et al.2016). Examples include the global volunteer mappingeffort that assisted the humanitarian response to the2010 Haiti earthquake (Meier 2012); the use of crowd-sourced mapping and social media for capturing com-munity information and communicating withimpacted people in response to cyclones and floods inQueensland in 2010 and 2011 (McDougall 2011; Tay-lor et al. 2012); and the role of VGI as an alternativeinformation source to authoritative disaster informa-tion during fires in Santa Barbara in 2007 to 2009(Goodchild and Glennon 2010) and Tasmania in2013 (Irons et al. 2015).

There are, of course, both great opportunities pre-sented by VGI in disaster management, as well aschallenges and implications. Studies have reported

timely information exchange and promotion of com-munity connectedness (Taylor et al. 2012); the col-lection of data in near real time without limitationsof other geospatial technologies, such as satelliteimagery being obstructed by weather (Triglav-�Cekadaand Radovan 2013); collection of complimentarygeospatial data in regions where other data are pooror absent (McDougall 2011); and the ability of indi-viduals to volunteer and participate from outside theimpacted disaster location (Whittaker, McLennan,and Handmer 2015) as some of the unique opportuni-ties presented by VGI. Examples of reported chal-lenges include issues of source credibility and dataquality (Ostermann and Spinsanti 2011; Goodchildand Li 2012), information and personal security(Shanley et al. 2013), data management, perceivedlegal concerns associated with privacy and liability(Scassa 2013), and the underrepresentation of partic-ular groups and individuals through the notion of thedigital divide (see Sui, Goodchild, and Elwood 2013;Van Dijk and Hacker 2013).

There is more to be said of the impacts of VGI indisaster management, however, beyond a simplebinary classification of opportunities and challenges.Disaster management follows a command-and-controlmodel, and these “do not easily adapt to the expandingdata-generating and seeking activities by the public”(Palen and Liu 2007, 727). VGI signals significantsocial changes related to power relations and tradi-tional systems of top-down, authoritative productionof geographic knowledge.

Drawing on the findings of three distinct yet relatedstudies on the opportunities, challenges, and implica-tions of VGI in the context of community preparednessfor potential bushfire disaster (also referred to as wildfireor forest fire internationally), this article aims to elucidatethrough critical discussion some of the broader implica-tions of VGI on (1) authoritative emergency manage-ment and (2) the disciplines of geography andGIScience. Separately, the studies are data rich as aresult of three independent survey efforts, including com-munity surveys, professional stakeholder interviews, andparticipatory mapping focus groups, and each presentsindividual findings of value. The collective insightgained through consideration of the studies together,however, provides a deeper understanding of the currentand future considerations of VGI for the professionaland academic fields of disaster management and GIS-cience. This article differs from previous work publishedin this field, including a review of VGI for disaster man-agement (Haworth and Bruce 2015), whereby

Implications of VGI for Disaster Management and GIScience 229

conclusions here are drawn from experimental andempirical case study–based research rather than a reviewof themes emerging from wide-ranging existing litera-ture, and the three studies that form the basis for analysisin this article (Haworth, Bruce, and Middleton 2015;Haworth 2016; Haworth, Whittaker, and Bruce 2016),in that the insights described here are not immediatelyapparent in the individual studies alone. It is onlythrough research experience, critical reflection, and col-lective analysis of the various unique perspectives pre-sented in each study that the wider implications of VGIare illuminated.

The structure of the article is as follows. The nextsection briefly describes each of the three studies,including a summary of their context, aims, methods,and findings. After that, I discuss recurring themesthat emerge through collective consideration of theworks, before turning attention to the broader policyand discipline implications of VGI. Finally, I detailimportant considerations associated with the studiesand offer directions for future work.

A Tale of Three Studies

The three studies were undertaken in an Australianemergency management context, focusing on the roleof VGI in community DRR in the state of Tasmania,where bushfire risk is high and established communityengagement programs (see Bushfire Ready Neighbour-hoods 2014) provided a useful setting in which toaccess community members and explore approachesusing VGI. Although VGI has relevance to all disas-ters, unlike other hazards that are less predictable,such as volcanic eruptions, flooding, or earthquakes,bushfire has a known annual period of heightened riskthat communities can prepare for and, like VGI, prep-aration is inherently centered on community partici-pation and thus fire provides a unique case studythrough which to examine the contributions of VGI.

Bushfires threaten communities each year and repre-sent one of the greatest environmental threats in Aus-tralia (Beale and Jones 2011). Although bushfires inAustralia are not new, recent pressures, including pop-ulation growth into areas of urban–bushland interface,where the greatest potential for loss of life and propertyexists (McAneney, Chen, and Pitman 2009), and theeffects of climate change increasing temperatures andsusceptibility to droughts have exacerbated risk condi-tions (Frandsen 2012). Major bushfire disasters havedevastated numerous communities in recent years (see,

e.g., Every et al. 2015; Irons et al. 2015; Rich et al.2016; Whittaker, Eriksen, and Haynes 2016).

Increasingly, communities and individuals are encour-aged to manage their own bushfire risk in line with thepolicy changes toward shared responsibility and commu-nity resilience. A key focus here is increasing communityand household preparedness. Preparing for disasters dra-matically reduces the risk of negative impacts on life andproperty. Research has shown, though, that the numberof people who actively prepare for disasters is significantlylower than the number of people at risk, and providingcommunities with relevant information alone is ineffec-tive in stimulating meaningful and active disaster prepara-tion engagement (Paton 2003). Finding new ways toengage communities is an important global issue, andcommunity participation is considered a fundamentalprinciple of DRR and resilience building (UNISDR2015). Questions related to the effectiveness ofapproaches incorporating VGI in promoting increasedcommunity engagement in DRR were central to each ofthe studies detailed next.

Study 1: Assessing the Potential Use of VGI in DRRthrough Community Surveys

The first study aimed to examine the potentialfor use of VGI in fostering community engagementin bushfire risk reduction in Tasmania based ontechnology uptake, community interest, and limita-tions to use. A survey was administered in twelveat-risk communities across Tasmania, issued bymail, in person, and online, with 154 completesurveys returned (for full methods, analysis, andfindings, see Haworth, Bruce, and Middleton 2015).The article argued based on empirical analysis thatthere is high potential for VGI to assist with indi-vidual and community disaster preparation, particu-larly through social media platforms like Facebook.Quantifying the need for increased communityengagement and trends in VGI technology uptake,usage patterns, and community preferences forinformation sharing methods in bushfire preparationcontributed to an evidence base for the use of VGIin disaster preparedness approaches. Critically, thesurvey identified important limitations to VGI userelated to spatial and demographic factors, trust ofonline information sources, and preferences formore traditional communication methods. Methodsof VGI generation and the efficacy of these techno-logical practices for promoting positive behavioral

230 Haworth

change and empowering individuals to engage inrisk reduction activities were beyond the scope ofthe survey.

Study 2: The Perspectives of EmergencyManagement Professionals

Having discussed the views of community membersin Study 1, the second study considered the perspec-tives of emergency professionals on the key opportuni-ties and challenges associated with VGI for disastermanagement through detailed semistructured inter-views with thirteen participants from a range of profes-sional roles and emergency management organizations(for full methods, analysis, and findings, see Haworth2016). Important insights were gained into how VGIpractices and traditional authoritative emergencymanagement could operate more effectively togetherin the future. Professionals identified the increasedreach of communications, local knowledge exchange,citizen empowerment, feelings of self-worth, and spa-tial awareness as significant opportunities for VGI.Demographics and the digital divide, infrastructurereliability, required resources, data quality and sourcetrustworthiness, and perceived legal and liabilityconcerns were revealed as key challenges. The articleproposed ways forward to more effective VGI utiliza-tion in disaster management, including recommenda-tions for greater valuing of citizen knowledge inauthoritative systems such as emergency management,tackling the digital divide and inequalities in knowl-edge production, improvement in data managementmechanisms, and further adoption of VGI in disasterpreparation and recovery.

Study 3: Testing VGI in Practice throughParticipatory Mapping

The aim of the third study was to evaluate whetherthe process of mapping local knowledge for bushfirepreparation as a shared practice with other communitymembers contributes to increasing an individual’sawareness and understanding of local bushfire risk, hisor her social connectedness, and engagement in DRR.This was accomplished through participatory mappingworkshops in three Tasmanian communities accompa-nied by participant questionnaires (for full methods,analysis, and findings, see Haworth, Whittaker, andBruce 2016). Participants each attended a workshoplasting four hours, which involved collaborative paper-based and digital mapping activities related to bushfire

risk and preparation in their local community. A smallsample size (n D 31) compared with larger survey meth-ods limited statistical generalizations to broader popula-tions, but the qualitative intensive workshop approachfacilitated interaction and collaboration between partic-ipants and provided information richness, with greateremphasis on the participation with others and the map-ping process over data production. A novel participatorymapping approach to creating, sharing, and consideringVGI was adopted in the workshops, which represents animportant contribution to current debates in the field ofVGI and GIScience. Together, workshop observationsand questionnaire results provided evidence that partici-patory mapping of VGI in bushfire management con-tributes to the promotion of social inclusion, capacitybuilding, and enablement of democratic participation.A major finding was that although the local knowledgeexchanged was of value to participants in their bushfirepreparedness, the social aspect of VGI appeared to bethe most engaging element of participatory mapping.Concerns around community mapping of VGI thatarose in the study included issues of data quality, pri-vacy, trust, and the underrepresentation of particularindividuals or groups in (1) the study, (2) communitybushfire management, and (3) mapping broadly, such asyouth, people experiencing socioeconomic disadvan-tage, people with dependents, and travelers.

General Discussion

By considering the findings of these three studiestogether, we can observe threads emerging throughoutthe work. Before moving to outline the implications ofthe findings in the next section, I here describe recur-ring outcomes that arise through consideration of thepapers collectively.

First, scale is important in VGI. Scale plays a role indetermining the volumes and types of data contributedand the user experience of contributors. Previous stud-ies have emphasized the large volumes of data contrib-uted from a dense global network of observers as anexciting opportunity presented by VGI. In contrast,this research shows VGI collected and shared on localscales to be effective for the aim of increasing commu-nity engagement in disaster preparation. In the partici-patory mapping workshops, information contributedabout local neighborhoods was the most personally rel-evant to participants, and the local-scale approachproduced additional outcomes beneficial to commu-nity disaster resilience, such as increased communityconnectedness. For DRR and the discipline of

Implications of VGI for Disaster Management and GIScience 231

geography, small data (as opposed to big data) col-lected as VGI about everyday lived experiences atlocal scales are valuable. Kitchin and Lauriault (2015)argued that despite the rapid growth of big data andassociated new analytics, small data and accompa-nying studies will continue to be an essential compo-nent of the geography research landscape, withincreasing importance as it combines with new andother data sets. They made comparisons to miningpractices, whereby big data are equivalent to open pitmining and small data involve mining narrow seams ofhigh-quality data with greater control over researchdesign and the ability to answer specific, targeted ques-tions through the telling of individual, nuanced, andcontextual stories (Kitchin and Lauriault 2015). Kwan(2016) emphasized that no geographic knowledge orresearch results associated with big data are obtainedsans mediation by algorithms. Algorithmic mediationproduces uncertainties in how data are generated,processed, and analyzed and potentially limits thetypes of societal questions that can be addressed, fur-ther highlighting the quality and value of traditionalor small geographic data (Kwan 2016).

VGI at broader scales with increasingly large vol-umes of diverse information becomes increasinglyproblematic to manage. Difficulty arises for big data incoping with abundance and exhaustivity (Kitchin andLauriault 2015), and management of volunteered datawas identified as a key challenge experienced by emer-gency management professionals (Study 2). Here, thebroader scale might negatively affect the usefulness ofVGI in disaster management if it cannot be effectivelymanaged. Broader scale VGI projects might also bedifficult to monitor and sustain into the future. Largenumbers of people engaged in VGI, however, havebeen shown to be useful for reducing errors, misinfor-mation, and the spread of rumors associated withonline information contributed from nonprofessionals,which were identified as important issues for VGI ineach of the studies.

VGI facilitates new scales of participation in disas-ter management, and I argue that the notion of whatis local, or “localness,” is increasingly distorted withthe advent of VGI practices and associated platforms.That geoweb technologies allow individuals to con-tribute to disaster management from outside disasterlocations and geographically bound communitiesraises questions about (1) conceptualizations of com-munity in disaster management with the growing sig-nificance of virtual and other nongeographiccommunities for community resilience (Mulligan et al.

2016) and (2) the legitimacy of individuals and theirknowledge as local. A person who no longer physicallyresides in a community but remains connected andengaged via platforms like social media could havevaluable knowledge to assist in emergency manage-ment, could maintain connections to people andplace, and might be affected emotionally or otherwiseby a disaster in that community. How does that indi-vidual’s local status differ from that of those who residein the geographically local community, and how doesthis affect community participation, disaster resilience,and broader social practices? The three studies havepresented arguments for increased utilization and valu-ing of citizen local knowledge in disaster management,which can be potentially enabled through VGI, butfurther work is needed to determine what exactly islocal in VGI and the implications of associated blurredunderstandings of localness.

Potentially influenced by scale, the second signifi-cant recurring topic is the notion of trust. In each ofthe studies, uncertainty in trust of unknown onlinedata sources was identified as a limitation to the appli-cation and usefulness of VGI. Community membersexpressed concern about the level of trust attributed todata from anybody other than official informationsources like government agencies for disaster commu-nication (Study 1). Emergency professionals reportedtrust of citizen-produced information as a concern forintegration of VGI into official systems (Study 2).Workshop participants were not confident that theycould trust people not to alter their contributed con-tent or use it in unintended ways (Study 3).

Trust in VGI is related to data quality, source credi-bility, and contributor reputation. These elements canbe difficult to discern in VGI and thus who partici-pates in VGI practices is important. Trust of thoseinvolved and the data they share is crucial for limitingor enabling the successful adoption of VGI, as evi-denced by these studies. VGI relies on reciprocatedtrust (McCall, Martinez, and Verplanke 2015), andlearning might need to be acquired from other exam-ples of online trust building, such as Wikipedia, eBay,or Couchsurfing, which each have mechanisms forassessing and assigning levels of trust to contributorswith verification often coming from other users.McCall, Martinez, and Verplanke (2015) raised thenotion of appointed versus self-selected volunteers.For disaster management, preferencing “trusted” localpeople with known expertise for VGI contributionsand community mapping participation over a systemthat allows anyone to volunteer contributions might

232 Haworth

have merit in increasing trust of VGI and its sourcesand therefore usefulness for some purposes. I argue,though, that some of the value of VGI shown in thesestudies, such as increased community connectednessand increased spread of information, would be sacri-ficed through such an approach. Assigning greaterprivilege and status to some community members asmore suitable for VGI contribution is at odds with theopportunities that VGI provides and appears to con-tribute to some of its weaknesses through the exclusionof particular groups and individuals. This leads me tothe third recurring topic.

Finally, demographics and who can and does con-tribute VGI are key considerations. Throughout thethree studies it was evident that the inability for someto participate in VGI practices is a limitation to theusefulness of VGI in bushfire preparation and disastermanagement. The studies showed that VGI representsonly a skewed picture of communities with varyingdegrees of participation of particular groups and indi-viduals on social media and in participatory mappingactivities. This inability of VGI technologies to repre-sent and reach “everybody” was a barrier to furtheragency adoption of VGI into authoritative emergencymanagement practices.

The converse of some people being left off VGImaps and excluded from any empowerment they mightprovide is those who are able to participate dispropor-tionately benefiting and the notion of elitism. Thepower dynamics that accompany VGI participationare of significance to local disaster management incommunities, through the potential amplification ofcommunity tensions and social divides, for instance,and to social systems more broadly. The contrastingabilities of VGI to empower and exclude, often simul-taneously, could exacerbate existing socioeconomicdivisions and raise increasing concerns about the roleof geotechnological advancements globally. Thisresearch examined the case of Tasmania, a relativelyhomogenous community, relative to other regions ofthe world where the differences between the rich andthe poor and those with high social capital and thosewithout might be even more pronounced. Thus, theseissues are potentially even more significant in othercontexts. Geographers, particularly those workingwith VGI and public participation, need to remainacutely aware of the digital divide and work towardreducing it, providing more people greater access tothe benefits of geospatial technologies and achievingincreasingly accurate and comprehensive pictures ofpopulations on maps.

Implications

Implications for Authoritative Disaster Management

VGI has had a remarkable impact on emergencymanagement in just a few years, and together the out-comes of these studies tell a story of opportunity, dis-ruption, and change. On one hand, the implications ofVGI for emergency management relate to positivechange and opportunities with technological advance-ments leading to increased citizen participation indisaster management. In particular, strengths of VGI inthe preparedness stage of disaster management weredemonstrated to include increasing community con-nectedness, local knowledge exchange, risk awareness,and engagement in disaster preparation. This representsa significant shift in how VGI is considered in disastermanagement. Previously, value in VGI was seen in cri-sis response for providing large volumes of informationfrom new data sources and on-the-ground intelligence,amplification of authoritative messaging through socialmedia, and mapping for humanitarian relief from loca-tions away from the disaster location, with very littleunderstanding or appreciation of the role that VGImight play in other phases of disaster management.

That VGI offers promise for achieving the goals ofdisaster management outside of immediate eventresponse, namely, increased community disaster resil-ience, has practical and policy implications. In currentand future community engagement initiatives, VGIthrough participatory mapping and social media siteslike Facebook should be increasingly used with greaterconfidence. The evidence base provided by this workdemonstrates that these technologies are already beingused by large portions of communities in diverse andsophisticated ways; communities are interested in theiruse for disaster communication; agencies recognize theopportunities they provide for improving disaster man-agement; and there are substantial benefits of commu-nity-scale mapping of VGI to community bushfirepreparedness and disaster resilience. Going forward, inpractice this might mean that social media and partici-patory mapping are used increasingly as engagementtools in current community development works, in asimilar fashion to the participatory mapping work-shops described in Study 3. In policy, this might meanformal recognition of the value of geoweb technologiesand of citizen knowledge in disaster managementthrough the proposal and implementation of new strat-egies and policies that better promote and harness thepotential of VGI.

Implications of VGI for Disaster Management and GIScience 233

On the other hand, tensions arise through the growthof VGI in disaster management. VGI and traditionaldisaster management are premised on contrasting modelsof information creation and dissemination. The first isrelatively spontaneous, unstructured, and created by lay-people, whereas the second is consistent, hierarchical,and produced by known experts. VGI practices disruptthe model of authoritative emergency management andsignify reduced control and regulation over information.Fundamentally, VGI is decentralizing, giving morepower to citizens and reducing the power of authoritativeagencies. In this context, the challenges of VGIhighlighted through this research, in particular issues ofquality assurance in heterogeneous data, source trustwor-thiness, data management, privacy, and ethical andpotential legal concerns, are especially pertinent indetermining how these sometimes-competing modes ofpractice can operate cooperatively and complimentarily.Until a policy shift or greater cultural change in emer-gency management transpires enabling agencies to fur-ther share control and to value citizens’ contributions,these challenges might be difficult to overcome and theopportunities of VGI might not be effectively realized.There might also be tensions lurking around privateownership of VGI platforms, such as social media sites,whereby official and government agencies are increas-ingly expected to utilize commercially driven resourcesto connect with citizens. Related are considerationsaround corporate gains and vulnerability of affectedcitizens in disaster management where those with com-mercial interests and vested profit stakes operate along-side government bodies.

I do not argue, however, that the disruptive changesor challenges of VGI negate the opportunities andbenefits for bushfire preparedness or disaster manage-ment. Rather, I seek to demonstrate that the implica-tions of VGI are complex and multidimensional,requiring further research and policy and organiza-tional change to best capitalize on the innovations ofrecent years. An important finding from interviewswith emergency management professionals (Study 2)was the notion of fear of the unknown and the fear ofchange and reduced control as a barrier to officialagencies’ participation in VGI practices. VGI is a phe-nomenon that is happening, not something that is pro-posed, and a choice not to engage with it would seemat odds with the goals of disaster management andshared responsibility. An implication of this researchis that agencies will need to participate more in thisspace to extract benefit from the technologies forthemselves in improving disaster management

strategies, to work toward a more holistic approach todisasters that values and works with communities, andto best deliver their service of ensuring communitysafety by engaging with people where they are, whichis often online through VGI platforms.

A central message in authoritative disaster manage-ment is that communities need to be disaster resilient.A key element of resilience is the ability to adapt tochange. In terms of VGI for increasing disaster resil-ience, based on the findings of these studies I arguethat there is a level of adaption required of emergencymanagement organizations. Agencies need to adapt tonew technologies for disaster communication, moresophisticated and detailed data from varying sources atvarying spatial scales, a lack of control over citizeninformation-sharing practices, actions and involve-ment in disaster management, and shifts in powerfrom the dominance of centralized, top-down institu-tional power to increased distribution of power amongvarious parties, including citizens. Failure of agenciesto adapt in culture and in policy might result in inef-fective emergency management strategies. Greateracceptance, encouragement, and exploitation of VGItechnologies and practices by disaster managementauthorities are key changes required. These studiesoutline important challenges to the increased utiliza-tion of VGI in authoritative emergency management,but they also demonstrate the strengths and opportuni-ties of VGI, rationale for further investigation andunderstanding, and ways forward for more effective useof VGI in disaster management. This includes acommunity-scale participatory mapping approach tousing VGI in the preparedness phase of disastermanagement.

I acknowledge, however, that VGI is not a stand-alone approach, and agencies already need to manageand triangulate information from various sources,often with insufficient resources. Further, communityengagement is just one of a suite of tasks undertakenby emergency organizations. Emergency managementhas many and varied functions, with procedures andpolicies often produced and refined through genera-tions of knowledge exchange and learning; it is unrea-sonable to expect that processes and behaviors shouldor even could have adapted to change at the rapid rateat which VGI technologies and practices haveadvanced. Acknowledging this partially explains someof the tension that exists between the differing systemsof VGI, disaster management, and authoritative sys-tems generally. VGI is both shaping and is shaped byauthoritative systems like formal emergency

234 Haworth

management. Adaption and effective uptake of VGIwithin disaster agencies might initially be isolated tosubsections of organizations and on small scales, suchas community engagement programs, as demonstratedin these studies to be beneficial for Tasmanian com-munities in the context of bushfire preparation. Thisapproach could serve as a model for research utiliza-tion and implementation of initiatives making use ofVGI going forward.

Implications for Geography and GIScience

Contributions of the research to critical geospatialknowledge theory are twofold. First, the conceptualiza-tion of VGI adopted in these studies advances under-standings and applications of the concept of VGIoriginally proposed by Goodchild (2007a) by empha-sizing VGI as more than a type of data but as a morecomplex social practice. In the VGI literature, empha-sis has been on VGI as data and on the individual as acontributor of VGI. Scholars have written about indi-viduals’ data creation activities and different ways inwhich individuals participate (Goodchild 2007a;Haklay 2013), motivations of individuals (e.g., Cole-man, Georgiadou, and Labonte 2009), the credibilityand trustworthiness of individuals as VGI sources(e.g., Flanagin and Metzger 2008), and the data accu-mulated from masses of individuals’ VGI practices.Although the individual is germane to VGI and thesetopics were considered in the papers, the studies alsoinvestigated the user experience of contributing VGIwith other community members, specifically for thecause of assisting in community DRR. It was shownthat the social aspect of contributing VGI (i.e., sharinglocal knowledge collaboratively with others) was morevaluable for individual and community bushfire prepa-ration and disaster resilience than the specific informa-tion mapped or shared. VGI should be consideredfurther as a powerful social process with implicationsfor society and for geography far beyond the opportu-nities presented by a network of individual observersor distributed data contributors. Conceptualizing VGIas a social practice can lead to important outcomesrelated to community connectedness and social cohe-sion, democratization, community participation,action, and positive behavioral change for a widerange of geographical questions.

Second, the studies challenge recent contestationsin the critical GIS literature made in relation to theclassification of VGI relative to other establishedfields, including participatory geographic information

systems (PGIS) and public participation GIS (PPGIS;see McCall, Martinez, and Verplanke 2015; Brown2017). Authors such as McCall, Martinez, and Ver-planke (2015) and Brown (2017) have claimed thatVGI is not participatory, that it lacks cultural informa-tion, that it is not empowering on an individual scale,and that it focuses more on volume of informationrather than depth. This work (in particular, Study 3)presents methods and findings in opposition to theseclaims.

In response to calls for VGI and citizen science tobecome more participatory (Haklay 2013; McCall,Martinez, and Verplanke 2015), the work of thesestudies harnessed PPGIS theory and collaborativepractices in using local spatial knowledge andencouraging information sharing through the use ofVGI methods. VGI, particularly on local scales cap-turing more cultural information and everyday livedexperiences through a participatory mappingapproach, as demonstrated in Study 3, offers promisefor empowering and engaging individuals. Thisapproach allowed for benefits of both VGI andPPGIS to be gained, such as rapid dissemination ofinformation through online platforms and conve-nient information contribution through tablets andsmartphones, as well as incorporation of participatoryvalues, depth of local and cultural information, indi-vidual empowerment, and increased social connect-edness. Whereas VGI has been presented as a form ofbig data (Burns 2015), this research highlights thevalue of VGI to geography as small data, with localand cultural knowledge shared being more critical tocommunity disaster management. The approachtaken here is not presented as gospel, and there willbe others, but it is a demonstrated case of how com-ponents of VGI and participatory mapping can becombined to deliver useful outcomes for both thepublic and the disciplines of geography andGIScience.

Table 1 presents a summary of the general implica-tions associated with VGI and those more specific toauthoritative disaster management and the disciplinesof geography and GIScience, as revealed through anal-ysis of the three studies in DRR and illuminated in thepreceding discussions.

Considerations and Future Work

Although the research presented in the three stud-ies makes substantial and timely contributions to DRR

Implications of VGI for Disaster Management and GIScience 235

and extant critical debates in the fields of geographyand GIScience, they remain in-depth studies of only afraction of those fields. There are limitations to thework in relation to scope, methodological restrictions,and practical realities, and the next paragraphs detailimportant considerations for future work.

The studies, in particular the community surveys(Study 1) and the participatory mapping workshops(Study 3), were limited by who volunteered to partici-pate. Study samples tended to overrepresent older, highlyeducated, wealthier, retired people without dependents,underrepresenting youth, people with children, andthose from marginalized or lower socioeconomic back-grounds. It has been reported elsewhere that participa-tory mapping studies rarely result in representativesamples and usually exhibit biases toward older, more for-mally educated male participants with higher incomesand underrepresent ethnic groups and minorities (Brownand Kytt€a 2014; Brown 2017). Further work is requiredto determine how to better include the underrepresentedin VGI and participatory mapping projects.

Longitudinal studies would enhance this work. Forinstance, repeat participatory mapping workshops ineach of the study communities would require furthertime and resources to extend the participating commu-nities and address the potential for participant fatigueand thus assessment of long-term community

engagement outcomes based on this research was notpossible. Related, evaluation of bushfire preparationengagement outcomes provided through VGI practicesin the studies was made via questionnaires assessingparticipants’ perceived outcomes. Follow-up studieswith participants would strengthen the work by per-mitting evaluation of outcomes based on actual behav-ior change or specific actions taken following the VGImapping exercises that would indicate increased bush-fire preparation; for example, the formulation of a writ-ten bushfire survival plan.

Further, studies of longer duration would alsoallow for more detailed appraisal of methods formaintaining and managing VGI initiatives andplatforms. The notion of VGI management andmaintenance was discussed with emergency man-agement professionals (Study 2) and communitymembers during participatory mapping workshops(Study 3), but these were both limited and predom-inantly speculative. Hence, the topic deserves fur-ther research. Future work should consider thelongevity of VGI platforms and practices, particu-larly as technology advances rapidly. The roles andresponsibilities of the public, technology providers,and governmental departments and related ethicalconcerns such as data storage, privacy, and thepotential for repurposing of data in unintended

Table 1. A summary of the key implications of volunteered geographic information

General implications Implications for disaster management Implications for GIScience

� Scale of participation influences theuse and value of VGI

� Increased citizen participation andcommunity connectedness

� VGI should be considered not just asdata but as a complex social practice(with social cohesion, democratization,community participation, and action aspotential outcomes)

� Trust limits the usefulness of VGI � VGI (including social media) should beincreasingly utilised in communityengagement for disaster risk reduction

� VGI, especially through a participatorymapping approach, can be more localand cultural

� Digital divide, exclusion, and elitismare key challenges shaping VGI

� Changes to the “local” landscape ofdisaster management through onlineparticipation

� VGI as small data can be moreimportant than big data

� Decentralization of power: reducedauthoritative control of information

� Top-down disaster management needsto adapt to change and work withcitizens using VGI

� VGI methods and value of geowebshould be formally included in disastermanagement policies

� Increased expectation to utilizecommercially driven, privately ownedresources

Note: VGI D volunteered geographic information.

236 Haworth

ways need to also be considered. Learning could betaken from citizen science (see Cooper 2016) interms of system design and maintenance for ongo-ing projects that effectively use and value citizenknowledge and large volumes of data.

The emphasis of this work has been on VGI as asocial practice, with much of the focus on how peoplecontribute and share knowledge and what benefits andchallenges these practices present for individuals andcommunities. The technological elements of VGI alsorequire further attention, however. Future work mightseek to consider data characterization and qualityassessment measures, system development to bettercollate and use various forms of big data provided byVGI, and the integration of citizen-produced data onthe geoweb in its varying forms into existing emer-gency management databases.

The research questions in these studies werebased on the premise that VGI practices can con-tribute to increasing community disaster resilience.It was beyond the scope of the work, however, todevelop detailed measures of resilience. Othershave recommended measures of resilience, oftenusing indexes, and future work might benefit fromexploiting these in assessing the contributions orimpacts of VGI. The concept of measuring resil-ience raises a number of questions, such as whethersome aspects of VGI are better than others;whether some components of resilience hold greaterweight; whether resilience is indeed a desirable out-come and, if so, how much resilience is enough;and what happens when an individual or commu-nity is deemed resilient. How would a resilience“badge of honor” affect future actions and supportprovided? These questions are both conceptual andpractical in nature, and further work is needed tounderstand them.

Finally, the studies examined the specific case ofVGI for bushfire preparation engagement in Tasmania.Although findings are applicable to broader questions,it would be advantageous to examine VGI in othercontexts. I see three key areas for further scientificinvestigation here.

First, it would be beneficial to explore the issuesraised in this article in other geographic contexts, suchas other parts of Australia, where community composi-tion and disaster preparedness might differ; othercountries with differing emergency management,governance, and political structures; or regions withdifferent socioeconomic circumstances, such as deve-loping nations.

Second, research into the opportunities, challenges,and overall usefulness of VGI could be undertaken forother phases of the PPRR disaster management cycle.Review of the literature revealed a saturation of studiesexamining VGI in disaster response, and the researchdescribed in this article considered disaster prepara-tion, but the application of VGI practices for disasterprevention and recovery should also be examined. Onthe surface, it seems that the valuable qualities of VGIin DRR, such as increased community connectednessand exchange of local knowledge, might also be ofvalue in disaster recovery, where communities are aim-ing to rebuild their lives together, often over thecourse of many years. VGI and participatory mappingcould facilitate community connections and resourcesharing and could act as an indicator of progress bynoting how community maps change throughout therecovery process. This potential warrants furtherinvestigation.

Third, similar questions to those in these studies ofbushfire could be asked for other hazards, includingfloods, drought, cyclones and other storms, and volca-nic eruptions, among others. This might also includenonenvironmental events, such as preparing for andresponding to acts of terror, human conflict, diseaseoutbreaks, or other crises. Research investigating VGIin any combination of these differing scenarios wouldfurther elucidate its strengths, weaknesses, and modesof best practice and broaden our understanding of thisrapidly growing but juvenile field.

Conclusion

The insights gained through this analysis of VGI inDRR into the wider implications of changing, grow-ing, and largely bottom-up practices of geospatialinformation creation and dissemination for traditional,authoritative, top-down systems have significancebeyond the field of disaster management. In particular,VGI can be seen as disruptive but also democratizing,where GIS is facilitating enhanced civic engagementand value is increasingly recognized in both expertand citizen information and practices for a range ofapplications, including mapmaking and the disciplinesof geography and GIScience. Much has been publishedon the opportunities presented by VGI, including indisaster management, but this article demonstrates amore complex “world of volunteered geography”(Goodchild 2007a, 211) with multidimensional impli-cations for citizens, GIScience, and authoritative

Implications of VGI for Disaster Management and GIScience 237

systems of geographic knowledge production and dis-semination. What and who VGI represents is depen-dent on factors such as the scale of participation;demographics; and existing social, cultural, and politi-cal systems that determine what is valued and wherepower lies, with often the already marginalized beingthe least visible in VGI. To what extent VGI can beuseful in a given scenario is dependent on whether sys-tems are established to best capitalize on the growingcollections of citizen-supplied data. If organizationsare not willing or able to adapt to the changes in infor-mation creation practices and increased citizeninvolvement enabled by VGI technologies, then theopportunities described could, in fact, become missedopportunities. A key area of opportunity for the futureof VGI highlighted in this article is that of VGI as asocial practice, where emphasis is placed on the poten-tial for increased social cohesion and the participatorynature of people collectively contributing geographicknowledge, rather than on individual observations orlarge volumes of data. That VGI technologies andpractices might contribute to a world of greater citizenparticipation in arenas such as disaster management orgeographic and scientific knowledge production is astirring suggestion, but the question remains, can thisbe a world for everyone?

Acknowledgments

I thank Eleanor Bruce, Kurt Iveson, Jon Corbett,Nadine Schuurman, and Deanne Bird for reading ear-lier versions of this article. I thank the editor andreviewer Mei-Po Kwan and one anonymous reviewerfor their constructive comments. I thank all of thestudy participants for their valuable contributions.

Funding

Parts of this research were supported by the Com-monwealth of Australia through the Bushfire and Nat-ural Hazards Cooperative Research Centre, theTasmania Fire Service, the International Associationof Wildland Fire, and the University of Sydney.

References

Abrahams, J. 2001. Disaster management in Australia: Thenational emergency management system. EmergencyMedicine 13:165–73.

Beale, J., and W. Jones. 2011. Preventing and reducingbushfire arson in Australia: A review of what is known.Fire Technology 47:507–18.

Brown, G. 2017. A review of sampling effects and responsebias in internet participatory mapping (PPGIS/PGIS/VGI). Transactions in GIS 21:39–56.

Brown, G., and M. Kytt€a. 2014. Key issues and research pri-orities for public participation GIS (PPGIS): A synthe-sis based on empirical research. Applied Geography46:122–36.

Bruce, E., L. Albright, S. Sheehan, and M. Blewitt. 2014.Distribution patterns of migrating humpback whales(Megaptera novaeangliae) in Jervis Bay, Australia: A spa-tial analysis using geographical citizen science data.Applied Geography 54:83–95.

Buckle, P. 1999. Redefining community and vulnerability inthe context of emergency management. The AustralianJournal of Emergency Management 13 (4): 21–26.

Burns, R. 2015. Rethinking big data in digital humanitari-anism: Practices, epistemologies, and social relations.GeoJournal 80 (4): 477–90.

Bushfire Ready Neighbourhoods. 2014. Home page.http://www.bushfirereadyneighbourhoods.tas.gov.au(last accessed 27 April 2017).

Coleman, D. J., Y. Georgiadou, and J. Labonte. 2009. Vol-unteered geographic information: The nature and moti-vation of produsers. International Journal of Spatial DataInfrastructures Research 4:332–58.

Connors, J. P., S. Lei, and M. Kelly. 2012. Citizen science inthe age of neogeography: Utilizing volunteered geo-graphic information for environmental monitoring.Annals of the Association of American Geographers 102(6): 1267–89.

Cooper, C. B. 2016. Citizen science: How ordinary people arechanging the face of discovery. New York: Overlook.

Council of Australian Governments. 2011. National strategyfor disaster resilience: Building our nation’s resilience todisasters. https://www.coag.gov.au/sites/default/files/national_strategy_disaster_resilience.pdf (last accessed 25March 2016).

Craglia, M., F. Ostermann, and L. Spinsanti. 2012. DigitalEarth from vision to practice: Making sense of citizen-generated content. International Journal of Digital Earth5 (5): 398–416.

Cretney, R. 2014. Resilience for whom? Emerging criticalgeographies of socio-ecological resilience. GeographyCompass 8 (9): 627–40.

Elwood, S., M. F. Goodchild, and D. Sui. 2012.Researching volunteered geographic information:Spatial data, geographic research, and new socialpractice. Annals of the Association of American Geog-raphers 102 (3): 571–90.

Every, D., A. Reynolds, L. Clarkson, C. Bearman, R. Mat-thews, L. Haigh, and D. Dawson. 2015. Capturing com-munity experiences in the 2015 Sampson Flat fire: Reportfor the South Australia Country Fire Service. Melbourne,Australia: Bushfire and Natural Hazards CRC. http://www.bnhcrc.com.au/publications/sampsonflat (lastaccessed 31 May 2016).

Flanagin, A., and M. Metzger. 2008. The credibilityof volunteered geographic information. GeoJournal72:137–48.

238 Haworth

Frandsen, M. 2012. Promoting community bushfire pre-paredness: Bridging the theory–practice divide. PhDdissertation, University of Tasmania, Hobart, Aus-tralia. http://eprints.utas.edu.au/15309/ (last accessed 1November 2014).

Goodchild, M. F. 2007a. Citizens as sensors: The world ofvolunteered geography. GeoJournal 69 (4): 211–21.

———. 2007b. Citizens as voluntary sensors: Spatial datainfrastructure in the world of Web 2.0. InternationalJournal of Spatial Data Infrastructures Research 2:24–32.

———. 2008. Commentary: Whither VGI? GeoJournal 72(3–4): 239–44.

———. 2009. Neogeography and the nature of geographicexpertise. Journal of Location Based Services 3 (2): 82–96.

Goodchild, M. F., and J. A. Glennon. 2010. Crowdsourcinggeographic information for disaster response: Aresearch frontier. International Journal of Digital Earth 3(3): 231–41.

Goodchild, M. F., and L. Li. 2012. Assuring the quality ofvolunteered geographic information. Spatial Statistics1:110–20.

Haklay, M. 2013. Citizen science and volunteered geo-graphic information—Overview and typology of partic-ipation. In Crowdsourcing geographic knowledge:Volunteered geographic information (VGI) in theory andpractice, ed. D. Z. Sui, S. Elwood, and M. F. Goodchild,105–22. Berlin: Springer.

Haklay, M., A. Singleton, and C. Parker. 2008. Web map-ping 2.0: The neogeography of the GeoWeb. GeographyCompass 2 (6): 2011–39.

Haklay, M., and P. Weber. 2008. Openstreetmap: User-gen-erated street maps. IEEE Pervasive Computing 7 (4): 12–18.

Haworth, B. 2016. Emergency management perspectives onvolunteered geographic information: Opportunities,challenges and change. Computers, Environment andUrban Systems 57:189–98.

Haworth, B., and E. Bruce. 2015. A review of volunteeredgeographic information for disaster management. Geog-raphy Compass 9 (5): 237–50.

Haworth, B., E. Bruce, and P. Middleton. 2015. Emergingtechnologies for risk reduction: Assessing the potentialuse of social media and VGI for increasing communityengagement. The Australian Journal of Emergency Man-agement 30 (3): 36–41.

Haworth, B., J. Whittaker, and E. Bruce. 2016. Assessingthe application and value of VGI and participatorymapping for community bushfire preparation. AppliedGeography 76:115–27.

Irons, M., D. Paton, L. Lester, J. Scott, and A. Martin. 2015.Social media, crisis communication and community-ledresponse and recovery: An Australian case study. Paperpresented at the Bushfire and Natural Hazards CRC andAFAC Wellington Conference 2014, Wellington, NZ.http://www.bnhcrc.com.au/publications/biblio/bnh-1621(last accessed 31 May 2016).

Kitchin, R., and T. P. Lauriault. 2015. Small data in the eraof big data. GeoJournal 80:463–75.

Klonner, C., S. Marx, T. Us�on, J. P. De Albuquerque, andB. H€ofle. 2016. Volunteered geographic information innatural hazard analysis: A systematic literature reviewof current approaches with a focus on preparedness and

mitigation. International Journal of Geo-Information5:103.

Kwan, M.-P. 2016. Algorithmic geographies: Big data, algo-rithmic uncertainty, and the production of geographicknowledge. Annals of the American Association of Geog-raphers 106 (2): 274–82.

Leszczynski, A. 2014. On the neo in neogeography. Annals ofthe Association of American Geographers 104 (1): 60–79.

Manyena, S. B., G. O’Brien, P. O’Keefe, and J. Rose. 2011.Disaster resilience: A bounce back or bounce forwardability? Local Environment 16 (5): 417–24.

McAneney, J., K. Chen, and A. Pitman. 2009. 100-years ofAustralian bushfire property losses: Is the risk signifi-cant and is it increasing? Journal of Environmental Man-agement 90 (8): 2819–22.

McCall, M. K., J. Martinez, and J. Verplanke. 2015. Shiftingboundaries of volunteered geographic information sys-tems and modalities: Learning from PGIS. ACME: AnInternational E-Journal for Critical Geographies 14 (3):791–826.

McDougall, K. 2011. Using voluntary geographic informa-tion to map the Queensland floods. Paper presented atthe Surveying and Spatial Sciences biennial confer-ence, Wellington, New Zealand.

McLennan, B., and J. Handmer. 2012. Reframing responsibil-ity-sharing for bushfire risk management in Australiaafter Black Saturday. Environmental Hazards 11 (1): 1–15.

Meier, P. 2012. Crisis mapping in action: How open sourcesoftware and global volunteer networks are changingthe world, one map at a time. Journal of Map and Geog-raphy Libraries 8:89–100.

Mulligan, M., W. Steele, L. Rickards, and H. F€unfgeld.2016. Keywords in planning: What do we mean by“community resilience”? International Planning Studies21 (4): 348–61.

O’Reilly, T. 2005. Web 2.0: Compact definition? O’ReillyRadar. http://s3.amazonaws.com/academia.edu.documents/38552727/OReilly_Radar_-_Web_2.0_Compact_Definition.pdf?AWSAccessKeyIdDAKIAJ56TQJRTWSMTNPEA&ExpiresD1475284132&SignatureD4L8f%2Bhc97drcsBK6BkYXX3JJYoY%3D&response-content-dispositionDinline%3B%20filename%3DWeb_2.0_Compact_Definition.pdf (last accessed 1October 2016).

Ostermann, F., and L. Spinsanti. 2011. A conceptual work-flow for automatically assessing the quality of volun-teered geographic information for crisis management.Paper presented at the 14th AGILE International Con-ference on Geographic Information Science, Utrecht,The Netherlands.

Palen, L., and S. B. Liu. 2007. Citizen communications incrisis: Anticipating a future of ICT-supported publicparticipation. In Proceedings of the SIGCHI Conferenceon Human Factors in Computing Systems, ed. M. B.Rosson and D. Gilmore, 727–36. San Jose, CA: ACM.

Paton, D. 2003. Disaster preparedness: A social-cognitiveperspective. Disaster Prevention and Management12:210–16.

Pearce, L. 2003. Disaster management and communityplanning, and public participation: How to achieve sus-tainable hazard mitigation. Natural Hazards 28 (2–3):211–28.

Implications of VGI for Disaster Management and GIScience 239

Prosser, B., and C. Peters. 2011. Directions in disaster resil-ience policy. The Australian Journal of Emergency Man-agement 25 (3): 8–11.

Raento, M., A. Oulasvirta, and N. Eagle. 2009. Smart-phones: An emerging tool for social scientists. Sociologi-cal Methods & Research 37 (3): 426–54.

Rich, J., A. Booth, A. Rowlands, and P. Redd. 2016.Bushfire support services and the need for evalua-tion: The 2013 Blue Mountains experience. TheAustralian Journal of Emergency Management 31 (1):8–12.

Ricker, B., S. Daniel, and N. Hedley. 2014. Fuzzy bound-aries: Hybridizing location-based services, volunteeredgeographic information and geovisualization literature.Geography Compass 8 (7): 490–504.

Scassa, T. 2013. Legal issues with volunteered geographicinformation. The Canadian Geographer 57 (1): 1–10.

Senaratne, H., A. Mobasheri, A. L. Ali, C. Capineri, and M.Haklay. 2017. A review of volunteered geographic infor-mation quality assessment methods. International Journalof Geographical Information Science 31 (1): 139–67.

Shanley, L. A., R. Burns, Z. Bastian, and E. S. Robson.2013. Tweeting up a storm: The promise and perils ofcrisis mapping. Photogrammetric Engineering and RemoteSensing 79 (10): 865–79.

Sieber, R. E., P. J. Robinson, P. A. Johnson, and J. M. Cor-bett. 2016. Doing public participation on the geospatialweb. Annals of the American Association of Geographers106 (5): 1030–46.

Sui, D. Z. 2008. The wikification of GIS and its conse-quences: Or Angelina Jolie’s new tattoo and thefuture of GIS. Computers, Environment and UrbanSystems 32 (1): 1–5.

Sui, D. Z., M. F. Goodchild, and S. Elwood. 2013. Volun-teered geographic information, the exaflood, and thegrowing digital divide. In Crowdsourcing geographicknowledge: Volunteered geographic information (VGI) intheory and practice, ed. D. Z. Sui, S. Elwood, and M. F.Goodchild, 1–12. Berlin: Springer.

Taylor, M., G. Wells, G. Howell, and B. Raphael. 2012.The role of social media as psychological first aid as asupport to community resilience building: A Facebookcase study from “Cyclone Yasi Update.” The AustralianJournal of Emergency Management 27 (1): 20–26.

Teague, B., R. McLeod, and S. Pascoe. 2010. 2009 Victo-rian Bushfires Royal Commission: Final report. Mel-bourne, Australia: State of Victoria.

Triglav-�Cekada, M., and D. Radovan. 2013. Using volun-teered geographic information to map the November2012 floods in Slovenia. Natural Hazards and Earth Sys-tem Sciences Discussion 1:2859–81.

United Nations International Strategy for Disaster Reduc-tion (UNISDR). 2005. Building the resilience ofnations and communities to disasters: Hyogo Frameworkfor Action 2005–2015. http://www.unisdr.org/2005/wcdr/intergover/official-doc/L-docs/Hyogo-framework-for-action-english.pdf (last accessed 2 April 2016).

———. 2015. Sendai framework for disaster risk reduction2015–2030. http://www.preventionweb.net/files/43291_sendaiframeworkfordrren.pdf (last accessed 27July 2016).

Van Dijk, J., and K. Hacker. 2003. The digital divide as acomplex and dynamic phenomenon. The InformationSociety: An International Journal 19 (4): 315–26.

Waugh, W. L., Jr., and G. Streib. 2006. Collaboration andleadership for effective emergency management. PublicAdministration Review 66:S131–40.

Whittaker, J., C. E. Eriksen, and K. Haynes. 2016. Genderedresponses to the 2009 Black Saturday bushfires in Victo-ria, Australia. Geographical Research 54 (2): 203–15.

Whittaker, J., B. McLennan, and J. Handmer. 2015. Areview of informal volunteerism in emergencies and dis-asters: Definition, opportunities and challenges. Interna-tional Journal of Disaster Risk Reduction 13:358–68.

Wisner, B., P. Blaike, T. Cannon, and I. Davis, eds. 2003.At risk: Natural hazards, people’s vulnerability and disas-ters. 2nd ed. London and New York: Routledge.

Zhou, H., J. Wang, J. Wan, and H. Jia. 2010. Resilience tonatural hazards: A geographic perspective. Natural Haz-ards 53:21–41.

Zook, M., M. Graham, T. Shelton, and S. Gorman. 2010.Volunteered geographic information and crowdsourc-ing disaster relief: A case study of the Haitian earth-quake.World Medical and Health Policy 2 (2): 7–33.

BILLY TUSKER HAWORTH is a Postdoctoral ResearchFellow in the School of Agriculture and Environment at theUniversity of Western Australia, WA, Perth 6009, Australia.E-mail: billy.haworth@uwa.edu.au. His research interestsinclude critical participatory GIS, volunteered geographicinformation, and the opportunities and challenges ofincreased citizen participation in GIS for applications suchas disaster management and livelihood security.

240 Haworth