Identifying success factors in crowdsourced geographic information use in government Identifying success factors in crowdsourced geographic information use in government © 2018 International Bank for Reconstruction and Development / International Development Association or The World Bank 1818 H Street NW Washington DC 20433 www.worldbank.org This work is a product of independent consultants working in collaboration with The World Bank. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights & Permissions The World Bank supports the free online communication and exchange of knowledge as the most effective way of ensuring that the fruits of research, economic and sector work, and development practice are made widely available, read, and built upon. It is therefore committed to open access, which, for authors, enables the widest possible dissemination of their findings and, for researchers, readers, and users, increases their ability to discover pertinent information. The material in this work is made available under a Creative Commons 3.0 By IGO License. You are encouraged to share and adapt this content for any purpose, including commercial use, as long as full attribution to this work is given. More information about this license can be obtained at: http://creativecommons.org/licenses/by/3.0/igo/ Any queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202- 522-2422; e-mail: pubrights@worldbank.org Attribution Please cite the work as follows: GFDRR. 2018. Identifying success factors in crowdsourced geographic information use in government. Washington, DC: GFDRR. License: Creative Commons Attribution CC BY 3.0. Acknowledgments This report was written by Mordechai (Muki) Haklay, UCL, Vyron Antoniou, Hellenic Army Geographic Directorate, Sofia Basiouka, Hellenic Ministry of Culture and Sport, Robert Soden, World Bank, Global Facility for Disaster Reduction & Recovery (GFDRR), Vivien Deparday, World Bank, Global Facility for Disaster Reduction & Recovery (GFDRR). Matthew Ryan, University of Southampton, and Peter Mooney, National University of Ireland, Maynooth. We would like to thank Madeleine Hatfield of Yellowback for editing. We would also want to thank Kate Chapman, Nicolas Chavent, Nama Budhathoki, Mikel Maron, Alex Barth, Celine Jacquin, Emir Hartato, Fraser Shilling, Andrew McKenna, Jim McAndrew, Abigail Baca, Yantisa Akhadi, Kate Logan, Amanda Malcolm, Sam Droege, Cristina Capineri, Antonello Romano, Liam Lysaght, Andrea Giacomelli, Alessandro Campanaro, Erin Korris, Linda See, Guillaume Touya, Ana-Maria Raimond, Elizabeth McCartney, Erica Hagen, Simon Roberts, Mark Iliffe, Svetla Borovska, Anthony Klemm, Tomas Straupis, Mark Herringer, Joaquín Huerta, Kiru Pillay, Bibiana McHugh, for their contribution to the case studies. Any errors or omissions that remain are solely the responsibility of the authors. Funding for this report was provided by the GFDRR Innovation Lab. Table of contents Part A: Policy summary 9 9 Introduction 10 Volunteered geographic information in government 11 Identifying factors of success in VGI projects 14 Summary of findings Part B: Main report 16 16 Introduction 18 Learning from the past 21 Research methodology 24 Analysis 25 1. Incentives/drivers 26 2. Scope and aims 27 3. Participants, stakeholders and relationships 28 4. Inputs 29 5. Technical and organizational aspects 31 Adopting VGI in government 35 Summary and way forward CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Appendix: Case studies 37 42 1. Mosquito Alert project in Spain 44 2. StatCan crowdsourcing 46 3. Monitoring Insects with Public Participation (MIPP) in Italy 48 4. Global Healthsites Mapping Project 50 5. North American Bird Phenology Program 52 6. Foul & Filthy Rivers, China 54 7. Farma Valley Community Map, Italy 56 8. Community river monitoring volunteer project, Scotland 58 9. Malawi flood preparedness 60 10. Food insecurity mapping, Bangladesh 62 11. Syria Tracker Crisis Map 64 12. iCitizen, South Africa 66 13. Community Mapping for Exposure in Indonesia 69 14. Haiti disaster response 72 15. Mapping of South Sudan 74 16. Crowdsourcing The National Map, National Map Corps, US 77 17. Crowdsourced flood resilience in Jakarta, Indonesia 80 18. Twitter use in Italian municipalities 83 19. National Biodiversity Data Centre Ireland 85 20. Towns Conquer gamification and Instituto Geográfico Nacional toponyms database, Spain 87 21. US National Park Service – Places Project 89 22. USGS’s “Did you feel it?” 92 23. US Census Bureau – Building an OSM community of practice 94 24. New York City open data initiative 96 25. Imagery to the Crowd, State Department Humanitarian Information Unit, US 99 26. Humanitarian OpenStreetMap Team mapping in Ulaanbaatar, Mongolia 101 27. Mapping schools and health facilities in Kathmandu Valley, Nepal 103 28. Informal settlement mapping, Map Kibera, Nairobi, Kenya 105 29. Skandobs, Scandinavian predator tracking system, Norway and Sweden 107 30. Corine Land Cover 2006 (CLC2006) in OpenStreetMap, France 109 31. FixMyStreet for municipality maintenance information, UK 111 32. FINTAN vernacular placenames project, Ordnance Survey and Maritime and Coastguard Agency, UK 113 33. Boston StreetBump, US 115 34. California Roadkill Observation System (CROS), US 117 35. Crowdsourcing satellite imagery in Somalia 119 36. Portland TriMet, transportation planner, Oregon, US 121 37. The Base Adresse National (BAN) Project 123 38. Citizen participation in urban planning, Kirtipur, Nepal 8 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 125 39. Land tenure in Tanzania 127 40. Open for Business, UK 129 41. International Hydrographic Organization’s Crowdsourced Bathymetry Database (CBD) 131 42. Disaster Management, Early Warning and Decision Support Capacity Enhancement Project in Indonesia 134 43. Summer Camp Guide 136 44. Ramani Huria 138 45. Natural Resources Canada-OpenStreetMap Synergy 140 46. Participatory mapping and decision support tools for disaster risk reduction, the Philippines 142 47. Government open data usage in Lithuania 144 48. Xalapa collaborative transport mapping 146 49. FloodTags 148 50. Open Cities, Sri Lanka References and sources 151 Part A: Policy summary This policy brief summarizes the findings of the research report “Identifying success factors in crowdsourced geographic information use in government” produced by the World Bank Global Facility for Disaster Reduction and Recovery (GFDRR) in partnership with scholars from University College London (UCL). This brief explains the report’s context, methodology, main findings and recommendations. Introduction The adoption of crowdsourced geographic data, no longer be a reason for the lack of government or volunteered geographic information (VGI), adoption of VGI. Nonetheless, significant challenges as a valuable source of spatial data is growing at remain for governments seeking to take full all levels of government. VGI is crowdsourced advantage of the benefits that crowdsourcing offer. geographic information provided by a wide range of participants with varying levels of education, This research used a case study approach to knowledge and skills.1 Despite some initial concerns understand factors that have contributed to the about data quality during early development of VGI success of government VGI efforts, some of which approaches, extensive research now demonstrates include supportive organizational or legal contexts, that the reliability and accuracy of VGI is suitable for the presence of local champions, and project design official or government use.2 Such concerns should elements. The 50 case studies of government PART A: POLICY SUMMARY 9 10 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT involvement with VGI spanned multiple sectors, including disaster risk management, urban planning, Box 1: USGS activities and environmental conservation. This report argues that, while there are no “one size fits all” approaches The US Geological Survey (USGS) is an example of an to successful project implementation, a number of organization with a set of well-established crowdsourcing common lessons can be gleaned from analysis of projects. These include base mapping with The National prior examples. Map Corps, collecting data on bird migration with the North American Bird Phenology Program, and the “Did you feel it?” Volunteered geographic project that documents public experience of earthquakes. information in government National Map Corps project volunteers are asked to collect Government organizations, from the local to and edit data about human-made structures and provide the national level, rely on geographic data for accurate and authoritative spatial map data for the USGS many of their operations, from planning new National Geospatial Program’s web-based ‘The National Map’. infrastructure development to maintaining order Volunteers edit 10 different structure types in all 50 states, and responding to emergencies. For this reason, including schools, hospitals, post offices, police stations and geographic information systems (GIS) are now other important public buildings. commonplace in governmental operations. VGI, which has gained recognition as a novel source of The North American Phenology Project ran from 1880 until data in the past decade, offers many advantages 1970 when it was discontinued due to lack of financial to governments who use GIS: VGI complements resources. In a new project, USGS is working with volunteers professional data collection by government to transcribe its records, curate the data, and make it agencies by providing data coverage for locations publicly available. and time periods not addressed by official data “Did you feel it?” asks volunteers to report how they collection initiatives (Box 1). experienced an earthquake. Contributors complete an online • VGI enables government agencies to capture form to describe the impact, and their experience is quantified and integrate local knowledge, which is not using the Modified Mercalli Intensity (MMI) scale. Contributors incorporated by official processes. do not need to have experience in seismology. The project makes up for the paucity of instrumental ground-motion • Engaging the public in governmental data in regions of low seismicity by providing a rich data pool processes, creating space for dialogue, and of observations, which are analyzed and used in maps and supporting efforts around transparency and graphics available to the public. data-driven decision making. • In some cases, VGI can be a cost- and time-effective alternative to standard data collection. • VGI approaches can support the development of beneficial community skills and capacities in areas such as preparedness for natural hazards (Box 3). PART A Identifying factors of success in sensitive situations such as disaster response; lack of resources for data collection; policy change VGI projects around governmental data, particularly related to The analysis behind this report was based on 50 open data initiatives; research and development; case studies. Each case features: and environmental monitoring through citizen science where members of the public work • a government agency at local, regional, or together with scientists to collect and analyze national level environmental information. • public outreach with a request for individuals to share their knowledge (e.g. vernacular placenames) or to create new data (e.g. map a place) • use of information and communication technologies Box 2: Methodology • a specific intended use for the The following steps were taken to understand the scope of VGI information created. projects in government: The cases were coded according to a set list of characteristics (e.g. type of partnership, existence An online survey about existing VGI projects in government of an organizational champion). The analysis was made available on a dedicated website, in consultation highlights six key elements of government VGI with GFDRR. The website already included 29 case studies projects: from a 2014 report. Experts were asked to complete the survey or provide information about projects directly to the • incentives/drivers to start a project research team. • scope and aims of the project The report team identified and researched other new case • participants, stakeholders and partnerships studies. • inputs such as technical and financial resources or training 72 new projects were identified and 21 selected for inclusion in the report (these were the most complete in terms of detail • technical and organizational aspects available), along with the 29 projects identified in 2014. In • challenges encountered total, 50 cases were analyzed in depth and are detailed in the Appendix. The first five topics were analyzed further using Qualitative Comparative Analysis (QCA) (Box The case studies were compared and analyzed using 2). We discuss the main findings in each of the Qualitative Comparative Analysis (QCA). QCA is a well- elements below. accepted analytical technique that uses Boolean algebra to compare and qualitatively study social phenomena. All cases Incentives to start a VGI project

were mapped against a detailed set of factors that could (positively or negatively) affect their outcome, and several The organizations studied in this report were QCA models were tested. found to start VGI projects for several reasons, including the lack of institutional data in time PART A: POLICY SUMMARY 11 12 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Analysis of these factors showed that, compared provided by a public sector organization; policy to other incentives, environmental monitoring development or reporting on implementation; through citizen science provided an especially and disaster risk management and response. strong foundation for successful VGI projects The analysis showed that base mapping3 was an in government. In other contexts, open data important component of almost all successful policies provided opportunities for successful cases. It was successfully combined with other collaboration between government and the aims such as disaster preparedness and response public around the maintenance or updating of or public service improvement. Collection of authoritative datasets. Importantly, VGI projects base map data for the area of interest is thus that were driven by a lack of resources to create considered to be an important part of successful information for government operations, were only VGI projects. successful when there was also a commitment by the government to invest in internal capacity Participants, stakeholders and development or other sustainability measures.

relationships Without such an investment, they tended to be

unsuccessful over the long term. Most successful VGI projects involve cooperation amongst a range of partners and stakeholders. Scope and aims The analysis examined the range of partners in relation to the success of the VGI projects. The scope and aims of the case study projects fall Specifically, it examined combinations of public into the following key categories: base mapping sector, private sector, international NGOs, local of an area of interest or entire country; updating NGOs and research organizations, and considered authoritative datasets, such as the US National the overall number of partners involved in the Map (Box 1); upgrading the quality of the services project. PART A Many successful projects included international NGOs, which highlights the effectiveness of NGOs with crowdsourcing experience working together with public bodies to achieve Box 3: Open Cities Kathmandu project goals. The resources of private sector The Open Cities Kathmandu project was launched in 2013 organizations are valuable, but in our case studies, to collect data on earthquake vulnerability using the they were more successful when partnered with OpenStreetMap platform. Nepal has long been recognized as intermediaries, such as experienced NGOs. one of the most at-risk countries in the world for earthquakes and other natural hazards. The Open Cities project created The absence of local NGOs in our case studies a partnership between local NGOs, the Nepal government, might be explained by the observation that, and several universities to perform detailed base mapping of apart from the area of environmental monitoring Kathmandu and collect structural data for over 2000 schools through citizen science, there are few well- and 350 health facilities for use in seismic risk assessment. established local NGOs with the expertise to act At the conclusion of the project, local participants formed as VGI project champions. It is likely that, where a new Nepali non-for-profit technology organization called they exist, they could play a similar intermediary Kathmandu Living Labs (KLL) which has continued working on role, as occurred in the aftermath of the 2015 projects related to open mapping and VGI. In the aftermath Nepal Earthquake (Box 3). of the April 2015 earthquake in Nepal, both KLL and the OpenStreetMap platform played pivotal roles in supporting Inputs emergency response. Governments can invest in VGI projects in a wide variety of ways. Project inputs included the release of existing data resources, such as through Technical and organizational aspects

open data efforts; direct investment, for example, The research also examined various technical and employing contributors or a government employee tasked with managing the project; investment in a organizational aspects of VGI implementation new technology, such as a website or mobile app to within government programs. These aspects facilitate contributions; development and delivery included the combination of VGI data with of training programs with key partners; and creating authoritative government data, concerns about research and citizen science initiatives linked to accuracy and quality of VGI data, the intended government agencies. use of VGI data, and the role of champions in the government entity. The analysis showed that incentivizing contributions of either dedicated government The most important finding here was the need staff time, or public participants, was employed by to formalize and standardize VGI before its use many successful projects when combined with the in government systems. The analysis showed introduction of a new technology, such as a mobile that many successful projects combined the data collection application. Training activities were presence of an active government champion and also important determinants of project success, mapping activities that were focused on creating but in some cases these could be replaced by new data (as opposed to maintaining existing, partnerships with expert communities. authoritative, data). PART A: POLICY SUMMARY 13 14 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Other challenges Second, the quality of the VGI and its application are key considerations to be addressed at the Some additional factors are also important in beginning of the project, and revisited regularly ensuring successful VGI projects in government. throughout its lifecycle. Government agencies First, there should be careful consideration of are responsible for providing authoritative data whether the activity is designed as a one-off event while also integrating public input, making or ongoing initiative. While many VGI activities accuracy and reliability areas of concern. There are conceived as single events, the longevity are multiple methodologies for quality assurance of updates and maintenance remains relevant. in VGI, which should be explored and integrated Regardless of the length of the project, public depending on the context and aims. interest and participation needs to be maintained through the life of a VGI project. Maintaining Summary of findings the data and relevant software is also a concern; therefore, resources should be allocated The analysis identified key factors at several across the whole life of a project to support levels in successful established VGI projects in these activities. government. PART A At the individual level , champions and change house process are methodologically unsuitable for leaders in public sector organizations are critical. the VGI context. Overall, government staff must Individuals who oppose public participation be open to adjusting their beliefs about “how in data collection and analysis for reasons of things are done”. information security or quality can block or Other Recommendations hinder the integration of VGI into government.

In addition to the findings summarized above, Discussions to understand and alleviate the the findings of the study also suggest that concerns of such individuals can assist in the projects seeking to help government entities take adoption of VGI. advantage of the potential that VGI offers should: At the organizational level , organizations that • Identify and support an already rely on external sources for spatial data organizational champion. show a greater potential for VGI adoption because staff of government agencies that produce data • Consider a project focused on improving may see the use of VGI as a threat to established public services or environmental monitoring modes of operation and sources of funding. as these have been particularly successful in adopting VGI approaches. Further organizational issues can be procedural, such as existing legislation and service delivery • Aim to collect base map data or basic obligations, or structural, such as responsibilities information, instead of attempting to update for data collection and use. This is linked to the authoritative datasets. need for having well-understood business models technology requires significant investment and consideration of how these will change with beyond building the technology itself, and the introduction of VGI. using established technology with a training program can be effective. Technical challenges should be also noted. For example, the ability to merge datasets that • Engage with a local or international NGO that have been changed by the public into existing has experience facilitating VGI projects, or government systems was found to be critical. This link with other government agencies that have analysis suggests that governmental organizations experience in this area. new to VGI should use more mature technologies, • Plan to address the organizational, technical or partner with experienced organizations on and worldview challenges within the capacity building activities. government entity in order to gain support for Finally, conceptual or “worldview” issues VGI across the organization. need to be recognized, as they also interact with the above categories. The adoption of VGI 1 The term VGI was established by Michael Goodchild in 2007, in his requires accepting a higher level of uncertainty, paper, “Citizens as sensors: the world of volunteered geography”. The terms crowdsourced geographic information and VGI are often attention to heterogeneity, collaboration among used interchangeably. diverse groups of participants, different ways of 2 For example, see Haklay, M. (2010). How Good is Volunteered communicating with the public, and different Geographic Information? operating procedures. For example, quality 3 The creation of the basic elements of mapping – streets and roads, public buildings and facilities, major landmarks and natural features assurance procedures that are suitable for in- PART A: POLICY SUMMARY 15 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Part B: Main report Introduction The role of volunteered geographic provide information that can be used to support information (VGI) as a valued and useful its wider adoption of VGI. To this end, the report source of information is growing at all levels of compiles and summarises lessons learned and government. Put simply, VGI is “crowdsourced” successful models from government projects geographic information provided by a wide range in different sectors and at different levels. The of participants with varying levels of education, research presented in this report was motivated knowledge and skills.1 While extensive research by the following interrelated issues: demonstrates the reliability and accuracy of VGI • Sources of VGI data such as OpenStreetMap compared to official or government produced (OSM) are increasingly important across a datasets, progression towards their adoption and range of thematic areas and user communities. wider use has yet to achieve its full potential. What this research does provide is a range of • The quality and consistency of VGI data have mechanisms for ensuring that crowdsourced been assessed by a range of studies and found information is fit for purpose. Therefore, suited to many tasks. Therefore, concerns concerns about data quality are no longer a about these issues should not prevent the use reason for the lack of adoption of VGI. However, of VGI as a valuable source of data. organizational practices, regulations and legal • Managing VGI projects and interacting with issues are more difficult challenges. VGI communities is a different and potentially The aim of this updated report2 is to review more complex relationship than governments governmental projects that incorporate VGI and have had with traditional sellers and resellers 16 PART B of geographical information systems (GIS) data. for disaster preparedness) and the release of government information to the public • Governments have begun engaging with VGI for improvement and its subsequent use communities in different ways and there is by government. much to learn from these experiences. In addition, there is one flow of information that This updated report follows the notion of is not covered in this report but is important in information flows established in 2014 where it the context of information flows in general in was observed that VGI projects rely on different order to understand the full picture: kinds of information flows, summarized below: • Government → Public. The flow of • Public → Government. VGI provided by the information from government to the public public to government authorities also has a is important, but is rarely implemented in long history pre-dating the web, e.g. calling VGI projects. This report does not focus on to report a problem at a location. This report the use of open data (data made available includes several examples of such cooperation by government agencies without charge or to illuminate specific aspects of VGI practice. restrictions to the public) as there are many • Government → Public → Government and examples of this in commercial and civic Public → Government → Public. Examples of society. This is also covered in the OpenDRI two-way cooperation between government Field Guide. and the public or civic organizations form the This report explores different aspects of core of this report. Again, collaboration with government use of VGI, such as the maintenance the public has a long history. This report refers of public space (streets, public buildings and particularly to the use of publicly contributed parks), or collecting data about education, health, information to make decisions and actions tourism and civic safety. It includes a set of case (e.g. civil protection agencies using map data Box 4: OpenDRI Hosted by the World Bank’s Global Facility for Disaster has worked to implement these ideas in over 50 countries Reduction and Recovery, the Open Data for Resilience around the world. VGI and crowdsourcing are key elements of Initiative (OpenDRI) works to bring the philosophies and the OpenDRI approach. These approaches help our projects practices of the open data movement to bear on the collect up-to-date and accurate information, build new challenges of building resilience to natural hazards and partnerships across sectors, and incorporate the perspectives impacts of climate change. Partnering with governments, of the public into studies of disaster and climate impacts and international organizations and civil society groups, this risk. The research in this report was funded by the OpenDRI program develops open systems for creating, sharing, and project as a means to support further understanding of the using disaster risk and climate change information to ensure opportunities and challenges that emerging technologies that a wide range of actors can participate in the challenge and practices such as VGI create for governments and their of building resilience. Since it was launched in 2011, OpenDRI partners in building resilience. PART B: MAIN REPORT 17 18 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT studies, which are included in the Appendix. The turns to the methodology of the research and following sections provide background on the use provides an overview of the case studies. This of geographical information by government, as is followed by an analysis of the findings and this has a long history that should be considered recommendations for improving the use of VGI in alongside contemporary trends. The report then government. Learning from the past The history of digital geographical information By the 1980s, with the introduction of is intertwined with government activities. The customisable and off-the-shelf software system commonly recognized as the first GIS was packages, GIS implementation became more created by the Canadian government in the 1960s common in organizations. This led to research to map land use and agricultural productivity into the processes that assist organizations in and suitability. Manual predecessors were in use implementing GIS and how to ensure that digital perhaps 100 years earlier. Moreover, the use of geographical information is used in an effective geographical information by government agencies way. Of particular importance and relevance at all levels – from the local to the regional, to this report are the work of Stan Aronoff and national and intergovernmental – continues to be his 1989 book, Geographic Information Systems: a one of the most significant applications of GIS. Management Perspective, and William Huxhold’s, PART B An introduction to Urban GIS from 1991. Both paid community in a similar way to VGI. Much of attention to the management of GIS projects, the work undertaken in VGI is closely related the importance of understanding the way to citizen science, which can be described as organizations work and the effort required for the scientific work that comes from the public successful GIS implementation. either in collaboration or under the direction of professional scientists (Silvertown 2009). The In the past 15 years, a revolution has taken place term citizen science entered the Oxford English in GIS that has led to dramatic change in the Dictionary only recently in June 2014 and is use and manipulation of spatial information. directly linked to participatory research and more The changes were driven by increased access specifically to participatory mapping, where “non- to geographic information on the internet, the professional scientists voluntarily participate increased use of connected mobile devices such in data manipulation for scientific projects”, as as smartphones, and new techniques for data Cohn (2008) and Silvertown (2009) referred to visualization and sharing. These have led to a it. Both VGI and citizen science are extremely major change in the way geographic information dynamic and rapidly evolving fields, while their is produced. Sui (2008) and others refer to this as intertwining and close relationship, referred to digital spatial data, which is collected and edited as geographic citizen science (Haklay 2013), is not by traditional data producers but by citizens, gaining ground on many fronts. who are not experts but are willing to disseminate their spatial knowledge and observations without While VGI projects may seem fundamentally any special invitation. different from governmental approaches, with their higher level of public engagement Spatial data capture, storage, management and informal participation mechanisms, there and dissemination are particularly relevant to are a number of parallels. This is because contemporary VGI projects. In the early days, the governmental systems to which VGI is a major part of any given GIS project was the integrated are often “enterprise systems” set conversion of paper maps to digital formats. The up along the same lines. Therefore, many early challenges of integrating varied data sources lessons from when GIS was first introduced into a coherent database reflect many of the to governmental organizations are similar to issues emerging from crowdsourced information. the findings discussed below. For example, as Since this early work, the implementation of with early GIS implementation, VGI use relies GIS projects has received ongoing attention, on specific individuals who act as “champions” with Roger Tomlinson’s 2013, Thinking about inside the organization and spearhead the effort GIS: Geographic information system planning for necessary to secure acceptance for this source of managers, and significant portions of leading GIS information. Another example is the opportunity textbooks such as Longley et al.’s 2015, Geographic that major events, such as disaster response, Information Systems and Science, providing the create in terms of rethinking current procedures latest summaries from nearly five decades and practices. Evaluations of organizational of practice. responses provide opportunities to reflect on the Discussion has also arisen around citizen way current systems are utilized and develop new science projects, which are not new but have procedures for data collection and use. Major recently attracted the interest of the research events and the response to them also facilitate PART B: MAIN REPORT 19 20 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT these projects are relevant when working with individuals who are experienced in implementing Box 5: Brabham (2013) best GIS in government and encouraging them to adopt VGI as a usable source of information. practices for crowdsourcing in In other words, while VGI has specific government characteristics of its own, it should not be seen as an unprecedented approach to collecting data. 1. Clearly define the problem and solution parameters. To conclude, there is significant interest in 2. Determine the level of commitment to the outcomes, how VGI can be used in government not only commit to communicate to the online community as an alternative source of geospatial data but exactly how much impact user-submitted ideas and also to record and reveal underlying conditions labor will have on the organization. on the ground. This can provide the necessary 3. Know the online community and their motivations. It information to address societal problems and is important to know whether a given crowdsourcing shape governmental decisions and policies. This application will appeal to participants. interest stems from the fact that VGI provides a bottom-up and tangible way of understanding 4. Invest in usable, stimulating, well-designed tools. citizens’ perspectives on spatial phenomena. 5. Craft policies that consider the legal needs of the organization and the online community. In 2013, Brabham produced a report, “Using Crowdsourcing in Government”, which outlines 6. Launch a promotional plan and a plan to grow and a more general overview of the potential sustain the community. for crowdsourcing in government. Brabham 7. Be honest, transparent and responsive. also attempts to classify crowdsourcing and understand when and how to deploy 8. Be involved, but share control. crowdsourcing in government. His analysis 9. Acknowledge users and follow through includes a small number of case studies and the on obligations. report concludes with ten best practices and 10. Assess the project from many angles. considerations for crowdsourcing. This summary is a helpful starting point on which this report builds. governmental policy change and thus provide the opportunity to increase recognition of the value of VGI, just as for GIS. Citizen science projects are focused on the participation of citizens in scientific observations, methods and analysis. The philosophy of these projects is thus strongly connected to public work and they have been linked to governmental initiatives in the past decade. As a consequence, many of the experiences and lessons from PART B Research methodology The starting point for this revised report was the harmonize the case studies to make them publicly lessons learned from the 2014 report. Interest in available through the blog. The 2014 case studies the report was significant: it has been downloaded were updated with their current status and any almost 1,800 times from 41 countries around new details. Box 6 provides a brief overview of the the world in approximately three years, which case studies. suggests it has been useful to researchers and The case study collation prepared the ground policymakers. In preparation for the 2017 update, for the Qualitative Comparative Analysis (QCA). the research team sought comments and reviews QCA is an established technique that uses from experts in the field regarding possible Boolean algebra to compare social phenomena. improvements. This feedback highlighted that the QCA therefore formalizes qualitative analysis. seven key factors shaping VGI use in government, QCA can reveal patterns of causal factors that highlighted by the first report (namely: incentives, lead to certain phenomena while respecting aims, stakeholders, engagement, technical the diversity and heterogeneity of each case. aspects, success factors, and problems) have GFDRR experts, the research team and a QCA developed, resulting in a new reality. The new expert met to implement the analysis. During a report focuses on changes to these factors and two-day workshop, they examined which factors on better documenting other factors, as well as and variables are likely to impact VGI use in evaluating them more rigorously. The public face government and searched for patterns. First, of the study was a dedicated website,3 which all cases were mapped against a detailed set of hosted the data as it was gathered. factors that could affect (positively or negatively) The methodology comprised of the following the outcome of each case. Next, several scenarios steps. First, in consultation with the OpenDRI were tested to adjust the QCA models and more team and according to expert recommendations, accurately highlight the causal factors. The the online survey from the first report was results were analyzed and discussed both during adapted to ask for more focused information the workshop and through online meetings. The (e.g. funding schemes and duration of projects). findings constitute the backbone of this new Experts were invited to provide input about report. Furthermore, a policy brief was created to projects either by completing the survey or allow policy makers to quickly focus on the most contacting the research team directly. In parallel, important elements of the study. the research team gathered information about new case studies that fall within the scope of the report, with a total of 72 new cases recorded. From the online survey and research process 21 cases were selected for inclusion (these were the most complete in terms of detail available) alongside the 29 cases identified in 2014. In total, 50 cases were analyzed. Next, the research team enriched the case studies through interaction with respondents to clarify information, fill in gaps and PART B: MAIN REPORT 21 22 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Box 6: Case studies Below are short descriptions of the case studies 14. Haiti disaster response. One of the most well-known crowdsourcing that provide the basis for this report (for full projects, which was launched following the Haitian earthquake in 2010. It aimed to fill the gaps left by the collapse of the national mapping details, see the Appendix): agency. 1. Mosquito Alert project in Spain. A citizen science project that aims to 15. Mapping of South Sudan. Addresses the need for an up-to-date map raise awareness of the tiger and yellow fever mosquito expansions, following the nation’s creation. The project uses Google Map Maker with two invasive species vectors of global diseases like Zika, dengue fever participation from the Sudanese diaspora and various organizations and chikungunya. that train people in the digitization of aerial imagery. 2. StatCan crowdsourcing. Launched by Canada’s national statistics 16. Crowdsourcing The National Map, National Map Corps, US. Involves agency (Statistics Canada) to fill gaps in national-level statistics on volunteers collecting and editing data about human-made structures buildings and their attributes with the participation of local people. in 50 states. 3. Monitoring Insects with Public Participation (MIPP) in Italy. Aims 17. Crowdsourced flood resilience in Jakarta, Indonesia. Encourages heads to increase knowledge of species distribution for the conservation of villages to identify critical infrastructure using paper maps and OSM of insects protected under the European Commission (EC) (similar to 13). Habitats Directive. 18. Twitter use in Italian municipalities. A research project into the profile, 4. Global Healthsites Mapping Project. Aims to produce an openly activity and use of Twitter accounts of Italian municipalities. The focus accessible global geolocated healthcare facility using OSM, openly was on the types of messages sent, revealing that culture and tourism available healthcare and crowdsourced location datasets. are the most common topics. 5. North American Bird Phenology Program. A citizen science project 19. National Biodiversity Data Centre, Ireland. Initiated to leverage the digitizing almost a century of bird migration patterns and population potential of outreach groups and the general public for biodiversity status, bringing back to life records from 1880-1970. data survey and observation encouraging the participation of non- 6. Foul & Filthy Rivers, China. Aims to reduce the percentage of polluted professional biodiversity scientists. waters with citizens’ participation. Its Blue Map app is directly linked to 20. Towns Conquer gamification and Instituto Geográfico Nacional government reporting platforms. toponyms database, Spain. Citizens played an Android-based game in 7. Farma Valley Community Map, Italy. Adopts a participatory a competition format to submit suggested corrections and updates to a methodology to collect local placenames missing from current official web-based map service operated by Spain’s National Mapping Agency. basemaps. 21. US National Park Service - Places Project. Uses tools developed by the 8. Community river monitoring volunteer project, Scotland. Asks local OSM community to allows national park visitors and staff to contribute communities to monitor how sediments move in burns (small streams) to mapping trails, tourist sites and other park infrastructure to keep and how this can influence flood management techniques. park maps up to date and accurate. 9. Malawi flood preparedness. The Humanitarian OpenStreetMap Team 22. US Geological Survey (USGS)’s “Did you feel it?”. Driven by a lack (HOT) facilitated community mapping, training the Malawi community of instrumental ground-motion data this website maps seismic and using satellite tracing and Global Positioning System (GPS) field activity reports from the public based on their perception of recent survey techniques. seismic activity in their local area. It is particularly useful in regions of low seismicity. 10. Food insecurity mapping, Bangladesh. Focuses on mapping the unmapped Khulna district, targeting agricultural production. 23. US Census Bureau - Building an OSM community of practice. US Census Bureau employees have built an active volunteer group of OSM 11. Syria Tracker Crisis Map. A recording mechanism for the names and contributors in the bureau. This group advocated for the incorporation locations of victims of the current war. of VGI practices into census operations. 12. iCitizen, South Africa. Aims to involve the local population in reporting 24. New York City open data initiative. The City of New York has released infrastructure problems by collecting geographic data points via geographic data which the company MapBox imports into OSM in mobile phones. partnership with the US OSM community. The city administration receives email updates as changes are made. 13. Community Mapping for Exposure in Indonesia. Aims to reduce vulnerability to natural disasters. It uses the OSM platform to create 25. Imagery to the Crowd, State Department Humanitarian Information thematic maps showing potential damage. Unit, US. High-resolution satellite imagery is made available to the OSM community for the digitization of ground features. Resulting data has been used in humanitarian operations. PART B 26. Humanitarian OpenStreetMap Team mapping in Ulaanbaatar, Mongolia. women and create authoritative land rights for them. Ulaanbaatar’s municipality embraced the volunteering work of university students for OSM data contribution towards the creation of 40. Open for Business, UK. A public authority supported mapping project a “Smart City”. that calls for private business owners to contribute details about the continuation or re-opening of their businesses after a major 27. Mapping schools and health facilities in Kathmandu Valley, Nepal. flooding event. A proactive, crowdsourced mapping project that was disrupted (yet continued on an occasional basis) due to the need to map another 41. International Hydrographic Organization’s Crowdsourced Bathymetry earthquake-struck area. Database. Aims to collect crowdsourced bathymetry data from trusted participants to create a more detailed and updated map of the seas 28. Informal settlement mapping, Map Kibera, Nairobi, Kenya. Map Kibera and oceans. was carried out in the most crowded slum in Nairobi, Kenya, in an effort to create an accurate picture of this dynamic area. 42. Disaster Management, Early Warning and Decision Support Capacity Enhancement Project in Indonesia. Enhances the capacity of disaster 29. Skandobs, Scandinavian predator tracking system, Norway and management agencies through crowdsourced data that will be used to Sweden. A cross-country project developed to collect observations for create and disseminate alert warnings to communities and populations wildlife management and research purposes. at risk. 30. Corine Land Cover 2006 (CLC2006) in OpenStreetMap, France. A data 43. Summer Camp Guide. Under the auspices of the local authority, policy change for CLC2006 enabled its one-off use by the French OSM summer camp managers add their facilities and services to a map, community to update and complete the OSM land use data. enabling citizens to select the place and time that best fits their needs. 31. FixMyStreet for municipality maintenance information, UK. Engages 44. Ramani Huria. University students are trained to provide the necessary the public to provide data about problems in the urban fabric, which are data for resilience and disaster reduction for the flood-prone area of then propagated to the local authority responsible for maintenance. Dar es Salaam. 32. FINTAN vernacular placenames project, Ordnance Survey and Maritime 45. Natural Resources Canada-OpenStreetMap Synergy. An early effort to and Coastguard Agency, UK. A project leveraging “professional” use OSM in collaboration with authoritative data to update and improve crowdsourcing to record vernacular placenames to improve response both datasets. time to distress calls. 46. Participatory mapping and decision support tools for disaster risk 33. Boston StreetBump, US. Based on the use of a mobile app to capture reduction, the Philippines. Training and technical assistance for local bumps or potholes on Boston’s streets, which are then reported to the government units to create basemap information and impact analysis authorities responsible for maintenance. using InaSAFE impact modeling software. 34. California Roadkill Observation System (CROS), US. Collects 47. Government open data usage in Lithuania. Combines conventional and crowdsourced observations of wildlife roadkills and then uses GIS and open data for generic mapping and raises awareness of open data. statistical modeling to predict hotspots, quantify impacts and develop remedial wildlife management actions. 48. Xalapa collaborative transport mapping. Collaborative project to map and document the transportation network of public buses. 35. Crowdsourcing satellite imagery in Somalia. An effort to map all shelters in the Afgooye corridor using satellite imagery to support 49. FloodTags. Short messages posted on social media are used to map estimates of the number and location of refugees. flood risk and help improve preparedness and response actions. 36. Portland TriMet, transportation planner, Oregon, US. The Portland 50. Open Cities, Sri Lanka. Crowdsourced data gathering based on public transportation agency both uses and supports OSM, aiming to accumulated experience in disaster preparedness plus further training provide multimodal transportation plans for citizens. of students and civil servants. 37. The Base Adresse National (BAN) Project. A collaborative project between the French national mapping association, postal service, municipalities and OSM France to create the most update-to-date and complete address database in France. 38. Citizen participation in urban planning, Kirtipur, Nepal. A project where volunteers use a video game to create three-dimensional models of future designs for their local public spaces. 39. Land tenure in Tanzania. A crowdsourcing project that collects parcel borders and land ownership data to fight social injustice against PART B: MAIN REPORT 23 24 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Analysis For the purpose of this report each project’s The analysis identified six key factors that need success (defined as evidence of data creation, managing for successful VGI projects: continued contribution, government use of the 1. Incentives/drivers data, and acceptance of crowdsourcing in the organization) has been rated and compared 2. Scope and aims against the aspects identified in the first report. 3. Participants, stakeholders and relationships First, this looked at the goals of the projects and the reasons for their implementation. Second, 4. Inputs organizational factors were considered, including 5. Technical and organizational aspects the impact of partnerships and/or input from the leading organization. Finally, the overall design 6. Other challenges of the project was considered. These led to The following sections look at each of these topics modelling, evaluated through the QCA analysis, and provide some insights from the individual to identify the most important combination of case studies. factors associated with a successful VGI project outcome. The details of the analysis are beyond the scope of this report and will be provided elsewhere. PART B 1. Incentives/drivers Research and development. This includes exploring new products or research There are many different factors that encourage directions. Examples are CROS, a university-led organizations to search for an alternative to initiative about wildlife roadkills, and Towns institutional data solutions and start VGI projects: Conquer, which explores gamification techniques and strategies to enhance VGI. A slightly different Lack of institutional data in time sensitive approach is taken by FINTAN in the UK, which situations. uses “professional” VGI to collect and preserve This is one of the most common drivers and expert local knowledge. its importance was shown particularly by the aftermath of Haiti earthquake, where there Environmental monitoring through citizen was an urgent need to provide mapping to science. facilitate humanitarian and first aid efforts. The Public participation in scientific data collection Kathmandu case study is a proactive example of is frequently related to monitoring the state of local and international stakeholders teaming up the environment rather than the creation of to provide basic mapping of schools and health new products/research trajectories (compared facilities in case of earthquake. to research and development, above). Examples include the USGS’s “Did you feel it?” public Policy change around governmental data, participation in seismic monitoring; public particularly open data. submission of biodiversity records to Ireland’s Policy changes and decisions to change access National Biodiversity Data Centre (NBDC); and and use of governmental datasets can provide Skandobs crowdsourced wildlife observations. the motivation, or the opportunity for wider public engagement in creating information. The QCA examined each of these drivers and their For example, the adoption of an open data influence on project implementation. In fact, each policy in New York City; the use of the released projects’ drivers can be categorized according CLC2006 data by the French OSM community; to different combinations of these factors. The and the US State Department Humanitarian results show that the most successful projects Information Unit (HIU) sharing high-resolution are those driven by a need for environmental imagery with humanitarian organizations and monitoring, either for management or research volunteer communities. (e.g. understanding species distribution). Low resources and the need to develop Alternatively, where the driver was different, national mapping infrastructure. successful projects benefited from the promotion Interestingly, this situation can be a driver and creation of open data by the government. for both well-established national mapping Such policy changes allow the utilization of agencies (NMAs), as in Canada where updates government datasets for new purposes, as in the are needed to mapping covering 10 million km2, case of the opening of the CLC2006 dataset. It and newly established ones, as in South Sudan might also be that a change in policy opens up where mapping of basic topographic features is an opportunity for governmental organizations needed to build the first data infrastructure of to reassess the way they operate, although this the new state. Similarly, the implementation of requires more detailed research. ambitious efforts like the “Smart City” project in Ulaanbaatar, Mongolia, need support from VGI. In other words, environmental monitoring and citizen science provide a good grounding PART B: MAIN REPORT 25 26 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT for successful crowdsourcing projects in of basic cartographic coverage, and Map Kibera, government, and open data policies can bring where the aim is to create a topographic backdrop opportunities for government agencies to of Nairobi’s largest informal urban settlement. collaborate with the public. Importantly, VGI Update authoritative spatial datasets. projects driven by a lack of resources and the A step further is the aim to support established need to create the information necessary for spatial infrastructures. Keeping datasets up to government operations tend not to be successful. date (mainly through change detection projects) is a crucial task, especially for NMAs as this 2. Scope and aims guarantees the quality of their services. Examples include Natural Resources Canada (NRCan), The scope and aims of the case study projects fall Towns Conquer, New York City’s open data into the following key categories: initiative, Ulaanbaatar’s “Smart City” and the US Basic mapping coverage. Census Bureau. In many cases the aim is as simple as the creation Upgrade public sector services. of a cartographic background for an area of Related to the above is the effort to collect new interest or entire country. In many developing and unrecorded spatial datasets to upgrade countries there is no spatial data infrastructure to the quality of the services provided by a public support humanitarian efforts, disaster prevention sector organization. Examples include the and general planning. Examples include South FINTAN project in the UK, which will improve Sudan, where the new nation was in urgent need HM Coastguard’s response to distress calls, PART B Portland TriMet, iCitizen in South Africa, Boston case demonstrates. However, in both cases StreetBump and FixMyStreet. basemaps are required to proceed, and this requirement can lead to success. Policy development or reporting. Local knowledge and the timely nature of participants’ observations are two of the most 3. Participants, stakeholders important VGI characteristics that stakeholders and relationships are trying to tap into. Examples that combine observation gathering and policy planning are Public sector and NGOs / international CROS, where observations about roadkills help organizations. stakeholders understand the phenomenon This type of cooperation is relevant to various and the best remedial actions, and Skandobs, contexts, including crisis management, where the where wildlife observations inform actions and active participation of all stakeholders is needed to address often difficult and complicated tasks. policies in accordance with new legislation. Those involved in such partnerships include Ireland’s NBDC and USGS’s “Did you feel the United Nations, World Bank, Humanitarian it?” are more report oriented and use the OpenStreetMap Team (HOT), NMAs, state observations to stimulate public engagement and mapping departments and local universities. analysis respectively. The following examples indicate the breadth of Natural disaster preparedness (proactive) such cooperative efforts: Haiti disaster response, and crisis management (reactive). Imagery to the Crowd, Ulaanbaatar’s “Smart As well as being a driver, one of the most City”, Kathmandu and Map Kibera (the latter important aims of existing projects is to build involving national and international bodies). spatial datasets for humanitarian purposes. This Public sector, private sector and NGO covers both crisis management after a natural cooperation. disaster, as in Haiti, and proactive creation of the An even more inclusive type of cooperation necessary infrastructure in an effort to minimize is the additional participation of the private the consequences of a future disaster, as in the sector. For example, MapBox joined forces with cases of Kathmandu, Indonesia and the Imagery the OSM community and the New York City to the Crowd project. government to support data migration and These five factors were examined in combination software provision, in turn gaining valuable against the success of a given project. The analysis insight into the data released. In another case, a shows that updating basic mapping coverage, number of international organizations had the in combination with either improving public active support of Google in terms of software and service delivery (but not addressing a disaster) user motivation to help the government begin the or in addressing a disaster (but not other policy mapping of South Sudan. development) provide the basis for successful Universities and research institutions. crowdsourcing projects. In other words, it is VGI is a relatively new and dynamic possible to consider policy development when phenomenon so there is much active research there is no current disaster (e.g. Indonesian undertaken by universities and research disaster preparedness), but public service institutions on its evolution (e.g. Towns improvement cannot be central while in the Conquer) and possible real-world applications midst of dealing with a disaster, as the Haiti PART B: MAIN REPORT 27 28 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT (e.g. Skandobs, USGS’s “Did you feel it?”, 4. Inputs Ireland’s NBDC and CROS, California). Releasing existing resources. Private and public sector initiatives. This can include the release of datasets (including The private sector can also lead on projects for visual databases) or license changes enabling VGI use in government. One such example is greater use for public users, including the OSM FixMyStreet, run by social enterprise mySociety; community, and private entrepreneurs. The most another is the involvement of the UK’s Ordnance characteristic cases are the Imagery to the Crowd Survey (which functions with market criteria, project, New York City open data initiative and despite being an NMA) in the development of use of CLC2006 in France. The general public and FINTAN for use by HM Coastguard. NGOs are keen to utilize datasets that support their aims, such as producing new map-based The analysis examined the range of partners art products or creating new digital service for in relation to the success of the project. In cyclists, and thus welcome such initiatives. particular, it looked at combinations of public sector, private sector, international NGOs, Direct investment. local NGOs and research organizations, as Directly employing contributors can help well as the overall number of partners. The government agencies ensure that the data outcome shows that successful projects collected meets their immediate needs, while involved an international NGO, either with still benefiting from participation in the broader public bodies, but without local NGOs and ecosystem. In both the Portland TriMet and research organizations; or with private and Community Mapping for Exposure in Indonesia public organizations, but without local NGOs. case studies, government resources were used This highlights the importance of NGOs with to hire in-house mappers to contribute to OSM. experience of crowdsourcing at the interface Community-based projects in areas of very low with government (e.g. HOT) working together income can also benefit from providing financial with public bodies to achieve specific goals. compensation to public participants to replace The resources of private sector organizations earning potential, as Map Kibera did in Kenya. (Google, Digital Globe) are valuable, but they work better with intermediaries, such Investment in new technology. as experienced NGOs. The absence of local The development of new mobile phone or web- NGOs might be explained by the observation based applications to collect data are a common that, apart from in the area of environmental aspect of crowdsourcing projects. Examples monitoring through citizen science, there include the USGS’s “Did You Feel It?” website, the are few well-established local NGOs with the mobile app for monitoring rivers in China and the expertise to act as VGI project champions. Skandobs app for collecting information. Other projects can rely on established technologies which are developed for other purposes or for a general purpose (e.g. OSM). Training. Some projects establish specific training programs, sometimes with local students, to develop skills and knowledge in crowdsourcing PART B geographic information (e.g. the case studies 5. Technical and from Dar es Salaam and Kathmandu). Training organizational aspects increases the local capacity to run and coordinate crowdsourcing activities. The formalization and standardization of VGI is critical to its adoption by government. This issue Research and citizen science initiatives. is both technical and organizational and includes Although giving the public the opportunity to the following aspects: participate in monitoring and research projects has a long and successful history, technological Combining conventional and open source advances, especially in the mobile domain, offer data. a different level of dynamic engagement. The Many crowdsourcing projects use a wide range case studies of Towns Conquer, CROS, USGS, of software applications, both closed and open Skandobs and Ireland’s NBDC all demonstrate source. The CLC2006 in France and NRCan case this type of engagement. studies prove that combining various tools in different applications can widen the technical The analysis showed that many successful horizons of an application and create new VGI examples rely on either direct investment opportunities. However, access to the knowledge and the introduction of new technology, or and experience of using these software tools direct investment and training where projects is not available to everyone and requires high included a relatively low number of partners, technical abilities. with the latter cases using existing technology and providing training to increase capacity. Data integration. The USGS case study is an example of the first During update operations, datasets often diverge scenario: as an organization with a mature but the users and recipients of a project should understanding of crowdsourcing, it is capable know which version is the correct one and how of starting new projects (e.g. transcriptions of to maintain a definitive copy. In some cases, and phenology records) and providing the resources especially where open source software is used, the to create a new website, recruit participants, format of data is not convenient for further use in and manage the overall process. Successful proprietary software packages, making it difficult OSM mapping for resilience is an example of to reintegrate. Therefore, the interoperability of the second scenario: these projects rely on the data formats is a significant issue, for example relative maturity of OSM technology, training the CLC2006 show that after the initial import materials and practices of training new users. of the dataset, the integration of further Finally, successful VGI projects with multiple updates was beyond the interests of the French partners tend not to involve training, the use OSM community. of new technology or direct investment – these Authority. are examples where government agencies tap The authority given to data resulting from VGI into an existing community with the expertise, projects is one of the most challenging issues for knowledge and disposition to assist (e.g. New its use in government. In a departure from an era York City open data initiative and the local OSM in which information is considered authoritative community). simply because it originates from a government organization, recognition of the inherent heterogeneity in geographical information PART B: MAIN REPORT 29 30 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT and the need to keep it up to date permeate to engage more actively in VGI and develop many of the case studies. However, because methodologies and best practices that would government agencies have both the authority be beneficial for all stakeholders, including and the responsibility to provide accurate and governmental organizations, NGOs, and the comprehensive information, this requires more public. This particular mode of engagement can control over the data and its quality. Many of be observed in the case studies of the US Census the case studies, such as the USGS or NRCan Bureau and NRCan. projects, show that governmental organizations The results of the QCA showed that successful need to put appropriate procedures in place projects are particularly defined by the to ensure that, regardless of the source, the presence of a champion in the governmental information released is accepted as trustworthy organization, when the project is not focused on and valuable. updating an authoritative dataset. While there Role of champions. are some rare examples of success for updates This type of engagement stems from individuals’ of an authoritative dataset (e.g. crowdsourcing prior knowledge of the nature and use of VGI the National Map project) these are in the and, often, from their cooperation with local context of mature projects and in organizations communities such as OSM volunteers. Their that accepted the use of crowdsourcing in their personal effort pushes their organizations operational practice. PART B Adopting VGI in government The analysis of the 50 case studies shows that Traditional GIS practices and concern over there are multiple routes to the successful organizational change. implementation of VGI projects involving It is important to highlight that VGI should not government agencies. Unless a government be seen as an activity that replaces the work of organization has already reached maturity in professionals, but as one that complements and the area of crowdsourcing (the best example in enhances it. In some cases, such as Haiti, the use this report is the USGS), identifying a champion of VGI was seen as a direct threat to the business and starting a project that will not address model of the NMA and the OSM data could not authoritative datasets directly is a good way be adopted without consideration of how this to ensure early success and start the process could be protected (see also Intellectual Property of organizational learning on how to use VGI. Rights issues). In other cases, VGI is perceived as Next, if the choice is to use new technologies, a challenge to existing procedures, funding and this should be accompanied by appropriate professional standing, which leads to a negative resources and investment. Using an existing response. There is also a need to integrate VGI technology that has been successful elsewhere processes, including issues of engagement and investing in training and capacity building and feedback to contributors, into established is another path for success. systems, practices and procedures. In some cases, current practice does not require such high Intermediary NGOs (e.g. HOT) are also levels of engagement so organizational transition important factors of success. These is required. This is particularly important organizations have the knowledge and skills to in emergency situations, where entrenched implement projects at the boundary between procedures are necessary to ensure an appropriate government agencies and the public, making response, and capacity to deal with uncertainty valuable implementation partners. Successful and complexity is reduced. If VGI is considered VGI projects also tend to be focused on public for use in crisis response it should be evaluated services or environmental monitoring. Detailed and tried in preparation exercises to ensure that it basemapping of roads, buildings and other is fit for purpose. fundamental data can also provide a solid starting point. Accuracy and reliability. The quality of the data and its application are Overall, there is no one, perfect methodology for among the main questions that must be answered a successful VGI project in government. However, at the beginning of, and during, a project. A there are a series of issues which need to be challenge for projects launched or supported by addressed when considering launching a VGI government agencies is that they are responsible project. Attention to these issues can increase the for providing authoritative data while also likelihood of successful adoption by government integrating input from the public. This makes and acceptance by the public. accuracy and reliability key issues. The case studies demonstrate different levels of accuracy and ways of assessing this (e.g. Community Mapping for Exposure in Indonesia, National PART B: MAIN REPORT 31 32 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Map Corps in the US). Accuracy varies depending clear solution to a significant problem relatively on the new data’s purpose and the existing data quickly. The main differentiation concerns the available as a reference. This is also relevant strategy behind each VGI project and how it to hardware issues in “passive” VGI, in which is recognized by governmental bodies and the data from sensors carried by members of the public, whether as a one-off crowdsourcing event public is shared with government agencies. For or as an ongoing initiative. Maintaining the data example, in the Boston StreetBump case study, and the relevant software is also an issue of the use of smartphone-based sensor technology concern. Financial resources should therefore be produced many “false positives” due to erroneous split across the whole life of a project rather than movement detection. As noted above, there are allocated only to the beginning stages. multiple methodologies for quality assurance in VGI continuation and sustainability. VGI, and these should be explored and integrated In many of the case studies examined there is according to the specific context and aims. no specific plan for longer term adoption of VGI Inherent coverage, temporal and by the governments involved, and some projects participation biases. are done on a short-term basis to accomplish Growing research into VGI projects demonstrates a specific task. Yet, other tasks might also that they can exhibit certain biases in the level benefit from using VGI as a methodology and of participation, as well as when and where the way of operation. Some projects also require activities take place. For example, there is a ongoing engagement and long-term work with problem of coverage bias towards populated and volunteers. However, adoption of crowdsourcing popular places (such as large urban centers) and by government is a process subject to resource crowdsourcing rarely includes a representative management and organizational change. Adopting sample of the general population. An example VGI is likely to require additional resources of a case study limited by such inherent biases in terms of managing the crowdsourcing is the Skandobs project, where the real areas of processes, the data collected and engaging with interest are geographically remote and present the communities involved. Interaction with physically challenging terrain. Special incentives crowdsourcing projects may be terminated may be needed to encourage public involvement by governments for a number of reasons: the from specific regions and known biases of this VGI champion may be redeployed or leave the kind should be considered during the design and organization, there may be a change in policy implementation of such projects. or resources may not be available to continue the engagement. Governments must consider One-off event versus ongoing initiative. long-term plans and assess the sustainability While many VGI activities are conceived as single of their adoption of crowdsourcing. Lessons events, the longevity of updates and maintenance from organizations that have now accepted needs careful consideration. Projects vary crowdsourcing as a way of doing business (e.g. significantly in this respect. For example, the Scottish Environmental Protection Agency and use of CLC2006 in OSM in France lasted many USGS) can help other bodies understand how to months and resulted in an outcome which integrate VGI into organizational thinking. remained unexploited; other projects lasting only a few days, such as the mapping of South Sudan and the Haitian disaster response, provided a PART B Maintaining public interest. Licensing and other Intellectual Property Public interest and participation need to be Rights (IPRs). maintained through the life of a VGI project. Concerns over data ownership and specific As noted in the previous section, the response licensing agreements are another obstacle to the from the participants in the activity is frequently adoption of VGI in government. Incompatibility linked to the framing, the trust that is given and in licensing should be considered at the outset the aims of the project. For example, researchers of the project. For example, the OSM license involved in iCitizen are concerned about how to requires data to be shared-alike and Google owns convince the public to use mobile phones and all data in Google Map Maker. IPRs are especially applications. The experiences of other projects important in geographical datasets because most indicate that solutions are divided between less of the value will not come from a single one but and more economically developed contexts. In from linking and matching it with other datasets. the former, attracting volunteers means offering This means that problems with derivation (the a small amount of compensation to replace time source that dictates the coordinates used to locate that could otherwise be used to generate income. an object on the map) can emerge and create In the latter, innovative techniques such as uncertainty about how the data can be used and gamification and clear targets can be helpful. under which conditions. PART B: MAIN REPORT 33 34 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Integrating data collection and use for provides an example of how this can be resolved, policy analysis or implementation. as an established organization dedicated to A key concern for public contributors to VGI is providing VGI services in crisis situations. More how and where their data and contributions are generally, there is a need to define ownership put to use. Often it is unclear if their information of the process and responsibility over its has been used at all. It is important that the management. The information should be clear connection between data collection and eventual so volunteers can easily identify and contact use by the government is made clear on a the responsible individual, especially in case regular basis. of emergency. Contact points. Conflict between channels of reporting. Open and clear lines of two-way communication In some projects, channels of reporting can be between government and crowdsourcing confusing and conflicting for participants. An communities are vital. Some skepticism around example is the relationship between FixMyStreet, VGI stems from difficulty contacting those which is run by an independent body, and involved, compared to when governments enter the helpline of the local authority. In other into a contract with a third-party company or cases, there are multiple channels, which can organization. When governments provide a be important in crisis situations but require general call for contributions from the public, management. At the same time, it is clear that there is often no way of contacting the leader different participants will prefer different or director of the crowdsourcing community channels so enforcing the use of a single channel and indeed such a person may not exist. HOT can be counterproductive. PART B Summary and way forward This report has reviewed VGI use in government The organizational level is also important. In and identified factors of success as well as contexts where information is mostly provided challenges. Brabham’s (2013) suggestions of by external sources (e.g. private sector) there is crowdsourcing best practices are relevant here, a higher potential for replacing this information such as the need to define the problem clearly, with VGI than if information is maintained ensure commitment from the organization internally when the use of VGI might be seen as and know the online community. This report a threat. This is despite evidence demonstrating identifies several factors specific to VGI projects that successful VGI augments, rather than in government. To summarize, the factors likely replaces, organizational activities. Further to influence the use of VGI relate to the following: organizational issues can be procedural, such as existing legislation and obligations for service • individuals, delivery, or structural, such as responsibilities • organizations, for data collection and use. Specific regulations are emerging to support crowdsourcing and • business models, citizen science in government (e.g. the US • echnical problems and American Innovation and Competitiveness Act of 2016, which explicitly promotes these areas • conceptual issues. to US federal agencies) and these can assist in At the individual level , champions and change convincing organizations to adopt them in their leaders in public sector organizations are critical. operations. In the USA and France there are examples of open data enthusiasts at local government level Next are issues that relate to business models. organizations use VGI. Such champions were – maintenance and dissemination have an incentive and will continue to be – critical in the adoption to use VGI to reduce costs, although additional of geographic information technologies in costs might be involved in communicating government. On the contrary, individuals with with the public and maintaining their interest. worldviews opposed to public participation in However, the use of open data and release of data collection and analysis can block or hinder highly valuable geographic information is a threat the integration of VGI into government. Attempts to the financial viability of other organizations to understand and alleviate the concerns of responsible for selling data. Private sector such individuals can assist in the adoption of organizations are also affected when they have VGI. However, these should not focus solely on committed to deliver public services based on issues of data quality or motivation, but also assumptions that emerging data can be used for address concerns about the status and need for profit. Yet, the extensive use of crowdsourcing professional expertise in an environment where by private sector organizations demonstrates VGI is used. that issues of data ownership and use are not insurmountable and can be accommodated in a way that satisfies both the participants and the organizations concerned. PART B: MAIN REPORT 35 36 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Technical problems include the ability to merge There has been some improvement in practice existing systems with datasets that have been compared to the 2014 version of this report, but changed by the public. This requires various misunderstanding between these two groups skills, from version management to object- remains in terms of time frames, work practices level metadata. In some cases, differences in and problem solving. formatting and fundamental data structure, as At the same time, this report has demonstrated well as semantic ambiguity, add to the challenges that collaboration between government of using VGI fully. In addition, the plurality of organizations and the public is possible and tools and channels through which information beneficial to both parties. Recent technological can reach an organization are significant and societal changes mean that opportunities for challenges. It can also be a challenge to introduce such collaboration are increasing. However, like a new technology (e.g. a new app or system). any cooperation between established institutions Government organizations new to crowdsourcing and external groups or individuals, VGI projects should therefore use more mature technologies, require attention and planning. Successful or team up with NGOs with experience in the projects are not “happy accidents” but evidence area to build their own capacity. of commitment and investment at individual and Finally, conceptual or “worldview” issues need organizational levels, which provide a return in to be recognized and also interact with the above the form of information that would otherwise categories. VGI use requires accepting a higher be difficult to obtain. Practitioners can use the level of uncertainty, attention to heterogeneity lessons identified from case studies across the and the need to work closely with diverse groups world in order to further develop this field. and communicate with the public. It also requires adjusting some of the concepts of “how things are done” – for example, quality assurance 1 The term VGI was established by Michael Goodchild in 2007, in his paper “Citizens as sensors: the world of volunteered geography”. procedures that are suitable for in-house The terms crowdsourced geographic information and VGI are often processes are methodologically unsuitable in the used interchangeably. 2 See Haklay et al. 2014 for the first report crowdsourcing context. This can be challenging 3 The website can be visited at https://crowdgov.wordpress.com/ to people who are used to working in an isolated and top-down manner. In addition, perceptions of VGI as professionally threatening should be accepted as genuine and reasonable, and need to be addressed as noted above. To conclude, this report has documented the early and maturing stages of VGI use in government. It has looked across 50 case studies but further research is necessary to explore the factors influencing the success and failure of VGI projects for government use in detail. Some of the problems, such as licensing terms, will require specific effort from both governmental organizations and crowdsourcing communities. Appendix: Case studies This part of the report presents a detailed Readers may find it helpful to use the quick analysis of the 50 case studies included reference table, which includes the key in the QCA study. The case studies are elements of each case study. from the 2014 edition of this report and about 20 new case studies were added. The case study information was compiled in conversation with the experts who ran the projects, from replies to the blog-based questionnaire circulated via social media and mailing lists, or based on research using secondary resources (websites, existing publications, messages on forums, etc.). APPENDIX: CASE STUDIES 37 38 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Case studies: structure and overview Each case study provides an example of the use of VGI by government or by the public, and summarizes the context, positive and negative outcomes, and main lessons. The case studies are presented in full in a later section and are all deliberately short, with the same structure. First, a summary table provides general information about the case study in the format below. Second, a written summary of the project covers the context, description of the project and discussion of the positive and negative aspects of the collaboration. Finally, each case study closes with bullet points indicating the most important lessons from the project. Sample case study table title Interaction type The flow of the data (either crowdsourced or authoritative), in one of the following categories: Public → Government, Government → Public → Government, or Public → Government → Public. Trigger event A specific event that might have triggered the data sharing (e.g. change in data license, natural disaster, etc.). Domain The area in which the datasets have been used. This includes an abstract characterisation of the general area (e.g. generic mapping) as well as information about the specific field (e.g. update of national topographic database). Organization(s) The organization(s) that initiated the data sharing process and those that have been actively involved in the whole project. Actors Interested parties/stakeholders that have contributed to, or benefited from, the data sharing process in any way. Datasets The datasets that have been shared and used by the public or project partners (including new datasets generated). Process The process followed to implement the data sharing, data integration and cooperation. Feedback The immediate result returned to the initiator of, and/or participants in, the data sharing process, if any. Goal The original goal of the project and reason for exploring crowdsourced geographic information. Side effects Any other issues or outcomes. Impact of the The impact falls into three categories: local, national and governmental. project Temporal pattern The longevity and maintenance of the project (i.e. ongoing or one-off initiative). Funding The organization which is behind each project and supports it financially. Contact point The person who has either provided details about, or been significantly involved in, the case study, when available. APPENDIX Quick reference The table below provides summary information on each case study by thematic category. Environmental Natural Updating Upgrading monitoring / disaster authoritative public sector Project title citizen science preparedness spatial datasets services Mosquito Alert project in Spain Yes - - Yes StatCan crowdsourcing - - Yes Yes Monitoring Insects with Public Participation Yes - Yes - (MIPP) in Italy Global Healthsites Mapping Project - Yes - Yes North American Bird Phenology Program Yes - Yes - Foul & Filthy Rivers, China Yes - - - Farma Valley Community Map, Italy - - Yes - Community river monitoring volunteer Yes Yes Yes - project, Scotland Malawi flood prepardness - Yes - - Food insecurity mapping, Bangladesh Yes Yes - - Syria Tracker Crisis Map - Yes - - iCitizen, South Africa - - - Yes Community Mapping for Exposure in - Yes - - Indonesia Haiti disaster response - Yes - - Mapping of South Sudan - - - - Crowdsourcing The National Map, National - - Yes - Map Corps, US Crowdsourced flood resilience in Jakarta, - Yes - - Indonesia Twitter use in Italian municipalities - - - Yes National Biodiversity Data Centre Ireland Yes - Yes - Towns Conquer gamification and Instituto - - Yes - Geográfico Nacional toponyms database, Spain US National Park Service – Places Project - - - - APPENDIX: CASE STUDIES 39 40 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Environmental Natural Updating Upgrading monitoring / disaster authoritative public sector Project title citizen science preparedness spatial datasets services USGS’s "Did you feel it?" Yes Yes Yes - US Census Bureau – Building an OSM - - Yes - community of practice New York City open data initiative - - Yes - Imagery to the Crowd, State Department - Yes - - Humanitarian Information Unit, US Humanitarian OpenStreetMap Team mapping - - Yes Yes in Ulaanbaatar, Mongoli Mapping schools and health facilities in - Yes Yes Yes Kathmandu Valley, Nepal Informal settlement mapping, Map Kibera, - - - - Nairobi, Kenya Skandobs, Scandinavian predator tracking Yes - - - system, Norway and Sweden Corine Land Cover 2006 (CLC2006) in Yes - - - OpenStreetMap, France FixMyStreet for municipality maintenance - - - Yes information, UK FINTAN vernacular placenames project, Yes Yes Ordnance Survey and Maritime and Coastguard Agency, UK Boston StreetBump, US - - - Yes California Roadkill Observation System Yes - - Yes (CROS), US Crowdsourcing satellite imagery in Somalia - Yes - - Portland TriMet, transportation planner, - - - Yes Oregon, US The Base Adresse National (BAN) Project - - Yes Yes Citizen participation in urban planning - - - Yes (Kirtipur, Nepal) Land tenure in Tanzania - - Yes Yes Open for Business, UK - Yes - Yes APPENDIX Environmental Natural Updating Upgrading monitoring / disaster authoritative public sector Project title citizen science preparedness spatial datasets services International Hydrographic Organization’s - - Yes Yes Crowdsourced Bathymetry Database Disaster Management, Early Warning and - Yes Yes Yes Decision Support Capacity Enhancement Project in Indonesia Summer Camp Guide - - - Yes Ramani Huria - Yes - Yes Natural Resources Canada-OpenStreetMap - - Yes Yes Synergy Participatory mapping and decision support - Yes - - tools for disaster risk reduction, the Philippines Government open data usage in Lithuania - - Yes - Xalapa collaborative transport mapping - - - Yes FloodTags - Yes - - Open Cities, Sri Lanka - Yes Yes Yes APPENDIX: CASE STUDIES 41 42 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 1. Mosquito Alert project in Spain Interaction type Public → Government Trigger event Concern about the Asian tiger mosquito and subsequently awareness of mosquito impact through the spread of the Zika virus. Domain Health Organization Mosquito Alert Actors CREAF (Centre de Recerca Ecològica i Aplicacions Forestals), ICREA (Institución Catalana de Investigación y Estudios Avanzados) and CEAB-CSIC (Centro de Estudios Avanzados de Blanes), Dipsalut (Organismo de Salud Pública de la Diputació de Girona) and FECYT (Fundación Española para la Ciencia y la Tecnología), Obra Social “la Caixa” and Redbooth Datasets New mobile applications for citizen contributions about the locations of mosquito observations, a well-designed website and Google Maps as a basemap. Process Six stages: (1) citizens report their observations, (2) experts validate them, (3) spatial and attribute data is stored in an online and dynamic map, (4&5) the research community and public sector view the results and (6) awareness is raised through education. Feedback An online and interactive map, scientific publications and educational seminars and workshops. Goal Interactive map with spatial and attribute information on mosquitoes, which can be used to raise awareness among citizens, experts, the research and educational community and public sector. Side effects - Impact of the project National Temporal pattern Ongoing Funding Dipsalut and FECYT Contact point Aitana Oltra (CREAF) APPENDIX Mosquito Alert project in Spain with Mosquito Alert. Numerous municipalities are directly connected to the project so the Mosquito Alert is a citizen science project project’s tools may be transformed into launched to fight the tiger and yellow fever preventative actions. mosquito expansions, two invasive species 6. Periodic education and awareness campaigns vectors of global diseases like Zika, dengue and to combat mosquito vectors of disease. chikungunya. The project has six stages: Mosquito Alert “learning by doing” activities 1. Without special training, citizens are and innovative workshops promote the asked to use a new mobile application to measures that must be taken in homes to (a) report any possible breeding sites, (b) prevent mosquitoes from proliferating. report sightings of tiger or yellow fever In terms of strengths, the project relies on mosquitoes, (c) validate photographs from smartphone cameras, which are widely available other participants, and (d) consult the map of and make data collection straightforward once citizen observations. The mobile application participants have downloaded the application. works well both with Android and Apple Linking to other projects through collaboration smartphones. Citizen contributions are with the European Citizen Science Association crucial as the submitted information may be increased the project’s potential reach and the key to generating a participatory alert expanded it to other locations. The connection system to improve management of these to environmental management means that species and minimize the risk of disease public authorities can deal with the source of the transmission. The submitted photographs mosquitoes. will be part of a database used for research, monitoring and control of the tiger and yellow The main challenge is the location of the small fever mosquitoes. project team in a research institute, which limits the resources available to grow and extend the 2. All tiger mosquito findings with photographs project to new locations. and breeding sites in the public area are validated by expert entomologists from Main lessons: Mosquito Alert. • Citizens can be given advice on coping with 3. Once validated, these observations appear health and environmental problems. on the map. The dynamic map includes information about the type of mosquito, • Projects can be extended to allow for data possible breeding sites and distribution of analysis by researchers, the public sector and participants. The interactive map uses Google schools after data validation. Maps as a basemap. • Emergent situations, such as growing concern 4. Mosquito Alert focuses on the development of over Zika, can provide the impetus for predictive methods and models that combine ongoing processes. citizen data with data from other authoritative • A straightforward procedure guarantees sources, such as monitoring or control viability and success. entities, universities, etc. 5. A toolkit is provided for public sector representatives interested in collaborating APPENDIX: CASE STUDIES 43 44 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 2. StatCan crowdsourcing Interaction type Government → Public → Government Trigger event - Domain Urban mapping Organization Canada’s national statistics agency (Statistics Canada) Actors Statistics Canada, OSM local community, municipalities and local people Datasets OSM for data editing and open source dashboard for monitoring. Process Local people fill gaps in national-level statistics on buildings and their attributes. Feedback Updated OSM in Ottawa and Gatineau. Goal Updated georeferenced attributes of buildings and valuable information for various Statistics Canada divisions. Side effects - Impact of the project Governmental Temporal pattern Ongoing Funding Statistics Canada Contact point - APPENDIX StatCan crowdsourcing tags contributed for each city. It also shows the proportion of buildings by type and the number of Canada’s national statistics agency (Statistics missing address fields. Canada) launched a pilot crowdsourcing project in October 2016 relying on the principle that Statistics Canada is interested in understanding individual citizens are experts in their local how a national statistical office can mobilize environments. Volunteers are asked to fill gaps contributors and existing technologies. They in national-level statistics on buildings and also hope the project will teach them about the their attributes. Taking advantage of their local possibilities and limitations of crowdsourcing. For knowledge, citizens can input the location, the time being, traditional surveying will continue physical attributes and other features of buildings but the project will help assess the potential of such as type of use. This crowdsourced data supplementing it with crowdsourced data. aims to fill gaps in national datasets and produce In the second year of the pilot, Statistics Canada valuable information for various Statistics intends to develop a mobile app that will allow Canada divisions. contributors to map on the go. Outreach will Crowdsourced data is collected and edited in be maintained and quality assessments will be OSM facilitated by a customized version of OSM’s conducted as more data are collected. iD-Editor. This adapted tool allows participants Main lessons: to seamlessly add points of interest and polygons to OSM although the project information • Crowdsourced projects of this kind are based underlines that OSM is a free tool that may not on the principle that individual citizens are reach accessibility expectations. The platform experts in their local environments. also includes instructions on how to sign up for • Crowdsourced surveying can fill gaps in and use OSM, allowing beginners as well as more traditional data collection. experienced users to contribute. • Open source dashboards can be used to Statistics Canada initiated the two-year pilot monitor crowdsourcing projects. project to understand the potential of data crowdsourcing for statistical purposes. To • Stakeholders need to dedicate time to promote the project, meetings and conference understanding how projects work and how calls with people from the local (Canada/Ottawa- they can benefit. Gatineau) OSM community took place before the project start. A webpage and communications campaign were launched in September 2016 to encourage citizens of the cities of Ottawa and Gatineau to participate. More than 200 people signed up for updates on the project within the first month. The project is monitored using an open source dashboard developed by Statistics Canada, which records the number of buildings mapped, the number of users and the average number of APPENDIX: CASE STUDIES 45 46 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 3. Monitoring Insects with Public Participation (MIPP) in Italy Interaction type Government → Public → Government Trigger event - Domain Environmental monitoring Organization National Centre Forest Biodiversity Actors National Centre Forest Biodiversity, State Forestry Corps, Ministry for the Environment, Land and Sea, Lombardy Region, Sapienza University of Rome, Roma Tre University, Agricultural Research Council (CRA) Datasets Occurrence data of nine species listed in the EC Habitats Directive. Process Citizens upload spatial attribute data and pictures of insects via the web or an app. After validation by specialists, data are saved on the project map and in the government database on protected species. Feedback Record on map. Goal Increase knowledge on the distribution of species. Side effects - Impact of the project National Temporal pattern Ongoing Funding European Commission LIFE Programme Contact point Alessandro Campanaro APPENDIX Monitoring Insects with Public This project has increased knowledge on the distribution of protected species by 20-30 Participation (MIPP) in Italy percent. Citizens have also gained awareness The Monitoring Insects with Public Participation of forest biodiversity, Natura 2000 guidelines, (MIPP) citizen science program is part of a beetle species and monitoring. However, project to develop standard, national-level according to the project coordinators, only monitoring protocols for the conservation some sections of society have participated of insects protected under the EC Habitats (e.g. nature photographers, park rangers, Directive. In many European countries the hikers) and participation will be analyzed by an protocols are standardized through national ongoing sociological study. Funding is currently guidelines but this has not been done in Italy. guaranteed but may not be provided by the Italian government when the EC funding period ends. MIPP is a State Forestry Service project co- funded by the European Commission under the Main lessons: LIFE Programme and conducted with a range • EC funding can support citizen of partners (see summary table). It operates science projects. mainly in five areas managed by the National Forestry Service: the Tarvisio Forest, the Natural • Various channels of training Reserve of Bosco Fontana, the Casentino Forests, encourage participation. the Castle of Sangro and the Natural Reserve • Validation of submitted data may improve Mesola Wood. accuracy and quality. A team of coordinators is The main objective of MIPP is to develop and test required for quick and reliable evaluation. methods for monitoring five beetle species listed • Only certain sections of society may become in Annexes II and IV of the Habitats Directive. involved and more social study is required to MIPP tests methods based on direct observations, understand this. genetic analyses and trapping. A second objective is the collection of fauna data gathered from citizens via the project website or smartphone application. From 2014 to 2016, over 2,300 records were collected and validated by specialists, with nearly 1,700 (73 percent) confirmed. Step-by-step instructions are given on the project website, advising contributors how to send reports, complete the form, update the map with spatial observations and submit attribute data about insects and the place of sighting. Participants can also upload a photograph of the insect and provide other relevant information. Technical support days, seminars and workshops support and disseminate the project. APPENDIX: CASE STUDIES 47 48 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 4. Global Healthsites Mapping Project Interaction type Public → Government → Public Trigger event Ebola epidemic in West Africa (2013-2016), among other issues. Domain Health Organization Global Healthsites Mapping Project Actors Rwandan Ministry of Health, Tanzanian Ministry of Health, International Committee of the Red Cross, World Health Organization, Health Informatics Society of Sri Lanka, UNICEF, OSM, British Red Cross, United States Agency for International Development (USAID) Datasets Imported from OSM and other trusted providers of geolocated health data and information. Produced a geolocated health care facility dataset with a limited set of attributes. Process Health facility data imported from OSM and exposed to interested parties with an invitation to improve the accuracy of the data. This curated data is then shared openly as Shapefiles and an open API (Application Programing Interface). The process will complete the data lifecycle and return enriched data back to OSM. The public can also edit the health facility data directly. Feedback Participants can log on and edit information about health facilities. The Global Healthsites Mapping Project will make the data accessible over the internet through an API and other formats such as GeoJSON, Shapefiles, KML, CSV. Goal Long-term curation and validation of global health care location data – essentially an open data commons of health facility data. Side effects Intended future import of updated health facility information back into OSM. Envisaged wider reuse of health facility data and information for applications and citizen use. Impact of the project Governmental Temporal pattern Ongoing Funding Intention between the healthsites.io platform and partners to seek out ways to sustain (in terms of financial, infrastructural and human resources) its continued development and improvement. Contact point Mark Herringer APPENDIX Global Healthsites Mapping Project the project has not yet deployed operationally. Operational deployment is essential as it will assist The mission of Global Healthsites Mapping Project in driving the feedback process. Organizations are is to help supply governments, NGOs and the currently cautious about sharing healthsite data private sector with accurate and up-to-date health as they are concerned with the liability and quality facility information. The project is in a pilot phase of VGI. and is inviting governments, NGOs and individuals to share data and establish an open data commons There is hesitancy to directly associate with of health facilities in the OpenStreetMap project. a Healthsite update. This is a fundamental requirement of open data. Unfortunately there The incentive and driver for this project was, is a lack of consolidated learning from end users among other emergencies, the response to the and to date there is no operational use case and Ebola outbreak in West Africa. This region has working agreement with end users in an organized low resources and a great need for infrastructure innovation process that generates and integrates support. Exposure to primary health care work and evidence to guide the project and define the social a goal of sharing this knowledge and insight in an business model. The underlying assumption is open development context initiated the project. that Healthsites will improve health outcomes The Ebola response then presented a real need for through the publication of accurate and up-to-date accurate health facility data. open health facility data. Until this assumption is tested it is difficult to quantify the value of accurate The dataset produced is a geolocated health care health facility data and make a business case that facility dataset with a limited set of attributes is aligned with the achievement of Sustainable focused on facility type, location and essential Development Goals. services available. Data is also imported directly from OpenStreetMap and other trusted provides of Main lessons: geolocated health data and information. The data is exposed to interested parties with an invitation to • Triggers like Ebola create a real need for improve its accuracy. accurate health facility data in areas with low resources and limited infrastructure. There are a number of positive aspects to this • Multi-partner projects can benefit from the case study. The curated data is then shared openly contribution of various actors, from launching under the Open Database License (ODbL) through a pilot project to active involvement in making downloadable Shapefiles and an open API. The the project work. process is working towards completing the data lifecycle and returning enriched data back to • Citizens can edit data by using social media OpenStreetMap. OpenStreetMap is seen as the identification for authentication. Thus, baseline repository for this data. Citizens can easily registration process may be bypassed. edit the data in the Healthsites databases by using • Organizations remain cautious about sharing their social media identification for authentication. data as they are concerned about the liability As an open source project, all data is published as and quality of VGI. open data. The public are continually invited to engage in an Agile development process. • This project has not yet deployed operationally which means that its side effects are still There are some downsides to the project. Project unknown. communications have been slow with some project partners. At the time of writing, June 2017, APPENDIX: CASE STUDIES 49 50 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 5. North American Bird Phenology Program Interaction type Government → Public → Government Trigger event Global climate change Domain Environmental monitoring Organization USGS Actors Federal government, academics, non-profits, individuals Datasets Transcription of historical data cards regarding when birds arrived. Process Historic cards from around North America (1880- 1970) are scanned (1.1 million or so), a computer program allows online volunteers to transcribe data into the database, computer algorithms are used to compare multiple transcriptions to create validated records, volunteers and USGS personnel work on validating locations not in gazetteers, observer names and bird names. Feedback Transcribed records. Goal Volunteers enter data and help run the system. Side effects - Impact of the project Governmental Temporal pattern One-off Funding US Federal government Contact point Sam Droege, US Geological Survey APPENDIX North American Bird projects main factors of success. Phenology Program However, this project has been severely affected by a lack of resources. First, due to a lack of The North American Bird Phenology Program support it often took years to correct problems (BPP), part of the USA-National Phenology with the algorithms that run the program. Network, flourished from 1880 until 1970 when Second, lack of funding has led to the loss of a lack of funding led to its closure. The project dedicated coordinator. was coordinated by the federal government and sponsored by the American Ornithologists’ Union. While cumbersome, the program has successfully The aim of the project was to record information entered hundreds of thousands of verified on first arrival dates, maximum abundance and records. Additionally, while the costs may have departure dates of migratory birds across North been such that it would have been cheaper to America using a network of volunteer observers. employ professional data entry personnel, the It exists now as a historic collection of six million data would not have been added to the database migration card observations, illuminating almost without the novelty of the data entry program. a century of migration patterns and population The program still has volunteers transcribing data status of birds. and a website/data entry portal at: https://www. pwrc.usgs.gov/bpp/. The historic records permit a detailed study of geographic and temporal shifts in arrivals and Main lessons: departures . The government has housed the historic records and now an innovative project • Citizen science has existed since at least 1880 has launched to curate the data and make them and may be a viable solution for a greater publicly available. The records are being scanned participation and wider data collection. and placed on a website, where volunteers • Dedicated volunteers can contribute high worldwide transcribe the records and add them to quality transcriptions of records in a reliable a public database. In other words, the scope of the way. current project is to update authoritative spatial datasets for environmental monitoring through • Lack of resources can be a significant obstacle citizen science. Citizens located in the Baltimore- to project viability. Washington area may work as office volunteers while other participants may digitize data online. Training is provided with detailed information via the website and email assistance. A series of maps indicating migratory birds across North America over the decades has been designed with the material digitized by the volunteers. The project has recorded millions of transcriptions by often extremely dedicated volunteers over many years. Most are high quality transcriptions with relatively few incorrectly entered pieces of information. Recruitment of dedicated project volunteers was among this APPENDIX: CASE STUDIES 51 52 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 6. Foul & Filthy Rivers, China Interaction type Public → Government → Public Trigger event China’s Water Ten Action Plan (April 2015) Domain Environmental planning Organization Institute of Public & Environmental Affairs (IPE) Actors People’s Republic of China (PRC) Ministry of Environmental Protection (MEP) and Ministry of Housing and Urban-Rural Development (MoHURD), Institute of Public & Environmental Affairs (IPE), local NGOs in the Green Choice Alliance (GCA) and general public Datasets Interactive live map of the locations of “foul and filthy rivers”. Process Through a series of meetings and communications, IPE officially launched a cooperative relationship with MEP and MoHURD to link IPE’s Blue Map app directly to the government reporting platforms that had initially been established via WeChat. Following integration, reports raised via the Blue Map app are directly sent to government ministries, whose responses are also sent directly to the app for display. Feedback Interactive real-time map of polluted urban water bodies across China’s cities, communication records of government responses to public reports, live tracking of government’s clean-up efforts. Goal 1. Enable the public to report polluted bodies of water not initially on government lists designated for clean-up. 2. Allow the public to effectively supervise government efforts to rectify polluted waters. Side effects Initial reporting mechanism led to the realization that many polluted waters designated for clean-up were not on track to meet government deadlines. Therefore, while the initial module in the app only included a function for the public to report polluted rivers not already designated for clean-up, a second function was incorporated for the public to upload “supervision” reports about on-site progress (or lack thereof). Impact of the project National Temporal pattern Ongoing Funding The Blue Map App (蔚蓝地图) is supported by SEE Foundation and Alibaba Foundation. Contact point Kate Logan APPENDIX Foul & Filthy Rivers, China progress toward clean-up, with an emphasis on if the waters are on track to meet their government- To address China’s severe water pollution set deadlines, many of which are as early as 2017 challenges, the Chinese government issued its and 2018. For those waters already deemed clean, Water Pollution Prevention & Control Action local partners may assess if the judgment appears Plan in April 2015, a national plan to improve to be accurate and may upload photographs to the quality of China’s rivers, lakes, aquifers support their observation. As the foul and filthy and other bodies of water. Part of the Water rivers policy is ongoing, the app’s platform will Ten action plan, as it is commonly called, is an continue to collect, collate and display reports initiative called hei chou he, which translates as and to help support the government in verifying “black and smelly” or “foul and filthy” rivers. The that their clean-up work is satisfactory to initiative aims to reduce the percentage of waters the public. in urban areas designated as “foul and filthy” to less than 10 percent of the total by 2020. Those The successes of the foul and filthy rivers remaining must be cleaned up completely by reporting module are largely a result of two 2030. The Ministry of Environmental Protection factors: 1) transparency (the platform is (MEP) and Ministry of Housing and Urban-Rural completely public so anyone can view the reports, Development (MoHURD) launched the initiative photographs and responses); and 2) interactive just after the 2016 Chinese New Year holiday. channels for public participation (any user In addition, the two ministries created a public can complete a report as part of an official and reporting mechanism via popular Chinese social responsive government mechanism, which allows media app WeChat, where citizens can report the government to tailor its clean-up efforts to polluted bodies of water to be included in the satisfy the public). clean-up list and receive a guaranteed response One of the shortcomings of the case study is the from the government within seven working days. difference in government responsiveness across Following a series of communications with MEP regions and limitations of public power to push and MoHURD, IPE’s integrated Blue Map app for greater responsiveness. Another limitation is created an interactive, live map module and achieving the sustained involvement of the public. reporting mechanism, which links directly to the Main lessons: two ministries’ back-end system. The module allows users located within 10 meters of a body • All reports made via an app and government of water to upload a photograph and complete a responses can be made publicly accessible. report, which will then be submitted directly to • Platforms can be updated in real time the two government bureaus. The photographs with new information about reports and text of the report will also appear on an and observations. interactive map on the app, and government responses are posted directly to the app as part of • Platforms like this enable the public to directly the report. interact with the government, providing an official channel for citizens to raise concerns. Building on its initial work to report polluted waters, IPE adjusted the module on the app to • Users need to be logged into an account before add a second option to report an observation. being able to upload a report. This option allows IPE to better collaborate with NGO partners to supervise on-the-ground APPENDIX: CASE STUDIES 53 54 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 7. Farma Valley Community Map, Italy Interaction type Public → Government Trigger event Location-based game contest. Domain Heritage preservation and land planning. Organization Attivarti.org Actors Attivart.org (NGO) and citizens of Torniella, Piloni and Scalvaia, small hamlets in southern Tuscany Datasets 1:10000 basemaps (authoritative), placenames (crowdsourced). Process Participatory methodology (meetings, interviews, surveys) to collect placenames missing from current official basemaps but relevant to local communities. Feedback Increased international and national visibility, extension of open data originated by Attivarti.org and its network of partners, improved knowledge about the Farma Valley. Goal Improve visibility of the Farma Valley, empower citizens via open data. Side effects Interest of other communities in Italy to initiate the same process (preliminary meetings have taken place). Impact of the project Local Temporal pattern Ongoing Funding Community-based funding (citizen-owned) but searching for funding from other sources. Contact point Andrea Giacomelli (info@pibinko.org) APPENDIX Farma Valley Community Three key positive factors of the project have been recorded to date: Map, Italy 1. The project has created a base to document The Farma Valley, located in southern Tuscany heritage from the three hamlets using an open about 100km south of Florence, is small (120 data paradigm; km2) and sparsely populated (less than 500 residents, mostly concentrated in three hamlets). 2. The dissemination of the dataset has created The valley used to be crossed by one of the an additional layer of interest about the valley roads between the coast and Siena. However, from tourists, researchers and professionals with the opening of a new highway in 1974, the (especially in the tourism and environmental hamlets have been progressively cut off from education sectors); the main flows of traffic (including tourism). 3. The project represents an effective case This has reinforced the valley’s reputation as of a value chain starting from research a natural paradise but has not simplified the efforts (related to the INVOLEN project life of local communities. The arrival of an participation) and resulting in the promotion expert in environmental geomatics has led to of the local economy. the development of innovative projects for the protection and promotion of the area. The whole process exhibits appropriate project management practices. This project started with the participation of the local communities, coordinated by Attivarti. There are few negatives to date. With the project org, around a contest concerning location-based developing during the coming months, and more games for environmental education, launched information being collected, work is needed by the EU-funded INVOLEN project. The scope on consolidating the relationship between of the project was based on an initial effort to the project and its institutional stakeholders. collect points of interest and trails known to the Currently, the project is in an active phase, which residents and to expose a significant number may conclude towards the end of 2017. of toponyms missing from the current regional This project could be replicated in other government basemaps in the area of Torniella. contexts as the technology is fairly simple, but To extend the coverage of the community- the key element is the mix of actors, linking a based dataset, subsequent efforts were made to scientific and technologic subject to a strongly interview older residents of the other hamlets. felt driver (the need to minimize the loss of Contributions from the public are represented by local knowledge). the Web Map Service (WMS) provided by Regione Main lessons: Toscana with the 1:10000-scale basemaps. • VGI can attract the interest of tourists, researchers and professionals. • Maintaining public interest is key, and a well- known challenge in other VGI contexts. • Drawbacks of a project may not be clear until it concludes. APPENDIX: CASE STUDIES 55 56 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 8. Community river monitoring volunteer project, Scotland Interaction type Public → Government → Public Trigger event The Conservation Volunteers (TCV) Scotland and Clackmannanshire Council wanted to improve volunteer understanding of the interactions between sediment movement, chokes and blockages over time in the context of flood risk management, and investigate how sediments in highly dynamic burns (small streams) may move over time. Domain Flood preparedness Organization Community river monitoring volunteer project. Actors TCV Scotland, Clackmannanshire Council, the Scottish government and citizen scientists Datasets Citizen contribution to Clackmannanshire Council’s database. The sediment monitoring augments Clackmannanshire Council’s existing datasets. Process Citizens use the Monitoring Sediment Movement and Blockages recording sheet, smartphones and cameras to report their observations and the TCV team and local authority comes back to them via email/phone call to inform them about the outcomes. Feedback Update of governmental datasets with crowdsourced observations. Goal Raise awareness of flood risk in the council area and get local communities involved in recording useful information about some of the Hillfoot Burns. Side effects The data source has fed into Clackmannanshire Council’s forthcoming Flood Risk Assessment (FRA) options appraisal report for Tillicoultry produced by an external contractor, JBA Consulting. Impact of the project Local and national Temporal pattern Ongoing Funding The Scotland Counts Partnership: Scottish government, Scottish Natural Heritage (SNH), Forestry Commission Scotland (FCS) and Clackmannanshire Council. Contact point Amanda Malcolm (TCV) and Stuart Cullen (Clackmannanshire Council) APPENDIX Community river monitoring community, and Clackmannanshire Council has embraced the opportunity to collaborate. The volunteer project, Scotland only limitation for some users was the need to With support and funding from the access the internet to submit data, where there Scottish government, TCV Scotland and was a lack of access to smartphones and internet Clackmannanshire are working in partnership connection, or IT skills. to deliver a citizen science community river The final report has been passed to flooding monitoring volunteer project. The project’s consultants JBA Consulting Ltd., and used in their aim is to help raise awareness of flood risk in “Flood Risk Options Appraisal in Tillicoultry” the council area and get local communities report for Clackmannanshire Council. involved in recording useful information about some of the Hillfoot Burns. Through the project, Due to its success, the project will continue local communities and volunteers are actively in 2017, with the TCV team commencing the recording data and taking photographs to monitor second phase of the project: Become a Flood how sediments can move in burns and how this Warden Volunteer. can influence flood management techniques. Main lessons: In this project Clackmannanshire Council assisted The Conservation Volunteers (TCV) – an • The scope of the project and requests to environmental NGO – to identify gaps in data volunteers should be simple, clear and upfront. linked to flooding and ways to fill these through • Short introductory meetings can be useful to a citizen science pilot project with volunteers. break down barriers and build confidence in The main target of the project is to create new new users. datasets for the local authority. • Long-term contact is most effective with Volunteers map their observations by taking a volunteers – it builds confidence and photograph from the same vantage point each encourages users to continue recording time they are out recording river information. and monitoring. The locations picked by the volunteers and TCV team are noted on their maps as a reference for the photograph locations and for consistency. Contributors are provided with feedback and thanks to indicate that the contribution has been successfully received by TCV Scotland and submitted to Clackmannanshire Council. If a blockage was reported by a contributor and removed by the local authority, the contributor received a follow up email/phone call to confirm that the debris had been cleared from the watercourse. TCV provide training and support for the volunteers. The project has been very positive, inclusive and well received by the local APPENDIX: CASE STUDIES 57 58 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 9. Malawi flood preparedness Interaction type Government → Public → Government Trigger event Natural disaster (flood) Domain Disaster preparedness, survey of households. Organizations Humanitarian OpenStreetMap Team (HOT), Global Facility for Disaster Reduction and Recovery (GFDRR), the Department of Disaster Management Affairs (DoDMA), Surveys Department Actors Surveys Department, Department of Disaster Management Affairs (DoDMA), Housing and Physical Planning, Water Resources, Crops and Land Resource Conservation, and local universities Datasets InaSAFE, OSM, Satellite Imagery (Bing & WorldView-2), MASDAP. Process Satellite tracing and field survey using GPS. Feedback Maps of infrastructure (accessible through Malawi Spatial Data Portal (MASDAP)) and training manual for capacity building and future mapping projects. Goal Improved flood preparedness and contingency planning for floods; improved coordination between government and non-government institutions (data sharing). Side effects Actors are aware that open source software (OSM, QGIS, InaSAFE) is able to support flood contingency planning in the region. Impact of the project Governmental Temporal pattern One-off Funding GFDRR Contact point Severin Menard, Maning Sambale, Emir Hartato & Francis Nkoka APPENDIX Malawi flood preparedness open source geospatial data portal based on GeoNode and Geoserver. Key geospatial data In 2014 the Humanitarian OpenStreetMap Team is available in the portal for download by the (HOT) was involved in a community mapping public. and training project with the Malawi Department This case study has three key positive factors. of Disaster Management Affairs (DoDMA) to First, there was a follow up of the project due to a prepare for flooding, a severe natural disaster. flood that critically affected the southern parts of From late July to late December 2014, HOT Malawi along the Lower Shire River. The response carried out a project in the Lower Shire, the large was instant and the project was amended to valley to the south whose districts, Chikwawa serve the specific needs of the situation. Second, and Nsanje, are the most flood-prone areas of the interns from the university become a focal point country. The project was funded by the World for OSM/QGIS/InaSAFE training. Third, Malawi Bank Global Facility for Disaster Reduction and Spatial Data Portal (MASDAP) now includes open Recovery (GFDRR), for which Malawi is one of source data (e.g. OSM). the nine African priority countries. In addition Among the challenges that HOT had to cope with to GFDRR and the Department of Disaster is that infrastructure development was required Management Affairs (DoDMA) the project to support the mapping efforts as internet engaged the Surveys Department (in charge of connection was limited, especially in remote and maps and geodata) and students, local people vulnerable areas. Mobile data was also expensive, and other relevant government departments also limiting its use for collection. Further, capacity participated. building had to include computer basics and The main initiatives of the project include an some participants struggled to keep up with Integrated Flood Risk Management Plan (IFRMP) the training materials. However, countries with for the Shire Basin; an open data, Geonode limited telecommunication infrastructures, based platform called Malawi Spatial Data Portal such as Malawi, still have the potential to adapt (MASDAP); and a specific needs assessment for crowdsourcing/VGI for geospatial data acquisition Nsanje (the Nsanje 2012 Floods Post Disaster if actors support one another. The HOT team Needs Assessment). strongly believed that this project was an incentive for local people to start mapping. To reach its target, HOT used: Main lessons: 1. OSM to trace the collected in-field GPS tracks by adapting its JOSM and Tasking Manager • A well-organized plan including funding and functions; NGO expertise can result in a successful outcome; in this case funding from the GFDRR 2. Satellite imagery as a driver for field and technical support from HOT worked well. measurements; • Successful programs can build on accumulated 3. InaSAFE, an open tool which produces experience from other projects (see also cases realistic scenarios for flood preparedness; and 13 & 17) and to adapted to the local context. 4. MASDAP, Malawi’s Spatial Data Portal, which • Limited resources in developing countries may is managed by a core team of administrators affect participation and project progress. from various agencies and is a free and APPENDIX: CASE STUDIES 59 60 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 10. Food insecurity mapping, Bangladesh Interaction type Government → Public → Government Trigger event - Domain Mapping the Khulna District in Bangladesh as a part of an agricultural program. Organization Mapping for Resilience Actors USAID, Geo Center, Khulna University, US students, YouthMappers, local people Datasets High-resolution satellite imagery, GPS tracks and OSM. Process Tracing on high-resolution satellite images, collection of spatial and attribute data and importing all data in OSM. Feedback OSM indicating the transportation network and the water bodies in the vulnerable area of Khulna. Goal Map the unmapped Khulna district and improve land management and agricultural production. Side effects - Impact of the project Local Temporal pattern One-off Funding USAID Contact point - APPENDIX Food insecurity mapping, human constructions in every private yard and the irrigation ditches can be easily misclassified Bangladesh as roads. To overcome these challenges, custom There are various crowdsourced projects imagery from the Image Geo Center was used currently running in Bangladesh to support alongside Google Maps. Satellite imagery used different aspects of daily life. This project is was taken during the dry season. Also, local among those completed, funded by the US Agency people contributed their knowledge. for International Development (USAID) and This project ran under the Mapping for Resilience supported by YouthMappers, students of Khulda program established by Geo Center – a branch University and local people. The project is a part of USAID that offers technical assistance – of the “Feed the Future” agricultural program and supported students in the US and Khulna run by USAID and its main target was to map the University to jointly work and map Khulna water bodies, human-made water catchments district through “mapathons”. It also provided and the transportation network. USAID (2017) high-resolution satellite imagery for the creation is interested in Bangladesh because itg “is one of of the basemap, which may be shared in near-real- the most densely populated and climate change- time on the OSM platform. This workflow and the vulnerable countries in the world”. philosophy behind it have been adopted in various The project ran in Khulna district and its aim was projects with success. to map the area and improve land management Main lessons: and agricultural production. Mapping was focused on linear and area features such as rivers, ditches, • Developed countries such as the US may ponds, buildings and the road network. Four fund crowdsourcing projects through categories were used for each of pathway line their aid budgets to solve a problem such features and water features (both area and lines). as malnutrition. Training of students and local people was based • Climate may be a great obstacle in on workshops and extra tips were given on the satellite imagery and should be taken project homepage. Editing of tracks was done in into consideration when choosing data OSM and made available to the public. Once the collection periods. basemap was finished using satellite imagery, the • Government ethical policies may affect a students visited the area of interest to collect project’s public outreach. attribute and spatial information such as the locations of bodies of water. USAID aims to use • Local software companies may cultivate this project to better understand who has access attitudes towards open source software. to enough nutritious food to lead healthy lives. Mapping proved difficult due to the poor quality of existing satellite imagery, the complex waterway system in rural areas, the dense population in urban areas and the monsoon seasons that alter the terrain and therefore the satellite images. YouthMappers underline that the waterway system is further complicated by APPENDIX: CASE STUDIES 61 62 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 11. Syria Tracker Crisis Map Interaction type Public → Government Trigger event - Domain Humanitarian and disaster mapping. Organization Syria Tracker (ST), part of Humanitarian Tracker Actors US activists, Ushahidi, numerous human rights organizations and local people Datasets Google Maps in HealthMap platform. Process Hybrid model in near-time data collection and manipulation by a small team of volunteers. Feedback A continually updated list of eyewitness reports from Syria, often accompanied by media links. Aggregate reports including analysis and visualizations of deaths and atrocities in Syria. A stream of content-filtered media from news, social media (Twitter and Facebook) and official sources. Goal Preservation of the name and location of every victim to build memory of the disaster. Side effects - Impact of the project Local with potential governmental Temporal pattern Ongoing Funding - Contact point - APPENDIX Syria Tracker Crisis Map and/or photographic evidence. Moreover, the turn-around time for a report to be mapped on Syria Tracker Crisis Map was first started in Syria Tracker is between 1-3 days. The ST team 2011 by US volunteers located in the area when collaborates with various organizations to reach the civil uprising began. The project is still its target, such as Women under Siege and The active six years later according to the online Standby Task Force. reports and mainstream news feed, in order to document human rights violations and fatalities Among its main benefits, the project has shown as a result of the civil war. Due to its nature and continuity and progression. To meet its goal severity, the project has attracted the attention of and record everything it uses multiple channels. volunteers who have submitted over 6500 reports To avoid fake news, crowdsourced reporting is anonymously between April 2011 and May 2017. cross checked. It is clear that what was at first a temporary need has been transformed into an The approach is based on a hybrid model where ongoing process and the project is designed to data is collected either through automated or meet this challenge. crowdsourced methods. The automated data is collected from English online media resources However, it is still unclear how this record will be such as online news websites and blogs. As the used in practice, and this is the main challenge for Syria Tracker (ST) team underlines, their data the project. The instigators of the project aimed mining platform, “draws from a broad range of for the data to be used by future governments sources to reduce reporting biases”. (Irevolution, to deliver justice for victims of the crimes 2012). The crowdsourced technique is based committed. Syria Tracker’s ultimate goal is not on participation by citizens who can submit simply to record plain numbers but preserve the eyewitness reports and share information via name and location of every victim so that these a webform, email, Twitter, Facebook, YouTube crimes will not be forgotten. The material from and voicemail. To protect the volunteers and the Libya crisis mapping project in 2011 was only encourage participation, detailed security recently received for further investigation by the precaution information is given on an instruction International Criminal Court (ICC) and Amnesty page alongside a practical guide. International (AI). Syria Tracker, a project of Humanitarian Tracker, Main lessons: is powered by Ushahidi’s cloud-based platform, • A hybrid model where data is collected using Crowdmap. To cope with data mining, the both automated and crowdsourced methods HealthMap platform has been modified to serve work well. this project’s needs. HealthMap platform uses Google Maps as basemap. Data manipulation • A disaster response project may be an is undertaken by volunteers who receive the ongoing effort and not a one-off event when crowdsourced reports and then translate, circumstances demand. georeference, code and verify them to make • Crowdsourced maps in conflict areas are them public. The reports are also filtered and necessary to ensure anonymity for users. duplications resulting form the hybrid approach are removed. According to Irevolutions (2012), • Social media offers near-real-time information ST and their volunteers have been able to verify for data collection so that the map can always almost 90 percent of the documented deaths be up to date. mapped on their platform thanks to video APPENDIX: CASE STUDIES 63 64 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 12. iCitizen, South Africa Interaction type Public → Government Trigger event - Domain Generic mapping with focus on local infrastructure issues. Organization University of Witwatersrand, LINK Centre Actors LINK Centre Datasets Multiple datasets per service delivery issue to be tracked. Process Collection of data points via mobile phones. Adoption of different ways of geotagging photographs in real time or via SMS or/and email. Feedback Generic and purpose-built maps for disaster preparedness. Goal Reporting and solving fundamental problems with basic infrastructure and services. Side effects - Impact of the project Local Temporal pattern One-off Funding - Contact point Dr Kiru Pillay, University of the Witwatersrand APPENDIX iCitizen, South Africa The designers of the application aimed to extract boundary data for South African provinces, districts In recent years, South Africa has seen a surge in and local municipalities and expose these on the political protest against slow service delivery. online map using GeoJSON data. This would While the United Nations Human Development automate the process of calculating the jurisdiction Index considers South Africa to be a middle- of any reported incidents. A live reporting engine income country, there is a large disparity in income and online social tool would also allow for two- distribution across the population. Social unrest is way communication between the web server an obvious consequence of poverty, high levels of (and its user base) and local municipalities and unemployment and service delivery backlogs. civil societies. In the context of these issues a new initiative was The main difficulty related to the acceptance of launched in 2014 to improve the daily life of citizens the project. One university found validity in the by collecting crowdsourced reports of service concept but was unable to commit resources for issues and passing them to the relevant authorities the development of the application. Going forward, for resolution. The iCitizen project aimed to give two difficulties were envisaged. The first was members of the public the ability to report on acceptance of the validity of the generated datasets fundamental problems with basic infrastructure and by local municipalities. The second was acceptance services. The researchers involved in the project of the use of mobile phone and applications by intended to contact local municipalities to discuss the public as an effective tool for voicing service the extent to which this project could be embedded delivery concerns. Even though the penetration rate in current initiatives around citizen monitoring of mobile phones is fast approaching 100 percent and evaluation. The main aim of the project was to of the adult population in South Africa, the use of give citizens an active voice. A secondary research mobile applications and GIS-mapping tools of this objective was to understand and identify the nature is largely untested. role of mobile phones in citizen-led monitoring and evaluation. The project has not been updated since it was documented in 2014. Members of the public would be able to report issues by forwarding geotagged photographs, Main lessons: sending in locations via SMS or reporting issues • Projects can be used for a variety of tasks at local via email. The first iteration of iCitizen was built level, not just that for which they were designed. on the Drupal open source content management system (CMS). As an enterprise CMS, it provided a • Using a range of software, programing languages lot of services out of the box, including membership and platforms can broaden a project’s horizons. management, image upload, taxonomy (category) • VGI applications face financial issues due to their management, user commenting, thorough user technological nature and the resources of the permissions, field APIs (application programing organizations involved. interfaces), views templating and reporting and HTML5 theming capability. Researchers were able • Concerns from agencies about the quality to extend the core functionality to include mapping of generated datasets and improving public enabled through geolocation, leaflet maps (using adoption of mobile applications are common OpenStreetMap as the Map Tile server) and a voting challenges. API allowing users to verify incidents. APPENDIX: CASE STUDIES 65 66 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 13. Community Mapping for Exposure in Indonesia Interaction Type Government → Public → Government Trigger event - Domain A priori disaster response. Organization Community Mapping for Exposure in Indonesia Actors Indonesian Disaster Management Agency (BNPB), Disaster Management Innovation (DMInnovation or DMI), Humanitarian OpenStreetMap Team (HOT), The Australian Community Development and Civil Society Strengthening Scheme (ACCESS), GFDRR, Pacific Disaster Center (PDC) Datasets Satellite imagery, GPS tracks and attribute data. Process Collecting Spatial and attribute data and tracing in OSM platform. Feedback • OpenStreetMap layer showing an up-to-date basemap of vulnerable areas. • Thematic maps showing damage in case of various physical disasters. • A standardised training curriculum for capacity building purposes. Goal Reduction of vulnerability to natural disasters. Side effects Deemed a successful example of disaster relief preparedness that could be applied in other developing countries. Impact of the project National & with affect to governmental body Temporal pattern Ongoing Funding AusAID & BNPB Contact point Kate Chapman, HOT, (Phase I and II) Yantisa Akhadi, HOT, (Phase III and IV) APPENDIX Community Mapping for The main aspect of concern is the quality of the results, which showed great variation. Other Exposure in Indonesia minor deficiencies were also noticed, such as a An example of an a priori disaster response, failure to maintain constant mapping volunteers this Indonesian mapping project began in early and the use of time-consuming technical 2011 and at the time of writing is still active. The methods in a few cases (e.g. Excel spreadsheets main idea behind the project was to use OSM in data collection or manual methods of data to collect previously unavailable data about manipulation). buildings and their structure in both urban and In the second phase, which focused on scenario agricultural environments and to use appropriate development for contingency planning (August models to calculate likely damage in case of 2012 – December 2013), the main aim was to use physical disaster. The combination of the impact OSM data to fill gaps and QGIS and InaSAFE models and the use of realistic data led to the for spatial analysis under a curriculum program development of open source risk modeling which comprise a series of training programs: software (InaSAFE), showing the affected people, beginner, intermediate and training of trainers infrastructure, and damage if disaster were to hit (ToT). a specific area. Based on recommendations from previous phases, The pilot phase (early 2011-2012) consisted of HOT (in coordination with AIFDR) proposed workshops offering training on the project and strategies to further expand the use of InaSAFE building construction as well as data collection and OSM in Indonesia. In this phase, the focus in urban and rural areas. The approach between turned to supporting the Expansion of InaSAFE rural and urban areas was slightly different, and OpenStreetMap in Indonesia (December 2013 although the result was similar. The original data – August 2016). The proposed activities focused was derived from paper maps, which were edited on harmonizing training materials with the needs by local people; satellite imagery, depending on of disaster management agencies while keeping availability; and GPS tracks. Data was edited using them up to date with the software version. HOT JOSM and Potlach2 web editor and then used in continues to support local disaster management QGIS. Urban areas were mapped by students who agencies in multiple provinces in the use of OSM, took part in a mapping competition. Rural areas QGIS and InaSAFE for contingency planning were mapped with ACCESS contributors and local development. Additional training materials on people. data validation have been developed to improve In terms of technical support, the project was OSM data quality in Indonesia. not only supported by HOT and OSM but also This phase introduced and promoted the use by open source software such as QGIS. The main of mobile data collection using smartphones to innovation in data collection was the private reduce the use of pen and paper during surveys. datastore, which offered a unique ID for each GeoDataCollect, the in-house mobile application, object. The final output has also been a success has been installed in more than 5,000 devices and in enabling local government to visualize where used by Jakarta Disaster Management agency as people are most in danger by combining local their official app for disaster reporting. wisdom with scientific knowledge to produce realistic scenarios for numerous different In phase IV, the main focus is on transferring physical disasters. APPENDIX: CASE STUDIES 67 68 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT knowledge from HOT to BNPB. Here the core Main lessons: aim is to foster disaster management innovation through open geospatial data (September 2016 • The collaboration of multiple stakeholders in - April 2018). This would enable BNPB to have disaster management is important. its own capacity to map disaster prone areas, • Capacity development is essential to be able to train local partners and build collaboration with take full advantage of technical innovations in local universities. This set of skills is important disaster management. as BNPB has already listed 136 priority districts/ • Universities can be strategic partners to cities of high risk of disaster, and they would collaborate with local disaster management like to map these areas before disaster strikes. agencies as well as developing knowledge and HOT is also working closely with DMI to develop skills on mapping and data collection guidelines on how to use OSM, QGIS and InaSAFE for all phases of the disaster lifecycle, • OSM can be used to support disaster such as for risk assessment and situational preparedness through damage impact awareness. This would be relevant to the goal of projection to physical infrastructure. institutionalizing InaSAFE in the BNPB. A second focus of the current phase is to build and maintain links between people and institutions in disaster management. The strategy is to leverage existing partnership with universities and bridge the connection with BNPB. HOT also expected to start building links with underrepresented groups, such as people with disabilities. These projects show how sustained engagement and funding can bring a crowdsourcing effort into maturity, with careful piloting and testing. However, it should be noted that the project requires the ongoing engagement of external actors. APPENDIX 14. Haiti disaster response Interaction type Public → Government → Public Trigger event A natural disaster (earthquake) and humanitarian crisis. Domain Generic mapping (topographic maps of the area) and purpose-built maps (disaster relief management). Organization Humanitarian OpenStreetMap Team (HOT) Actors United Nations, NGOs, National Haitian Mapping Agency, National Center of Spatial Information, Haitian civil society Datasets Historic maps, CIA maps, high-resolution imagery from Yahoo, paper maps and GPS tracks. Process Tracing in OpenStreetMap (OSM) platform from different data sources and collecting GPS tracks. Feedback Topographic and purpose-built maps for the management of supplies of medicine and food, and location of settlements. Goal Facilitate disaster response management. Side effects The datasets created have not been used by the NMA but by international aid organizations (UN, USAID). Impact of the project National & with affect to governmental body Temporal pattern One-off Funding NGOs Contact point Mikel Maron APPENDIX: CASE STUDIES 69 70 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Haiti disaster response One legacy of the project was the subsequent establishment of an OSM community in Haiti Haiti was dramatically affected when a 7.0 COSMHA (Communauté OpenStreetMap magnitude earthquake hit the capital city of Port- Haiti), which remained active for several years au-Prince on 12 January 2010. Death toll estimates following the earthquake and conducted mapping range from 100,000 to 200,000 people. More activities throughout the country. The widespread than 250,000 residents were injured and more availability of OSM information in the country than 30,000 buildings collapsed or were severely was useful during the response to Hurricane damaged. When the magnitude of the disaster Matthew (October 2016). became clear, the main issue for those responding to the disaster was that the only available spatial The disaster response’s success lies in four data was poor and last updated in 1960s. The main factors: time, cost, high participation of local mapping agency collapsed in the earthquake, volunteers and official trust. The sensitization with the loss of most of the skilled employees. of the public to the Haitian crisis led to high An updated map was urgently needed for the participation of volunteers and immediate distribution of supplies and identification of mobilization worldwide. The contribution collapsed buildings, damaged infrastructure and of NGOs and other official partners and the medical stations. release of conventional datasets as reference maps for tracing without license restrictions The Haiti disaster response is a successful were vital to success. Government support was example of geographic information being made inevitable due to the critical circumstances and open by official partners, enhanced by public limited resources. volunteers and returned to government for action (although the government was reluctant about Although, the project is characterized as the involvement of volunteers). The first imagery successful, there were a few weaknesses. All the was loaded on the OSM platform within 48 hours. responding organizations lacked experience Sixty people were trained and more that 700 and awareness of the operational norms in contributed to the mapping, among them people humanitarian response, which constitute an from UN agencies, NGOs, National Haitian operational framework for the accountability of Mapping Agency, National Center of Spatial different sources. Deficiency of coordination led Information (CNIGS) and Haitian civil society. to the duplication of data. The national mapping Historic maps, CIA maps and high-resolution agency, CNIGS, also never really engaged imagery from Yahoo were first used for tracing with OSM, although official data was given in OSM to improve the basic maps. Volunteers supportively at the beginning of the project. with paper maps and GPS completed the second phase of tracing. As a result, OSM was used as a default basemap for responding organizations and the Haitian government. The effort also inspired efforts to crowdsource the analysis of satellite and aerial imagery by the World Bank, a project in which more than 30,000 damaged infrastructures were identified and classified. APPENDIX Main lessons: • This was the first large-scale crowdsourced mapping exercise for humanitarian reasons and shows its successful use in responding to disaster. • An integrated methodology of this kind follows four steps: spatial data contributed by official providers, supplemented with GPS tracks, integrated into OSM and updated by a large number of volunteers worldwide. • Time, cost and official trust of data by NGOs and other official partners are key to success. • Lack of coordination and experience between different actors can lead to duplication of data and waste of resources. • Differentiation between conventional and government data in terms of engagement to the project did not prevent success. APPENDIX: CASE STUDIES 71 72 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 15. Mapping of South Sudan Interaction type Public → Government → Public Trigger event On 9 July 2011 South Sudan became Africa’s 54th nation after its official independence. Although it is the newest nation, it is poorly mapped. Domain Generic mapping of a poorly mapped area and thematic maps of essential features like roads, hospitals and schools. Organizations Google, NGOs along with the World Bank, United Nations Institute for Training and Research (UNITAR) Operational Satellite Applications Programme (UNOSAT) and Regional Center for Mapping of Resources for Development (RCMRD) Actors The Sudanese diaspora, Google, the World Bank, UNOSAT and RCMRD Datasets Updated satellite imagery covering 125,000 km2 (40 percent of the UN’s priority areas) uploaded to Google Earth and Maps. Process Workshops and editing on Google Maps via Google Map Maker. Feedback Generic and thematic maps covering important points of interest such as schools, hospitals and roads. Goal To engage and train the Sudanese diaspora and other volunteers worldwide to participate in Google Map Maker. Side effects Mapping the poorly mapped South Sudan so that the infrastructure and economy of the country could be developed and humanitarian aid provided to the local population. Impact of the project National Temporal pattern One-off Funding World Bank Contact point - APPENDIX Mapping of South Sudan The activity was constructed as a one-off event and while the maps are now available on Google After years of political instability, South Sudan Maps, the demise of Google Map Maker in early became a new nation on 9 July 2011 after its 2017 means that only the Google Local Guides official independence. Although South Sudan program is open to data contribution, though the is expansive and the newest nation, it is poorly data is open for download by NGOs and research mapped. Maps are particularly important for bodies on the Google website. the development of the infrastructure and economy of the country and the distribution of Among the main factors in the project’s success humanitarian aid. are not only the enthusiasm and inspiration of the Sudanese diaspora, which encouraged Google, with the aid of World Bank, UNOSAT them to convey their experience and knowledge and RCMRD, recognized this need and started to other people, but also the interest the local the project for the creation of better maps of government showed in the project. The project’s South Sudan by supporting communities to map approval by local government and its impact in schools, hospitals, roads and more with Google decision making policies is noticeable. Another Map Maker. The project was launched with a innovation of the project is the principal role and series of events to disseminate the purpose of the contribution played by the Sudanese diaspora. mapping and inspire and train participants. The Through VGI projects, local knowledge can be first event was in April 2011 at the World Bank shared worldwide and from different parts of the headquarters in Washington, DC, with a satellite world, not only from the area of interest. At the event in Nairobi at the same time. The next event same time, among the main weaknesses of the was in September 2011, held by the South Sudan mapping is that local people were not actively National Bureau of Statistics in Juba. involved. The project lacked research in the field, and did not use GPS or open source software, To aid their work, updated satellite imagery of although Google’s involvement guaranteed high the region, covering 125,000 km2 (40 percent of participation levels. the UN’s priority areas), was uploaded to Google Earth and Maps. In the last event volunteers Main lessons: worked together and made hundreds of edits in less than four hours. The process is simple: • Crowdsourcing projects can be coordinated citizens edit using available web tools and and implemented from a distance. their local knowledge and, after approval, edits • Participation of volunteers and transmission become visible to all Google users worldwide. of motivation to others are key factors in The mapping was used by the Satellite Sentinel successful crowdsourcing applications. Project, a collaborative project focused on human rights violations and human security concerns • Inspiration for other projects and improved in Sudan and involving Google, the Enough applications can be beneficial to areas Project, Not On Our Watch, UNITAR, UNOSAT, of interest. DigitalGlobe, the Harvard Humanitarian Initiative • Acceptance by local government as an and Trellon. opportunity for decision making policies and humanitarian aid can escalate the impacts of a VGI project. APPENDIX: CASE STUDIES 73 74 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 16. Crowdsourcing The National Map, National Map Corps, US Interaction type Public → Government → Public Trigger event The National Map Corps project became available nationwide. Domain Generic mapping (structures data in the public domain is used to update The National Map geospatial databases and US Topo maps). Organization US Geological Survey (USGS), National Geospatial Program Actors Public, local, state and federal agencies including the USGS Datasets USDA National Aerial Imagery Program (NAIP) imagery, National Map base layers, ESRI world imagery, Alaska community photos, the national structures database and US Topo maps. Process Using crowdsourcing techniques, the USGS National Map Corps VGI project known as “The National Map Corps (TNMCorps)” encourages citizen volunteers to collect and edit data about human-made structures to provide accurate and authoritative spatial map data for the USGS National Geospatial Program’s (http://www.usgs. gov/ngpo/) web-based The National Map. Feedback Updated structures are contributed in real time. Databases are updated in near real time. Goal Maximize limited resources while continuing to support the National Geospatial Program by leveraging volunteers with local knowledge to update The National Map and USTopo maps. Data collected is in the public domain and freely downloadable. Side effects - Impact of the project National Temporal pattern Ongoing Funding US federal government Contact point Erin Korris, Elizabeth McCartney (nationalmapcorps@usgs.gov) APPENDIX Crowdsourcing The National Map, improve baseline structures data; that further review by advanced volunteers willing to provide National Map Corps, US peer review improves the data further; and that VGI is not new to the USGS, but past efforts sample-based inspection by USGS personnel can were hampered by available technologies. Over monitor these processes. the last two decades, the USGS has sponsored TNMCorps volunteers continue to support various forms of volunteer map data collection the National Geospatial Technical Operations projects, including the Earth Science Corps where Center (NGTOC) mission of the “acquisition and volunteers annotated topographic paper maps, management of trusted geospatial data, services, the collection of GPS points using handheld GPS and map products for the Nation”. After the devices and, finally, web-based technology to TNMCorps project fully expanded to all 50 states input data in 2006. Despite these efforts, and as late in 2013, the project continued to grow. In valuable as the updates were, technology could 2014, a national data quality study was conducted not keep pace with decreasing USGS resources which showed that volunteers were continuing and changing priorities, and the VGI effort was to provide high quality data even as the project suspended in 2008. In 2010 – 2012, the perfect expanded from a volunteer corps consisting storm of improved technology, social media and primarily of GIS students and professionals to continued decreasing resources once again made the general public. More than 230,000 points had crowdsourcing an attractive option. been edited by the end of 2016. After several pilot projects to determine the Successful crowdsourcing is not without viability of bringing back the volunteer mapping challenges, some of which include volunteer program, The National Map Corps volunteers are recruitment, volunteer engagement and successfully editing 10 different structure types participant motivation. The National Map in all 50 states, including schools, hospitals, post Corps endeavors to meet these challenges offices, police stations and other important public using gamification techniques and a mixture buildings. Using National Agricultural Imagery of traditional and social media. Gamification Program (NAIP) imagery as the primary base includes easy on-ramping, virtual recognition layer, volunteers collect and improve data by badges, friending, map challenges, social media adding new features, removing obsolete points interaction and a tiered editing approach. Using and correcting existing data. Edits are contributed these techniques has been successful. The through a web-based mapping platform and are National Map Corps continues to see substantial incorporated into The National Map geospatial increases in the number of volunteers and databases and ultimately US Topo maps. volunteer contributions to The National Map. Data quality is a common concern regarding VGI. Other challenges continue to exist and include: In order to address those concerns, an analysis organizational resistance to accepting data from of a pilot project over the state of Colorado was volunteers as being “good enough” to populate conducted in 2012, and a national data quality national databases; and working through issues study was conducted in 2014. For all structure for which there is no well-established policy feature types, volunteer involvement was regarding government accepting data from found to improve positional accuracy, attribute citizens. One example is the requirement to accuracy and reduce errors of omission. The obtain approval for conducting a “survey” from studies demonstrated that volunteer edits APPENDIX: CASE STUDIES 75 76 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT the Office of Management and Budget as part of the Paperwork Reduction Act even though the project is not really conducting a “survey”. Main lessons: • Adoption of challenging techniques such as gamification has been successful and attracts volunteer interest. • Evaluation of the quality indicated that participation improves accuracy and reduces errors. • Organizational resistance to accepting data from volunteers is one of the major challenges for VGI projects of this kind. • Key factors to successful crowdsourcing include building on experience, leveraging existing technology and having the support of key individuals in the organization. APPENDIX 17. Crowdsourced flood resilience in Jakarta, Indonesia Interaction type Public → Government → Public Trigger event Disaster management agency of Jakarta wanted to have better basemap and data for flood planning and reporting. Domain Mapping for disaster management. Organization Jakarta Disaster Management Agency (BPBD DKI Jakarta) Actors Indonesian Disaster Management Agency (BNPB), Jakarta Disaster Management Agency (BPBD DKI Jakarta), Department of Foreign Affairs and Trade (DFAT) Australia, United Nations Office for Coordination of Humanitarian Affairs (UNOCHA,) University of Indonesia (UI), Humanitarian OpenStreetMap Team (HOT), GFDRR, MIT Urban Risk Lab, USAID, Pacific Disaster Center (PDC) Datasets OpenStreetMap (OSM) data of sub-village (Rukun Warga) boundaries, religious, health, sports and government facilities, schools, roads, flood locations. Process Collecting Workshops, field survey using GPS and mobile devices, satellite tracing, crowd-reporting using social media. Feedback Urban village leaders received paper poster maps of their villages. Public able to send and receive flood reports and alerts in near real time to help them make time-critical decisions about their safety during flood emergencies. Goal Improve geographic information available for flood management Improve geographic data sharing across formal and informal sectors. Side effects Crowdsourced flood resilience in Jakarta has made others interested in the idea of crowdsourcing and using community mapping to collect base data and record data at a relevant scale. Impact of the project National & with affect to governmental body Temporal pattern Ongoing Funding NGOs Contact point Kate Chapman, Emir Hartato, Yantisa Akhadi, & Etienne Turpin APPENDIX: CASE STUDIES 77 78 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Crowdsourced flood resilience in Jakarta’s citizens to report the locations of flood events using the social media network Twitter. Jakarta, Indonesia The pilot study contributed to a public web-based Jakarta, Indonesia, is a megacity that has frequent real-time map of flood conditions powered by seasonal flooding issues. In 2012, Jakarta’s CogniCity Open Source Software. It produces disaster management agency (BPBD DKI megacity-scale visualisation of disasters using Jakarta) needed better data to prepare for the OSM basemap, crowd-sourced reporting, and flood season. DFAT-Australia, UNOCHA, HOT, government agency validations in real time. The GFDRR and University of Indonesia assisted in project demonstrated the value of social media the process. in disaster management as an operational tool to provide decision support in the event of disaster. The original aim was to ask the heads of the 267 urban villages the location of their critical Since its debut in 2013, the PetaJakarta.org infrastructure, then ask university students to platform has grown into a single robust platform help with technical mapping. Impact analysis that integrates local knowledge from various using InaSAFE open source impact modeling crowdsourcing tools (mainly social media and software was then performed as part of a instant messaging) and formal knowledge from contingency planning process. The data has government agencies. The project expanded to been used to create government maps to report PetaBencana.id with support from Massachusetts flood conditions and village heads have used Institute of Technology (MIT) Urban Risk Lab, poster maps to plan logistics when responding USAID, and Pacific Disaster Center (PDC) by to flooding. the end of 2016. Since then, the PetaBencana. id platform has been used by millions of Jakarta The project also created an open dataset that can resident users to make time-critical decision be used for a variety of analyses at the village, about safety and navigation during emergency district and provincial levels. Using an open flood events; it has also been adopted by the platform means that anyone can use the data National Emergency Management Agency (BNPB) and it can be updated easily as the information to monitor flood events, improve response times, becomes outdated. The data collected is useful and share time-critical emergency information for flood management, allowing the government with residents. of Jakarta to show more detailed maps than previously available and increasing demand for additional mapping at a higher resolution. One negative aspect of the methodology used is that while it did collect the data very quickly, urban village officials do not have an easy way to update their area as the data changes. In 2013, Jakarta’s disaster management agency (BPBD DKI Jakarta), together with SMART Infrastructure Facility (University of Wollongong) and Twitter Inc. conducted a pilot study to develop PetaJakarta.org platform, enabling APPENDIX Main lessons: • Collaboration between different stakeholders, depending on their knowledge and capacity, means stakeholders can contribute to specific tasks and stages of crowdsourced disaster resilience. • Open data can be used at different levels of decision making policies such as village, district and provincial levels. • Difficulties in keeping data up to date is one of the most important viability concerns. • Data can be used in a variety of ways, including by government agencies for the creation of maps. • Sharing information and coordinating data between residents and government agencies fosters equitable and innovative practices of crowdsourcing for disaster resilience. * This case study was in the 2014 report under the title “Flood preparedness through OpenStreetMap, Jakarta, Indonesia”. The title has been changed as more agencies are currently involved in the process. APPENDIX: CASE STUDIES 79 80 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 18. Twitter use in Italian municipalities Interaction type Government → Public → Government Trigger event - Domain Social Media Organization Ladest Lab, University of Siena (Italy) Actors University (UNISI); Twitter Italia; Anci (Associazione Comuni Italiani) Datasets New dataset creation including Twitter profile names and metrics (Tweets, followers, following, activation date); integration with other datasets. Process Data was collected manually for each of the 7,981 (ISTAT, 2017) Italian municipalities, then integrated with official census population data (ISTAT) by municipality code and with Large Urban Zones (Eurostat) to develop Twitter profile performance analysis. Feedback Official presentation of the results in public meetings; international publications. Goal Analysis of municipalities’ performance on the social platform to foster participation and efficiency. Side effects More municipalities activated Twitter profiles. Creation of an on official forum #pasocial (pa = public administration). Definition of dedicated professional skills to deal with social platforms in public administration. Impact of the project National Temporal pattern Ongoing Funding - Contact point Cristina Capineri & Antonello Romano APPENDIX Twitter use in thanks to their external economies based on social proximity or on a stronger sense of place Italian municipalities compared to high metropolitan fragmentation. The University of Siena in collaboration with The first and most active (in terms of Tweets Twitter-Italia and Anci (Associazione Nazionale sent and followers) municipalities on Twitter Comuni Italiani) started a research project are those which started with the activation of in 2013 to investigate the adoption of Twitter “civic networks” (municipality websites) in the profiles by Italian municipalities. After the first late 1990s, showing the relationship between survey, other updates took place in 2015 and the adoption of these kinds of technologies. It is early 2016 to analyze the spatial and temporal interesting to observe the case of 11 neighbouring diffusion process. municipalities sharing the same profile and thus creating a sort of “tweeting district” which The research identified that, at the time of the benefits from the common management of first survey (November 2013), only 461 of the social communication. 7,981 Italian municipalities had Twitter profiles, approximately 6 percent; three years later (early The survey analyzed the messages sent, since 2016) the number of profiles had reached 881 activity profiles must be assessed not only in municipalities (approx. 11 percent) showing relation to the amount of Tweets and followers an increase in institutions joining the social but also with respect to the quality and type of platform. The profiles’ activity is also increasing information sent. The latter includes simple as demonstrated by Twitter metrics: 49 percent messages for informational purposes up to increase in Tweets sent in 2015 compared to more complex messages addressing planning the previous year; as well as 73 percent more and territorial management. The research team followers and a 20 percent increase in following. classified the hashtags used by most municipal profiles into several categories and found that The geography of municipal Twitter profiles in the most widely represented information related Italy seems to reflect the urban structure of the to culture and tourism, followed by geographical country, which is mostly made up of many small information, utilities and weather forecasts. More and medium sized cities. The research highlighted recently several profiles have introduced news that 5 percent of the profiles have been activated about open data. Messages about governance are by large towns with more than 100,000 still limited, demonstrating that the potential inhabitants (note that large municipalities of Twitter as a collator and distributor of represent only 1 percent of the total); 40 percent information on complex issues around which by small to medium municipalities; while 55 to initiate debates and discussions has not been percent had been activated by municipalities realized. Only a few municipalities have used with less than 10,000 inhabitants. Furthermore, Twitter for emergency and risk management. while the larger municipalities attract the most Nevertheless, it is worth noting that the news followers (66 percent), the small and medium about L’Aquila’s severe earthquake of 2009 was municipalities are the most active (77 percent first announced through Twitter before other of the Tweets). This demonstrates that reduced media. Institutions still face the challenge of population size is not a barrier to the spread collecting contributions produced by their of social applications but may in fact be an community and filtering and validating them advantage or a driver; small municipalities to improve governance. These tools therefore also seem to have built a tweeting atmosphere APPENDIX: CASE STUDIES 81 82 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT provide an important opportunity to establish and consolidate good governance based on efficiency, transparency, simplification and on the development of collaborative solutions for critical issues. Main lessons: • Social media use is growing in many contexts and represents an opportunity for collecting and analysing VGI. • Demographics may be reflected in social media use but it can still be an efficient way to access information about less populated areas. • Social media may be used for some types of communication more than others but has an application in disaster scenarios. APPENDIX 19. National Biodiversity Data Centre Ireland Interaction type Public → Government Trigger event NBDC strategy to encourage the submission of biodiversity records by the general public. Domain Update and submission of National Biodiversity Database (NBD). Organization National Biodiversity Data Centre (NBDC) Ireland Actors NDBC and the Irish biodiversity community (especially university researchers) Datasets Existing NBDC database. Process Users enter their observations through the appropriate HTML forms on the NBDC website. Observational data is checked internally in NBDC and then made available for access and visualisation on the online maps and charts. Data submitted in any other file format is converted to the appropriate format to be added to the database. A free mobile phone app allows for real-time, in the field, data capture. This feeds observations directly to the online system. Feedback Any data submitted is mapped and viewable immediately. Contributors can visualise their own data using their own email address. Once checked and validated, the validated data are added to the NBD. Goal To leverage the potential of outreach groups for data survey and observation thus widening the base from which observational data may be obtained. Provide a suite of thematic data capture systems partner organizations can use for their own data capture needs, which are dynamically linked to the core system. Side effects - Impact of the project National & with affect to governmental body Temporal pattern Ongoing Funding NBDC Contact point Dr. Liam Lysaght, National Biodiversity Data Centre, Carriganore, WIT West Campus, Waterford, Ireland APPENDIX: CASE STUDIES 83 84 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT National Biodiversity Data at NBDC. Data is submitted through the web application, mobile phone app or by email Centre Ireland (for large and possibly incorrectly formatted The Irish National Biodiversity Data Centre observations). Data is quickly checked and made (NBDC) initiated this project to leverage the available for access and visualisation on an online potential of outreach groups and the general map. Those who submit their data to the system public for data survey and observation. This will be able to access this data in the future. A new widens the base from which observational online capture system will be released before mid- biodiversity data may be obtained in Ireland. May, 2017 which will provide far more visibility of The NDBC also launched this project to initiate all recording efforts, and will include leagues of stronger engagement with non-professional records by region, by taxonomic group and even scientists and heighten the understanding of by recorder. biodiversity related matters among the general The NBDC has also built a species browser, which population in Ireland. There is a good tradition of brings together information on the ecology of the observational work being performed by volunteer species with maps of empirical data feeds from community groups in Ireland. the NBDC’s own data capture system and that of Since June 2012 there have been over 213,000 GBIF and links to the Encyclopedia of Life. citizen science records submitted, validated and The NBDC ensures that contributors’ efforts stored in the NDBC databases. Approximately are well recognized and advertised online. 5,000 records are submitted per month. These Frequent workshops also report on the progress are complemented by data provided from other of the project, the uses the data is being put sources, such as NGOs, researchers and learned to, and how the project can be sustained and societies. Currently the entire database has over improved. Events such as Bioblitz (held annually) 4 million records. NDBC also makes this data involve the general public in biodiversity data available to the Global Biodiversity Information collection events which have an aspect of fun Facility (GBIF), a network of 90 participants and competition. worldwide working on an open biodiversity data infrastructure funded by governments. It allows Main lessons: anyone, anywhere, to access data about all types of life on Earth, shared across national boundaries • Feedback for contributors is important and via the internet. The NBDC is Ireland’s GBIF can be established in many forms. node and contributes Irish data to the more • Providing several pathways for users to than 700 million biodiversity records mobilised contribute their biodiversity observations through the GBIF data portal. improves accessibility. For the submission of observational data there • Allowing contributors to provide as much (or are species and site forms, with the latter as little) information as they have available allowing a more detailed recording of biodiversity means there can be great diversity in the information. The species forms include birds, resolution of data provided, requiring manual amphibians and reptiles, bumblebees, dragonflies, and automated controls. etc. The data from these forms and collections are extracted and merged with other datasets APPENDIX 20. Towns Conquer gamif ication and Instituto Geográf ico Nacional toponyms database, Spain Interaction type Public → Government Trigger event Funding opportunity from AGILE and EuroSDR project on Crowdsourcing in National Mapping 2013. Domain Validation of a national toponyms database. Organizations Universities (University of Nottingham UK and Universitat Jaume I of Castellón, Spain), ESRI Europe (sponsor), Instituto Geográfico Nacional (IGN) Spain Actors IGN, University of Nottingham, UK, and Universitat Jaume I of Castellón, Spain Datasets IGN national toponyms database of Spain. Process Users provide updates to the database while playing a game for rewards, contributions are checked by the national mapping agency before being incorporated into the national database. Feedback Users played the game to gain points. Points were maintained in a league table format and when a user gained enough points they became the mayor of that particular region. The more validations the user provided, the more points they gained. Goal Volunteer validation of a national toponyms database. In Spain it has taken over 10 years to implement a model to standardize the nomenclature of municipalities, yet today there are still conflicts with the names of some places, especially in regions with two languages. Side effects Validated gamification techniques for public update and management of important national databases. Impact of the project National Temporal pattern One-off Funding - Contact point Dr. Joaquín Huerta, Universidad Jaume I, Castellón, Spain APPENDIX: CASE STUDIES 85 86 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Towns Conquer gamification and two languages. Instituto Geográfico Nacional This project did highlight that gamification toponyms database, Spain techniques (when properly designed and thought out) could provide a very good platform The Towns Conquer application was developed from which bodies like IGN could involve the when the research team were awarded a small public in updating and managing important funding prize from a joint funding venture national databases. The game also provided the between AGILE and EuroSDR. This funding mechanism to motivate users to participate. allowed the support of one person to work in the university, in collaboration with a nominated The project was stopped in 2014 and was not national mapping agency (NMA). The aim of this subsequently rolled out on a larger scale. There project was the validation of a toponyms database were a number of issues which prevented this provided by the Spanish Instituto Geográfico roll out. These were mainly centered around Nacional, with 136,454 entities requiring organizational practices, which made it difficult validation. The public is involved through mobile to deploy the project on a national scale, and and web-based gamification techniques, which funding and resource allocation for both the are used to persuade users to contribute their future development of the Android software and amendments to the given toponyms database. support of the gamification approach into the The goal is for citizens to amend well-known future. placenames while playing a game and using their Main lessons: knowledge of their local area or other areas/ regions in Spain. • Funding can drive innovative VGI projects and instigate new modes of engagement. An Android-based game was developed which was linked to a web map service at IGN. This • Links between government agencies and allowed IGN Spain to deliver the toponyms research centers can generate the resources database on suitable basemaps for the mobile required for VGI projects. application. Users signed up for free. There were • Gamification can be a successful way of some prizes at the end of the game (with a time engaging the public in VGI projects. limit on the number of months) for the citizens who had gained the most points (conquered the • Data quality issues can be addressed most placenames) during this time. Placenames through checking processes done by the submitted to IGN via the gamification software official organization. were checked by an IGN official before being submitted and updated in the database. This funding allowed the NMA to test out crowdsourcing and gamification as a means of updating a national database. In Spain it has taken over 10 years to implement a model to standardize the nomenclature of municipalities, yet today there are still conflicts with the names of some places, especially in regions with APPENDIX 21. US National Park Service – Places Project Interaction type Public → Government → Public Trigger Event Licensing and data validation concerns. Domain Tourism, natural features Organization National Park Service Actors National Park Service Datasets OpenStreetMap Process The edits are currently only done by Park Service employees, so there is no validation in effect. This may change in the future. Feedback Internal Park datasets digitized from park maps. Goal To create an up-to-date map of all the parks for viewing and use by the public on the parks’ websites. Side effects None yet, but data quality issues are anticipated. Impact of the project - Temporal pattern - Funding - Contact Point Jim McAndrew APPENDIX: CASE STUDIES 87 88 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT US National Park Service – there are plans to visit the parks and train users. Some parks with GIS departments have already Places Project started synchronizing their own information and The US National Park Service (NPS) does not have been pushing to use the system for all of have a comprehensive dataset of geographic their webpages. information describing tourist infrastructure and The biggest challenge is that OSM does not natural features for all of the 400 or so individual handle traditional GIS well. Tools are therefore territories managed by the NPS. Some parks being built to synchronize existing ESRI databases have their own GIS departments and produce with the new database using the ESRI REST API. excellent data, while other parks are small and The aim is to make contributing to the system as do not have these resources. The aim of the NPS easy as possible for the existing GIS departments, Places project is to allow non-technical users to which do the bulk of GIS work in the parks. While add and modify important landmarks in the parks the goal is to get non-technical people involved, in a single map. This map could then be used as a existing GIS departments remain the core of basemap for all National Park Service web maps. the project. The NPS Places project (https://www.nps.gov/ Main lessons: maps/tools/places/) uses the OpenStreetMap (OSM) platform because of the robust open • It is possible and sometimes desirable to source tools available. This includes the backend use the OSM toolset to facilitate voluntary API and rendering formats as well as the easy-to- mapping activities without using the use iD editor. It does not use the OSM database dataset itself. due to licensing restrictions on OSM data. The • The OSM platform can be empower non- NPS maintains its data separately. technical users to start modifying maps. The project is designed to collect point data The iD editor is flexible and easy to use. The from NPS employees that will be displayed on rendering tools, such as TileMill and Mapnik, most of the maps on the NPS website. There are allow the data to be updated in real time. plans to expand this project to allow edits from • The success of the system is predicated on the public and to use park employees to verify existing technical experts. Non-technical the information before it is published. There is users will require more training to get their currently no easy way to extract information from contributions into the system. the NPS Places Project although its API is open and fully documented, meaning that experienced • Live feedback (e.g. map updates) is important users could do this. for motivating participants. The biggest success of this project is that the NPS is able to collect contributions from non-technical park staff. This includes rangers and maintenance managers who know the parks better than anyone but whose knowledge has not been captured due to a lack of dedicated GIS staff. However, there are few non-technical users to date although APPENDIX 22. USGS’s “Did you feel it?” Interaction type Public → Government Trigger event Paucity of instrumental ground-motion data in regions of low seismicity. Domain Provision of ground-motion data to US Geological Survey Community Internet Intensity Map for public information about earthquake activity. Organization United States Geological Survey Community Internet Intensity Map (popularly known as “Did you feel it?”, or DYFI) Actors US Geological Survey Earthquake Hazards Program Datasets Various products are developed by the USGS from the DYFI data and other earthquake sensing and monitoring programs. Process Citizens experiencing earthquake activity can log on to the DYFI website and submit their observation. The intensity of the earthquake is quantified using the Modified Mercalli Intensity (MMI) scale. There is no need for participants to have experience in seismology. Maps and graphics are generated automatically by the DYFI system and made available to the public. Feedback Participant submissions are acknowledged and DYFI provides personalized feedback to the contributor in the form of a computed intensity for their submission. Users can provide contact details in the input form. Goal Outreach to citizens to become part of the seismic monitoring network and to allow the USGS to continue to learn and understand more about earthquake shaking and intensities. Side effects The data collected from DYFI is made compatible with ShakeMap, a product of the US Geological Survey Earthquake Hazards Program in conjunction with regional seismic network operators. Impact of the project - Temporal pattern - Funding - Contact Point David Wald, US Geological Survey, wald@usgs.gov APPENDIX: CASE STUDIES 89 90 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT USGS’s “Did you feel it?” the perspective of human and structural response on a qualitative scale from 1 (not felt) to 10 (very The US Geological Survey’s Community Internet heavy damage), based on descriptions such as Intensity Map (more commonly referred to “felt indoors” (MMI = 3) to “felt by all, windows, as “Did you feel it?” or DYFI) is a website that dishes, glassware broken, weak plaster cracked” automatically maps reports from citizens about (MMI = 6) to “some structures with complete their perception of recent seismic activity in collapse” (MMI = 9). DYFI allows anyone to their area. When a citizen feels a tremor they can report their earthquake experience as there is visit the DYFI website, report their location and no need to have experience in seismology. This describe their experience by answering a short citizen reporting allows the USGS to continue series of questions (with drop-down menus). to learn and understand more about earthquake DYFI is combined with a large network of sensors shaking and effects. In addition, the USGS can use worldwide and these additional citizen reports the data to quickly map the intensity of shaking allow USGS to develop a more detailed map of in the affected region, which can help inform earthquake intensity. Over 4 million earthquake emergency responders, the media and the public intensities have been submitted to DYFI for in the immediate aftermath of an earthquake. thousands of earthquakes; the data and DYFI maps are available to browse online in the archive DYFI has wide appeal and the quality and vast section of the website. There is an option for quantity of the data have been used to address citizens to give first-person descriptions of how longstanding issues in earthquake ground-motion the earthquake affected them. However, it is science. Such issues have been difficult to address made clear on the form that if the USGS uses this due to the paucity of instrumental ground-motion qualitative information the citizen will only be data in regions of low seismicity (Atkinson and referred to as “the observer”. Contributors can Wald, 2007). Prior to DYFI, intensity maps were see their report displayed on the DYFI webpage. rarely made for US earthquakes of a magnitude Maps and graphics are generated automatically by less than about 5.5; now intensities as low as the DYFI system and made available to the public. magnitude 2.0. are routinely reported for the The interactive DYFI map also interfaces with smallest felt earthquakes nationwide (Atkinson 1km2 geocoded boxes, color-coded according to and Wald, 2007; Wald et al., 2011). DYFI data the USGS ShakeMap/DYFI intensity scale. has been useful for rapid post-earthquake information and “they are also robust and of The “Did you feel it?” form interface is easy to surprisingly high utility”. The data collected offer use. Citizens can submit observations by selecting the potential to not only describe ground-motion their location automatically, from an address or effects qualitatively but to be used in quantitative with the map interface. The form mostly uses scientific studies. Atkinson and Wald (2007) drop-down lists asking for feedback on: situation state that “the key to the usefulness of the data is when the earthquake occurred, experience of simply this: they make up in quantity what they the earthquake (shaking strength, duration, may lack in quality. Because there are so many reaction), earthquake effects (e.g., objects move responses, stable statistics on average effects are or fall off shelves), and damage to buildings. The produced, illuminating ground-motion trends and intensity of the earthquake is quantified using the allowing effective correlation and calibration with Modified Mercalli Intensity (MMI) scale. MMI more-quantitative ground-motion measures.” measures the intensity of ground motions from Further, Wald et al. (2016) recently noted that APPENDIX many academic papers are based on the data from the project (over 2000 in Google Scholar search) and that sometimes this is the only data available for parts of the US. Main lessons: • Quality assurance/control: website applications can be used to filter out erroneous responses. Wald et al. (2011) note that the uncertainty of the intensity reported decreases when a higher number (>3) responses are obtained per location; more responses gives a more reliable average. Project leaders can reserve the right to manually exclude any responses they deem erroneous or unnaturally low or high. • Input forms should be easy to use and avoid the use of specialized terminology. • VGI data can be valuable in filling in gaps in existing data collection. APPENDIX: CASE STUDIES 91 92 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 23. US Census Bureau – Building an OSM community of practice Interaction type Government → Public Trigger Event Internal champion Domain General mapping Organization United States Census Bureau Actors United States Census Bureau Datasets OpenStreetMap Process Mapping events and outreach in the Bureau. Feedback Internal project, no explicit feedback planned. Goal To build a group of mappers and supporters of OpenStreetMap in the US Census Bureau. Side effects - Impact of the project - Temporal pattern - Funding - Contact Point Steven Johnson APPENDIX US Census Bureau – Building an Advocates face a number of challenges in promoting the adoption of OSM in the census. OSM community of practice Many in the Census Bureau, who have traditional The US Census Bureau has no formal interaction GIS backgrounds, have difficulty accepting with OpenStreetMap. However, a small number of OSM’s open source model of data production. In internal champions worked to build a community government datasets, databases are understood as of mapping enthusiasts to create support for authoritative but OSM offers no such assurances. incorporation of OSM into the bureau’s work in The OSM community also has no central point of 2013 and 2014. contact to provide support. Thus, learning how to engage with the distributed OSM community Advocates for the use of volunteered geographic would be an important part of any adoption. information in the US Census Bureau’s programs Finally, the census could potentially use OSM says that there a number of ways in which the as a data source for TIGER as well as the master platform could be useful. Data collection is address file but the Census Bureau would then expensive and OSM could offer a way for the need to release this as public domain information, census to partner with citizens to help identify which is not permitted by the current areas where change has occurred and, perhaps, ODbL license. collect basic location about the neighborhoods where they live. OSM could also be used Main lessons: as a tool for helping census professionals • Building support for VGI in government better understand issues that census field organizations requires time and willingness to representatives encounter in their work. For learn different models of creating, validating example, by participating in OSM, they would and using data. have to engage with ambiguity in tagging systems as well as better understand the geography of the • Hands-on exposure to OSM tools is a useful areas to which they are assigned. way of helping individuals understand the platform and its potential value. In order to help grow the OSM community in the Census Bureau, advocates organized mapping • The potential for adoption depends on the events during lunch hours every other week presence of an active champion, and without during June and July of 2013. Participants were it, the activity stops. given basic information about the OSM platform and community, taught how to map using OSM tools, and sent out in small teams to survey the area around the Suitland Federal Center, where the Census Bureau offices are located. In addition, advocates gave internal presentations about OSM and its applicability to the census’s mission, and took colleagues to OSM conferences and meetups. However, as soon as the advocates left the organization, the activities stopped. The Census Bureau has no current crowdsourcing activity as part of their preparation for the 2020 census. APPENDIX: CASE STUDIES 93 94 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 24. New York City open data initiative Interaction type Government → Public → Government Trigger Event Adoption of open data policy. Domain Local authority Organization New York City GIS Department and Department of Information Technology and Telecommunications (DoITT) Actors OSM community, New York City government, Mapbox Datasets Building footprint, addresses. Process Data import into OSM and crowd maintenance in OSM platform. Feedback Daily changes. Goal Leverage volunteers to help keep authoritative data current. Side effects - Impact of the project - Temporal pattern - Funding - Contact Point Alex Barth, Mapbox. APPENDIX New York City open data initiative changes in the city may necessitate updates to the authoritative city dataset. In September 2013, New York City released over 200 government datasets to the public as part Following the import of the city data into OSM, a of a broad open data initiative to “improve the feedback loop between the city and the volunteer accessibility, transparency, and accountability OSM community allows the government and the of City government” (City of New York, n.d.a). public to work together to create and make use Using the web-platform Socrata, the data is made of up-to-date and high quality spatial data. The available for download or through APIs that import process for large and complex datasets allow software developers to construct mobile like addresses and buildings is a complex process and web-based applications that incorporate that requires technical resources, significant this information. This data release continues an labor, and solid coordination between the OSM aggressive open data push by the city government community and others involved. In this case the that began in 2011. New York City’s Open Data information released by DoITT was up to date law, signed by Mayor Bloomberg in March 2012, and of high quality, but the same cannot be said mandates that all city agencies “systematically of all municipal datasets. There has also been a categorize and make accessible in ‘open’ formats great deal of communication between the city all public at no charge” (City of New York, n.d.b) government, the OSM community and the people before 2018. To date, over 1,100 datasets have working on the import, which is critical to the been made available on the city’s open data portal success of these kinds of efforts. This example of and numerous applications have been built that cooperation between local government and the address issues ranging from transportation to volunteer OSM community around a targeted food safety and the environment. dataset could be expanded by other projects seeking to address a wider range of open data. In partnership with the Department of Main lessons: Information Technology and Telecommunications (DoITT), the OpenStreetMap community and • Open data initiatives can increase government Mapbox imported city building footprint and accessibility and transparency. address point datasets into the OSM database. • Government datasets can be imported and All work was coordinated on Github and OSM then kept up to date by the OSM community. mailing lists and completed in 2015. The project This can be especially valuable in large and history can be accessed on Github: dynamic urban spaces. https://github.com/osmlab/nycbuildings/issues. • Cooperation between the OSM community These critical datasets, which are necessary and government agencies can create direct to support a wide variety of data analysis and links between VGI and authoritative datasets. visualization projects, can be difficult and A loop can allow both groups to work together expensive to keep up to date in a city as large to create and make use of up-to-date and high and dynamic as New York. Thanks to software quality spatial data. developed by Mapbox, the New York City GIS • Technical resources, significant labor, department now receives daily emails detailing and solid coordination between the OSM changes to OSM building or address information. community and others involved is required These emails allow the GIS team to quickly assess for success. updates in OSM to augment city processes used to identify where new construction or other • Data quality can vary and municipal datasets may lack the detail developed in this example. APPENDIX: CASE STUDIES 95 96 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 25. Imagery to the Crowd, State Department Humanitarian Information Unit, US Interaction type Government → Public → Government Trigger event Creation of new initiative, MapGive. Domain Open data for humanitarian relief programs and sustainable development; generic mapping. Organization US Department of State Actors Humanitarian Information Unit Datasets Satellite imagery. Process Manual digitization of roads, buildings, other features; public diplomacy and engagement. Feedback Access to vector OSM data for local and international organizations. Goal Increase participation in the volunteer mapping community and make it easier for users to create OSM data that can support humanitarian and development efforts. Side effects - Impact of the project - Temporal pattern - Funding - Contact point US Department of State, MapGive Team APPENDIX Imagery to the Crowd, State development programs such as the President’s Malaria Initiative (PMI) and the President’s Department Humanitarian Emergency Plan for AIDS Relief (PEPFAR). For Information Unit, US example, in the 2013 Typhoon Haiyan disaster in the Philippines, IttC-published imagery Facilitating access to high quality aerial and of the affected areas supported over 1,600 satellite imagery for volunteer mapping volunteer mappers who contributed nearly 5 communities can have dramatic impacts on million imagery-based edits to OpenStreetMap humanitarian and development activities. (OSM) coordinated by the Humanitarian Such imagery is often prohibitively expensive OpenStreetMap Team. These edits provided or only available under licenses that would detailed information on the location and extent prevent their broader use. With this in mind, of pre-event infrastructure and provided the basis the US Department of State’s Humanitarian for a preliminary damage assessment. As is the Information Unit (HIU) launched an initiative in case in the aftermath of other natural disasters, 2012 called Imagery to the Crowd (IttC), which organizations like UNOCHA, MapAction, the makes high-resolution imagery – purchased and World Bank, the American Red Cross and others licensed by the US government from providers used the OSM data created through these efforts like DigitalGlobe – accessible to humanitarian for the humanitarian response. Another project, organizations for use by the volunteer mapping implemented in partnership with the Global communities that support them. To build on Facility for Disaster Reduction and Recovery the success of this initiative, in 2014 the HIU and USAID, organized volunteers in Nepal, the and the Bureau of International Information United States, Germany and the United Kingdom Programs created MapGive, a public diplomacy to digitize roads and building footprints in the initiative founded on open mapping and citizen Kathmandu Valley of Nepal. Kathmandu is one engagement. MapGive makes it possible for of the most seismically at-risk cities in the world large numbers of volunteers to contribute to and the data created through this program helped the growing global open mapping community inform an earthquake retrofitting program and leverages partnerships in the humanitarian led by the government of Nepal and disaster and development communities to direct response planning by USAID, which ultimately volunteer efforts to support priority mapping played a critical role in the response to the 2015 projects worldwide. earthquake. MapGive continued to operate in In addition to leveraging the imagery publishing different activities and events in 2016. workflow developed for IttC, MapGive provides These examples demonstrate that sharing US an array of support to new and novice mappers government-purchased imagery with volunteer in the volunteer community via its website. mapping communities has resulted in the creation This includes beginner tutorials on mapping, of valuable spatial data that is accessible to toolkits for hosting mapathons and links to governments, international agencies, and the training resources. Since the launch of IttC public. Moreover, technical and policy efforts at and MapGive, the US Department of State has the Department of State resulted in an increase supported humanitarian response operations in the speed and quality with which imagery is in the Philippines, Nepal, Haiti and many served to the volunteer mapping community. The other locations worldwide. In doing so, it technical workflow for processing and managing has partnered with multilateral international APPENDIX: CASE STUDIES 97 98 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT the imagery is available on the Department’s Main lessons: GitHub page and is completely open source. At the programmatic level, in cases like USAID’s • Direct government initiative can grow work in Nepal, this also becomes a full-cycle the volunteer mapping community, example where a US government agency makes especially when supported by the release direct use of the data created as a result of the of high-resolution imagery, purchased and imagery release. In other instances, the users licensed by the government for use by are UN agencies or not-for-profit organizations humanitarian organizations and volunteer working towards primarily humanitarian ends. mapping communities. MapGive was included in the Quadrennial • Government initiatives of this kind can result Development and Diplomacy Review and the in partnerships with numerous humanitarian Open Government Partnership’s Third National and development programs throughout the Action Plan, which are recent fundamental US world, including responses to natural disasters government policy documents on open data, open and health campaigns. government and diplomacy. • The release of data is most successful when accompanied by a robust open source methodology for processing and hosting (e.g. in the OSM Tasking Manager (tasks.hotosm. org)). • Government initiatives can achieve demonstrable results by creating projects that produce data for use by the government itself, as well as its international partners. APPENDIX 26. Humanitarian OpenStreetMap Team mapping in Ulaanbaatar, Mongolia Interaction type Public → Government → Public Trigger event A greater vision to create a “smart city”. Domain Topographic mapping Organization Ulaanbaatar City Governor’s Office Actors World Bank/ICT, HOT, Mongolian University of Science and Technology, city officials, Mongolian Land Management, Geodesy, and Cartography Department Datasets Aerial and satellite imagery (Bing Maps), field survey. Process Training core group of people in field data collection, mapping a part of the city, creating conditions for the project to continue by setting up an OSM community. Feedback Topographic maps of the city. Goal Map a part of the city, create an OSM community and train local people to continue the mapping effort in order to support the vision to transform Ulaanbaatar into a “smart city”. Side effects – Impact of the project National Temporal pattern Ongoing Funding Public Contact point - APPENDIX: CASE STUDIES 99 100 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Humanitarian OpenStreetMap Another event was a two-day mapathon, during which over 20,000 edits were made to OSM Team mapping in Ulaanbaatar, using unmanned aerial vehicle (UAV) imagery Mongolia as a base layer. The Governor and Mayor of Ulaanbaatar embraced this effort, presenting In March 2013, the local authorities of awards to the three top teams, and intend to Ulaanbaatar in Mongolia announced their vision incorporate publicly collected OSM data in future to turn their city into a “smart city” by 2020. policymaking. The datasets created could also be The aim is to enable city residents to access used by city’s planning department to augment information regarding public services, provide a their own mapping. consolidated list of public service assets and allow citizens to send comments, reports and requests Main lessons: to relevant officials via the internet. To support this effort a project to map the city under OSM • Building an OSM community from scratch is guidance was funded by World Bank/ICT. likely be time and resource-intensive, and slow to begin with. The process started with basic training on tracing • Spatial datasets and mapping products might features from imagery and moved to the handling suffer from low quality, at least until an active of GPS devices in data collection. The datasets OSM community forms. were documented by field papers and ground photographs which facilitated data management • Local tagging requirements might be missing before the final uploading. A consensus on the from the overall OSM project or might be best tagging practice had to be achieved in order considered as outliers when it comes to to describe geographic features that do not exist normalizing the dataset into a spatial product in other areas of the world and are therefore not (e.g. Shapefiles, import into a database documented elsewhere. schema, etc.). An important challenge was that the city • Both governmental officials and private of Ulaanbaatar was undergoing a great sector organizations recognize the value and reconstruction phase and thus the available potential of VGI and open data. imagery datasets might not provide up-to-date • The existence of active volunteering information, raising quality concerns. On the communities can easily embrace the use of positive side, after five weeks the project had: innovative and largely disruptive solutions created an OSM community to continue the such as the replacement of satellite imagery mapping project, improved the awareness of local with UAV images. officials of the use of VGI and open data and also caught the attention of the private sector, • Continuous training of new volunteers which can enhance OSM community efforts and organizations, plus the support of by providing resources while at the same time events such as mapathons, can support drawing more official attention to them. project sustainability. • Communicating the results of volunteered The project continued on a volunteer basis. One efforts to the public sector is important for of the main events was when a group of student recognizing the merit and the potential of such interns spent part of their spring semester helping projects. the remote mapping of areas around Ulaanbaatar. These students continued their effort during the fall semester by joining technical projects. APPENDIX 27. Mapping schools and health facilities in Kathmandu Valley, Nepal Interaction type Public → Government → Public Trigger event Increasing natural disasters worldwide, including earthquake in Haiti. Domain Generic mapping of schools and health facilities. Organization Department of Education and Kathmandu Living Labs (KLL) with support from the World Bank / GFDRR Actors Citizens, Department of Education, KLL, The World Bank/GFDRR, Nepal Risk Reduction Consortium, NSET Datasets Aerial imagery from Bing and HIU, and list of schools and hospitals from government were used. Datasets on road network and other points-of- interest with focus on school and health facilities have been generated. Process Initial data about schools and hospitals were mapped using a variety of techniques and the results were presented to authorities and discussed. Feedback Interactive thematic map (for schools and hospitals) showing structural and non-structural attributes. More detailed online map of Kathmandu Valley. Goal To increase resilience of schools and hospitals to earthquake risks. Side effects - Impact of the project Local Temporal pattern One-off Funding World Bank Contact point Nama Budhathoki APPENDIX: CASE STUDIES 101 102 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Mapping schools and health In April and May 2015, two high-magnitude earthquakes struck Nepal, killing nearly 9,000 facilities in Kathmandu Valley, people and destroying over a half a million homes. Nepal Scarce and outdated geographic information hindered government understanding of their Nepal is considered one of the countries needs and assets, in turn making it difficult for most exposed to natural hazards, especially earthquakes. The capital city of Kathmandu has them to effectively prepare for, or manage, a experienced rapid urbanization in the last decades natural disaster, handle logistics and support and is considered to be vulnerable to earthquakes medical care, shelter, food and water needs. as the majority of the houses do not meet In this context, the KLL personnel and the earthquake safety requirements. In Kathmandu, volunteer community focused on supporting the local stakeholders have recognized the danger disaster management efforts and then on the and have tried to be proactive by initiating an reconstruction work, discontinuing the initial effort to create an OSM map of the city. The project. However, the information gathered aim is to provide a critical resource for disaster from the schools and health facilities mapping risk mitigation and emergency planning. In 2012 project proved crucial and helped inform the World Bank’s South Asia Region launched humanitarian responders and support recovery the Open Cities project to create an asset and efforts. The information was also helpful in exposure database for urban areas and facilitate determining which facilities must be retrofitted to its use for urban planning and disaster resilience. withstand tremors. As part of this project, Open Cities Kathmandu Main lessons: was started in November 2012 as a pilot initiative. • Being proactive is key to ensure that an area is The process was to use GPS, paper field maps, satellite imagery, web and mobile technology to prepared for future natural disasters. collect exposure data of schools/colleges and • A solid mapping background is needed health facilities in the field. This was followed by for relief efforts following disaster. a rigorous data validation process, which ensured Creating or updating an existing map is of that data were accurate. Open Cities Kathmandu great importance. has to date mapped over 130,000 buildings and • A well-managed and coordinated effort collected exposure data for 2256 educational and 350 health facilities. to drum up public support can provide valuable input from both local and The outcome of the project has sparked a international contributors. policy-level discussion about ensuring the safety • As well as a short, intensive mapping effort, of schools and health facilities in emergency it is also vital to create a community that will situations and encouraged some government continue the task to complete or update the agencies to explore ways to integrate VGI in their maps. workflows and others to share their datasets with the public. As part of the Open Cities program, more than 1,500 people received training on OSM procedures and a large number of presentations were delivered to universities in an effort to build a robust OSM community. APPENDIX 28. Informal settlement mapping, Map Kibera, Nairobi, Kenya Interaction type Public → Government → Public Trigger event - Domain Generic mapping of the biggest informal urban settlement area and thematic mapping of security, water sanitation, health and education. Organization Map Kibera Actors Map Kibera team, GroundTruth Initiative and various partners Datasets OSM Process Field mapping in OSM with GPS devices or mobile phones after training workshops. Feedback Thematic maps for the management of supplies in infrastructure (roads, paths, railway tracks), health (locations of medical facilities), education (locations of schools), water (locations of wells or spigots) and sanitation (locations of pit latrines, sewers). Goal Map the unmapped Kibera and actively involve local people. Side effects - Impact of the project National Temporal pattern Ongoing Funding GroundTruth Initiative Contact point Erica Hagen, GroundTruth Initiative APPENDIX: CASE STUDIES 103 104 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Informal settlement mapping, Among the main successes is the project’s acceptance by local government, by which it was Map Kibera, Nairobi, Kenya embraced from the beginning. At the end of the The homepage of this project welcomes visitors by project, Map Kibera representatives presented the stating that, “Kibera in Nairobi, Kenya, was a blank analysis to government officials. The negotiation spot on the map until November 2009, when young between the two sides had a positive impact on Kiberans created the first free and open digital map the community, which became recognized as a of their own community.” The welcome message real neighborhood, and residents gained new summarizes the main idea behind the project, which technological knowledge. is to map one of the biggest informal settlements of The project faced various challenges. First, the the world by putting marginalized communities on voluntary participation model was unrealistic in the map. Kibera – local people face great survival issues so Map Kibera was launched in 2009 by Mikel Maron a small daily compensation was given for their and Erica Hagen with initial funding from Jumpstart participation. Second, residents found it hard to International, an NGO specializing in community- understand the benefits of participation and the based mapping. The first phase, which lasted three potential impact of the project. Finally, NGOs found weeks from October to December 2009, involved 13 it difficult to cooperate and share information. They young people who were trained to collect and edit had learned to work separately and competitively for GPS tracks. OpenStreetMap (OSM) was used to a long time, which meant that voluntary work was create a dynamic and easily editable map and QGIS divided into small pieces, for different purposes. software was adopted to do further analysis and As of May 2017, the project is still active with part create specialty maps. ARCGIS, a non-open source of the original team. Interest has also extended to software, TileMill and other MapBox products were various fields, such as the mapping of more than also used. 350 schools. MapKibera also played a critical role From February to August 2010, mappers had the in the elections of 2013 where citizen mapping opportunity to enhance points of interest such as pinpointed major issues such as crime hot spots. water, public toilets, schools, police stations and Its intervention led to a parliament member clinics. It also included two other mini projects: committing to addressing those specific issues. Voice of Kibera and the Kibera News Network. The Main lessons: first offered the opportunity to submit reports, write articles and add breaking news with the aid • Slum mapping can be achieved by young local of Wordpress blogging and Ushahidi software. people relatively quickly. Work could be sent by SMS and published after • Compensation may be needed to improve approval by an editorial team. The second is a video participation in locations where participants journalism initiative offering more local people face great survival issues. the opportunity to participate, ensuring the wider acceptance of the project and hence its longevity. A • Innovative methods such as SMS, voice and YouTube channel also reports on priority issues. video reporting can support the appeal of mapping projects. Since then, the map has been updated in real time by users who report service gaps and incidents. As the data constantly changes, the project makes use of the OSM platform, which crowdsources updates and corrections to improve accuracy. APPENDIX 29. Skandobs, Scandinavian predator tracking system, Norway and Sweden Interaction type Public → Government → Public Trigger event Policy requirements for wildlife monitoring in Scandinavia. Domain Species population observation and tracking. Organization Rovdata (an independent part of the Norwegian Institute for Nature Research (NINA)) and the Swedish Environmental Protection Agency (EPA) Actors Rovdata in Norway and the Swedish Environmental Protection Agency Datasets Citizens submit their observations to Skandobs. A Google Maps-based visualisation tool is also available to allow map-based visualisation of the Skandobs database. Process Participants can submit observations using either smartphone applications or the Skandobs website. Geographical data is required with each observation and anonymous submissions are not allowed. Quality checks are performed by Skandobs staff. Feedback The number of observations in the database is updated on the website every 15 minutes. A table provides summary information of the total number of observations for the current year and month. League tables also show top contributors by individual and municipality. Goal To collect observations for lynx, wolverine, brown bear and wolf locations and population sizes to increase knowledge about species populations in Scandinavia. Side effects Increasing expectations from NGOs, the media and the public who want information on numbers and distribution of carnivore populations in their countries. Impact of the project International Temporal pattern Ongoing Funding Public funding Contact point - APPENDIX: CASE STUDIES 105 106 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Skandobs, Scandinavian predator User observations are subjected to internal validation and when this is completed they are tracking system, Norway assigned a validation status, which appears in the and Sweden list of observations. Only rated observations are also added to Rovbase, the primary database for Skandobs was developed to collect observations national population monitoring data in Norway for lynx, wolverine, brown bear and wolf locations and Sweden. The number of observations in and population sizes, driven by new legislation, the database is updated on the website every international conventions and directives. Reliable 15 minutes with tables summarizing the total observations help to inform management number of observations plus totals for the year objectives including long-term conservation and month. Observation totals are also provided and population level management. Monitoring for key species. is complicated and expensive for many reasons, including the geographical size of the area under There are concerns about the potential bias in the observation but involving the public has been observational data. For example, because there is very successful. greater population density in southern Sweden, there will be more observations than in northern In January 2017, there were more than 20,000 Sweden. This problem of the distribution of registered users of Skandobs, presenting a greater observations/observers is a challenge in citizen growth in the Swedish territory. There were about science and VGI more generally. 7,800 Norwegian registered users and about 12,500 registered users from Sweden. There are Main lessons: more than 16,000 registered users, while in the last seven years the system has received more • Feedback can be provided to participants than 14,000 reports of large carnivores, with through real-time summaries of the types and more than 3,000 observations from July 2016 quantities of contributions. to January 2017. More observations have been • Nations can work together on VGI initiatives recorded in the Swedish territory (9,500 reports to meet policy requirements and reduce costs. against approximately 4600 in Norway) but this is an expected bias since there are more people and • VGI initiatives can work well in carnivores in the Swedish territory. transboundary situations. • VGI projects can improve volunteer and public The Skandobs database is jointly maintained awareness about the topics of concern. by two national agencies, Norway’s Rovdata and Sweden’s EPA. Citizens can submit their • Observational bias due to population observations at any time, using iPhone or distribution remains an inherent challenge Android smartphone applications or through in VGI. the website. The addition of photographs and other information is encouraged. The Skandobs database is made available for download via a search interface on the website. A Google Maps-based visualisation tool is also available to allow map-based visualisation of the Skandobs database. APPENDIX 30. Corine Land Cover 2006 (CLC2006) in OpenStreetMap, France Interaction type Government → Public Trigger event Change in the license policy of the Corine Land Cover dataset. Domain Generic mapping (update of land cover/ environmental datasets). Organization European Environment Agency (EEA) Actors OpenStreetMap France, EEA Datasets 44 land cover classes for France. Process CLC2006 data not overlapping existing OSM data have been imported to the dataset (account for ~60 percent of the land). The CLC2006 typology has been adjusted to match OSM. Feedback – Goal The goal was to update the OSM database with land cover information, mainly in rural areas, as the contribution in such areas is limited. Side effects Other OSM communities followed the example and have integrated CLC2006 into their OSM database. Impact of the project National Temporal pattern One-off Funding None Contact person Dr. Guillaume Touya APPENDIX: CASE STUDIES 107 108 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Corine Land Cover 2006 data up to date, the creation of a false impression that such datasets are more accurate or more recent (CLC2006) in OpenStreetMap, than they actually are, the need to address semantic France inconsistencies and differences in the level of detail. Corine Land Cover (CLC) is a European All these factors have probably contributed to the Commission-supported program that aims to subsequent stance of the French OSM community provide land cover data for 39 European countries. The image production for the land cover digitization towards newer CLC datasets (CLC 2012, released was centrally coordinated by the European at the end of 2014) as the prevailing solution was Environment Agency (EEA) and the actual data that the bulk import should be considered only production was undertaken by EEA member states as a starting point and now the community prefer to “benefit from local knowledge”. CLC2006 not to use CLC datasets at all. Moreover, the includes 44 land cover classes. community is urged to remove the CLC identifiers and to update the “source” tag by adding the actual As permitted under the release terms of CLC2006, the French OSM community imported the CLC2006 source of imagery that was used in order to update into the OSM database. However, only about 60 the polygons. percent of the original was automatically imported, Main lessons: which was for those areas where there was no OSM data. The rest was not imported because it was in • The existence of active public communities conflict with existing land cover polygons created facilitates take up of opportunities to work on manually by OSM contributors. This was deemed open source data. more efficient as the OSM community realized that land cover polygons created by OSM contributors • Data integration should not be considered were more accurate than CLC2006. Moreover, easy or straightforward. This should also be as OSM contributors now have access to high- made clear to volunteers as any integration resolution Bing aerial imagery, their data should be initiative might mean large workloads with even more accurate than the CLC, which is based moderate results. on IMAGE2006. • Considerable expertise among volunteers is The integration of CLC2006 to the OSM datasets required for success. instantly enriched the latter with data regarding 60 percent of the French territory. Land cover • Successful, community-led efforts can be classification based on imagery interpretation needs replicated by others now experience in solving considerably more expertise than road classification problems has been built and often shared. and in general attracts fewer contributors than • Multiple datasets often have semantic the “high-profile” urban fabrics. However, in this inconsistencies and temporal accuracy should case, the land cover parcels imported serve as first-class photograph interpretation keys to aid the be addressed during integration OSM community. • The difficulties of authoritative and The integration brought to light a number of crowdsourced data conflation might deter issues. First, importing authoritative data into a VGI communities from relying on, or using, VGI database brings both positive and negative authoritative data and instead try to create the endogenous issues, such as the failure to keep the data needed on their own. APPENDIX 31. FixMyStreet for municipality maintenance information, UK Interaction type Public → Government Trigger event – Domain Local authority/municipality maintenance. Organization mySociety (originally developed with central government funding) Actors – Datasets Originally, the website used government datasets: postcodes, basemaps, local authority boundaries, contact details and email addresses of relevant personnel in local authorities. Process Problem reported on a website using the location’s postcode, which creates an email that alerts the local authority to the problem. The authority can respond to the complaint on the website. Feedback Update to participants provided through the website and by email that can be triggered when a person registers. Goal Provide online tools for residents to report local problems to their local authority and follow up the exchange with the public body in an open way. Side effects The application’s source code has been released under the GNU Affero GPL software license and has been used in other countries as well. Impact of the project Global Temporal pattern Ongoing Funding Local authorities funding and donations Contact point – APPENDIX: CASE STUDIES 109 110 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT FixMyStreet for municipality adopted it so far. maintenance information, UK Today, FixMyStreet continues its success and its codebase has been used to set up similar sites in FixMyStreet is a web-based application, which more than 20 countries worldwide. FixMyStreet enables the public to report local problems (e.g. staff also work in partnership with the authorities abandoned vehicles, graffiti, potholes). Issues to develop new features that make it as useful and reported by users are propagated to the relevant simple to use as possible. local authorities by email. Users reporting a problem are contacted by FixMyStreet after four Main lessons: weeks to check if the issue has been resolved. • Given technological advances it should be FixMyStreet was built and administered by expected that the public will find their way to mySociety and is free to the public. However, web applications of this kind. there is also a FixMyStreet for Councils • Authorities should develop processes and application, which is a paid version adapted to methods to integrate data from the public local authority needs for handling the problem in order to avoid issues of data duplication, reports. FixMyStreet enables the public to voice confusion and misunderstandings. concerns regarding local issues and has succeeded in engaging people and opening a channel for • Public VGI initiatives that address community public input into problem solving. Moreover, it problems might overshadow existing has resulted in public value creation both direct authoritative structures. Flexibility in (i.e. social gain through an immediate relationship embracing such initiatives might minimize with the user that reports the problem and his/ future problems for local authorities and her neighbors) and indirect (i.e. social gain that is maximize impact on society. dispersed to the entire community). • Successful examples of public participation However, there are some issues of concern can increase pressure on authorities for data especially when it comes to the data created: sharing under flexible license schemes. the cycle of public data creation, propagation, consumption and diffusion back to the public creates inconsistencies. More specifically, councils themselves have an independent channel for local problem reporting and thus the launch of FixMyStreet created a parallel channel that often simply duplicates problems already known to the council or, in the worst case, confuses the authorities as descriptions of the same problem appear slightly different. Additionally, when the council fixes the problem (which might take more than four weeks), it is not able to report progress on the issue in the application. This is the driving force behind FixMyStreet for Councils but as this is a paid version, few local authorities have APPENDIX 32. FINTAN vernacular placenames project, Ordnance Survey and Maritime and Coastguard Agency, UK Interaction type Government → Public → Government Trigger event Increasing recognition of the use of vernacular placenames and their importance to emergency services. Domain Emergency operations such as search and rescue. Organizations Ordnance Survey (OS) and Maritime and Coastguard Agency (MCA) Actors OS (research staff, data capture staff), MCA (coastguards), volunteer coastguards, general public Datasets OS topographic maps; address and placename gazetteers containing current known placenames. Process FINTAN application developed by OS to enable MCA and their volunteers to enter, locate and classify vernacular placenames; OS then assessed the quality of the data. Feedback Many new vernacular placenames that are assessed and quality controlled by OS. Goal Improve ability of the Coastguard to locate people when contacted in emergency situations; improve OS database of placenames for use in new and existing products. Side effects Tensions in areas where English is not the first language – some groups unwilling to report English names even when in common usage by local people. Impact of the project - Temporal pattern - Funding - Contact point Glen Hart, Ordnance Survey, glen.hart@ordnancesurvey.co.uk APPENDIX: CASE STUDIES 111 112 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT FINTAN vernacular placenames current known placenames have been used. Special care has been taken to allow other emergency project, Ordnance Survey and services that use different reference systems to work Maritime and Coastguard Agency, to a common geography and terminology. UK Through FINTAN, HM Coastguard and their Two of the most prominent UK governmental volunteers can enter, locate and classify vernacular agencies, the HM Coastguard of the Maritime and placenames of features such as beaches or rocks to Coastguard Agency (MCA), responsible for the update the existing database. Personal connection initiation and coordination of all civilian maritime with participants has been helpful in encouraging search and rescue operations, and Ordnance Survey participation, facilitated by coastguards inputting (OS), the UK’s national mapping agency, have joined data from their local volunteers, and the specialist forces to create an up-to-date dataset of vernacular application has been targeted at their specific placenames. Vernacular placenames are those in interest (as opposed to more generic approaches common usage irrespective of whether they are like OpenStreetMap). While the future of FINTAN official names or not. was promising in 2014, the project is now supported by one permanent staff member at the OS, who is There has been an increasing recognition by the responsible for code maintenance and debugging. To OS of the need to capture richer and more detailed date, the project has not been upscaled. vernacular placenames. At the same time, the UK Coastguard were in the process of closing On the negative side, tensions in areas where approximately half of their coastguard centers and English is not the first language have been observed recognized that a lot of local knowledge could be as some groups are unwilling to report English lost. In that context, both sides recognized the names even when these are in common usage by mutual benefit of capturing vernacular coastal local people. placenames using the knowledge of the coastguards Main lessons: (professional crowdsourcing) and local coastguard volunteers (local people who can assist in certain • Crowdsourcing can be a valuable tool when it coastguard operations). comes to existing knowledge preservation. An updated database of vernacular placenames is • Professional crowdsourcing might be equally a priceless tool when it comes to providing swift or even more productive compared to general response to life-threatening situations. One of the crowdsourcing practices when collecting data most important parts of the response process is to for specific areas or subjects. understand the position of the person in difficulty • Local and regional interests and perceptions and often official mapping products might not might introduce biases to the data collected. provide all the necessary information. It is common for people in need to use local names that do not • Two-way data flow can be a win-win situation exist in official gazetteers. for all involved parties as well as the general public, and is facilitated by shared interests FINTAN is an OS application that enables the and benefits. crowdsourcing of vernacular local names of coastal areas. FINTAN includes topographic mapping on a wide range of scales provided to the Coastguard to enable the recording of placenames. Additionally, address and placename gazetteers containing APPENDIX 33. Boston StreetBump, US Interaction type Public → Government Trigger event In 2012 Boston Mayor Thomas Menino announced that the city could take a more proactive approach to road maintenance by engaging the city’s citizens and their smartphones. Domain Road network maintenance. Organization Mayor’s office, City of Boston Actors City of Boston Roads and Public Infrastructure Division Datasets There are no specific datasets in use. Process Using the accelerometer and GPS devices on smartphones, users with the StreetBump application automatically report “bump” or pothole information to the City of Boston. These reports are collected in a database. The Roads and Public Infrastructure Division carefully monitor these reports. If a specific area receives many reports, an engineer will physically examine that location. Feedback Users of the StreetBump application can gain points (“street cred”) for each pothole they assist in reporting, which is subsequently fixed or repaired by the city engineers. Goal To overcome the manual, antiquated method of reporting potholes from citizen complaints or manual survey by city inspectors. Side effects The City of Boston uses this information to plan long-term investments in road and street infrastructure repair, upgrade and construction. Impact of the project Local Temporal pattern Ongoing Funding Public Contact point Information extracted from various websites. APPENDIX: CASE STUDIES 113 114 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Boston StreetBump, US by people with cars. This directed repair crews to wealthier neighborhoods, where people were Boston’s Mayor’s Office of New Urban Mechanics more likely to carry smartphones and download (MONUM) pilots experiments that offer the the app. The project ended in 2015. potential to improve the quality of city services. New technology – from smartphones to GPS Main lessons: – and a resurgent spirit of civic engagement • VGI can be used to address urban service have created increased opportunities for closer issues such as problems with roads and connection and communication between the city street infrastructure. government and its citizens. • “False positives” in reports can be effectively Using the motion-sensing capabilities of handled by applying a limit to the number of smartphones, volunteers who download Boston’s reports which are required before action is StreetBump application automatically send taken by the government agency. information to the city about the condition of the streets they are driving on. When their cars hit • The types of devices and their hardware a pothole their phone sends the accelerometer capabilities must be considered when data to a server application, which combines the planning to use VGI. Some technologies can information from many other phones to pinpoint be unreliable, over-sensitive or unsuited to problem areas on the streets. If three or more particular applications. bumps occur at the same location, the City of • Social biases can occur when there are fiscal Boston will physically inspect this obstacle and prerequisites in the use of technology. assign it to a queue for short-term repair or • Careful study of the data contributed might record its location to assist with longer-term reveal patterns or other sources of problems repair planning. that usually pass unnoticed by the authorities There are some problems around the reporting of “false positives”. The use of phone accelerometers means that other vibrations felt/absorbed by the phone can be incorrectly calculated as a pothole or bump in the road surface. Now the app uses accelerometers, smartphone GPS and machine intelligence to successfully detect problems while keeping false positives under 10 percent. The City of Boston plans to open source the StreetBump code so that it can easily be adopted by other cities. Another issue is that the app is tedious to use since users need to record any given road trip – even more demanding is the fact that the app cannot run in the background and thus limits access to other apps (i.e. navigation, telephone, etc.). Finally, social bias has been recorded since the app can only run on smartphones and be used APPENDIX 34. California Roadkill Observation System (CROS), US Interaction type Public → Government Trigger event A university-based education and research project. Domain Development of a database of user contributed observations of roadkill incidents. Organization Road Ecology Center at UC Davis Actors Members of the public are asked to contribute. No experience is necessary. Anonymous contributions are possible. ~50 percent of active contributors are professional biologists Datasets CROS developed a spatial database to store all of the roadkill information and on which GIS analysis is performed. Process A user can contribute details of incidents of roadkill in California (location, type of animal, photographs, etc.). Contributions are uploaded on the CROS website then submitted to the CROS database and displayed on the public web map interface. Feedback The contributor of roadkill information is provided with feedback through the project’s website. In addition, data is periodically analyzed to produce “hotspot” maps shared with users and the public. Goal To understand the ecology of wildlife-road interactions, wildlife behavior and how transportation contributes to wildlife loss. This includes the application of GIS and statistical modeling to predict roadkill hotspots, to measure the factors contributing to roadkill, to quantify impacts, and to estimate benefits of different remedial actions. Side effects The California Department of Transportation (Caltrans) and consultant organizations use the spatial information to support mitigation planning. Impact of the project Local, national and international Temporal pattern Ongoing Funding None, supported by volunteer effort by contributors and staff. Contact point Dr. Fraser Shilling APPENDIX: CASE STUDIES 115 116 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT California Roadkill Observation inform transportation planning in the state of California. System (CROS), US One of the main obstacles to the project has been The California Roadkill Observation System the tension between agency scientists’ desire to (CROS) system (http://www.wildlifecrossing.net/ use the data to inform transportation decision california/) can be used to record observations making and the concerns the same agencies from reporters out in the field who come across express regarding public access to information identifiable road-killed wildlife. According about WVC. Moreover, it is possible to litigate to the Humane Society of the United States, against public transportation organizations if over a million animals are killed every day on an individual is injured on a public highway/ US highways. road in a place where the organization knew, or The CROS website provides a systematic and should have known, there was a hazard (such consistent way for data input using mainly forms as wildlife crossing the road). As a result, these and drop-down lists. The observations recorded organizations oppose data collection and public include information such as the type of animal dissemination, even though their staff contribute and/or species found, where the roadkill was to the database and use the data. located, when it was found, how long it might Main lessons: have been dead, pictures of the roadkill, and various other additional details about road or • Internet-based training materials can be traffic conditions. Information about where provided to allow contributors to undertake wildlife vehicle collisions occur, what animals training in scientific methods of surveying. are involved, on what kinds of roads, and other • The user interface for contribution should data can help inform policy, management, and be easy to use and include widgets such financial investment in reducing roadkill. CROS as predefined lists and clickable maps to has been successful in gathering a large amount of decrease the opportunities for erroneous data and by June 2017 it had gathered over 54,100 data submission. road observations about 425 species, contributed by 1,340 observers. • Photographs submitted in addition to basic geographical data can provide useful visual The observations are used in a geographic context information. information system (GIS) to find stretches of highway where wildlife-vehicle collisions (WVC) • Tensions may exist between agencies due occur more frequently (high density) and places to legal issues that might arise from public where there are statistically significant clusters dissemination of the data. of WVC (hotspots). The use of data also includes GIS and statistical modeling to predict roadkill hotspots, measure the factors contributing to roadkill, quantify impacts, and estimate benefits of different remedial actions. The research organization collecting this data at the University of California, Davis, hope to use this data and their GIS analysis of the roadkill problem to APPENDIX 35. Crowdsourcing satellite imagery in Somalia Interaction type Public → Government Trigger event - Domain Humanitarian/disaster mapping Organizations Standby Volunteer Task Force (SBTF) and UNHCR Actors Standby Volunteer Task Force, UNHCR, DigitalGlobe, Tomnod and Ushahidi Datasets Satellite imagery provided by the Standby Volunteer Task Force in Tomnod platform. Process Volunteers identify shelters by using satellite imagery. Data is stored in Tomnod platform and is evaluated using theCrowdRank algorithm. Data is then pushed to a dedicated UNHCR Ushahidi platform for further analysis. Feedback Thematic maps to identify and count shelters in the Afgooye corridor in Somalia. Goal Map and count shelters in a sensitive area in order to help refugees and internally displaced people. Side effects - Impact of the project Local Temporal pattern One-off (seasonal) Funding UNHCR Contact point - APPENDIX: CASE STUDIES 117 118 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Crowdsourcing satellite imagery of charge. The algorithm determined which tags had the most consensus across volunteers. This in Somalia built-in quality control mechanism is a distinct This was a humanitarian project to geolocate advantage of using micro-tasking platforms like all shelters in Somalia’s Afgooye corridor with Tomnod. The tags with the most consensus were the aid of satellite imagery provided by the then pushed to a dedicated UNHCR Ushahidi Standby Volunteer Task Force (SBTF). UNHCR, platform for further analysis. SBTF and students DigitalGlobe, Tomnod, SBTF and Ushahidi are from the American Society for Photogrammetry the main organizational bodies cooperating so and Remote Sensing were among the volunteers that crowdsourcing can take place with the aid who contributed to this project. of volunteers. The aim of the project is to map The strength of the project was in the rapid all shelters by dividing them into three main recruitment of volunteers, and the collaboration categories: large permanent structures, temporary with data providers with access to high-resolution structures with a metal roof and temporary imagery. At the same time, there are challenges in shelters without a metal roof. The rule set creating a systematic data classification system describes the shape, color, tone and clustering for buildings that rely only on satellite imagery, of the different shelter types. The project was and therefore access to high-resolution imagery divided into two phases: a trial and an official continues to be needed. launch where specific instructions were given to participants. The goal of the project is to test Main lessons: the feasibility of crowdsourcing rapid shelter enumerations of internally displaced persons to • Innovative techniques and algorithms can be support population estimates. The process cannot used to evaluate volunteers’ work. be replaced by an automated system of image • Customized platforms and high-resolution identification because such a system could not satellite imagery may significantly support a identify the type of shelter. crowdsourced project. The satellite imagery methodology was selected • A trial period may indicate crucial factors to instead of on the ground survey because access avoid. to the area of interest is limited. The main task was to find out how many people are in the shelters and need humanitarian aid in order to inform decision making around logistics and planning policies. During the project, over 253,700 tags were created and more than 9,400 shelters visually identified after the processing of about 3900 satellite images. Satellite imagery was provided by DigitalGlobe to meet the requirements for high resolution. Tomnod also offered its CrowdRank algorithm to triangulate the data and a dedicated crowdsourcing platform for imagery analysis free APPENDIX 36. Portland TriMet, transportation planner, Oregon, US Interaction type Public → Government → Public Trigger event Need for a new generation of data that traditional proprietary datasets cannot provide. Domain Transportation data Organization Tri-County Metropolitan Transportation District of Oregon (TriMet) Actors OSM and public transportation authorities. Datasets The streets and trails are fromOpenStreetMap; transit routes and schedules from TriMet; and elevation data come from the US Geological Survey. Process The project uses OSM data to provide intelligent services to the public and at the same time improves OSM data. Feedback Improved services to the public. Goal Improve the provision of public transportation services. Side effects Development of an ecosystem of transportation applications. Impact of the project Local Temporal pattern Ongoing Funding Public Contact point Bibiana McHugh APPENDIX: CASE STUDIES 119 120 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Portland TriMet, transportation transportation system in which personalized mobility will be a key feature. TriMet is working planner, Oregon, US with the OSM community and other local The Tri-County Metropolitan Transportation jurisdictions to improve OSM data, in particular District of Oregon (TriMet) serves approximately on sidewalks, to support the next generation of 1.5 million people in the 533 mile urban portion trip planning. of the three-county Portland metropolitan The intertwining of OSM and TriMet has area. In an effort to keep the services provided sparked the creation of 57 TriMet apps and the to the public at the highest possible level, development of several OSM wiki pages dedicated TriMet is incorporating a range of services to Open Trip Planner and TriMet applications such as multimodal trip planning and service with directions and best practices for mapping change analysis. and tagging. TriMet needed a new generation of data that Main lessons: traditional proprietary datasets cannot provide. To solve this problem, TriMet turned to OSM, • Strong commitment to open source since it provides data about pedestrian paths software, open data and crowdsourced data and bicycle routes that can enable TriMet to can create a full spectrum of useful web/ provide multimodal services to the citizens of mobile applications Portland. OSM datasets are now used by all • Cooperation with sources of crowdsourced internal systems and applications that need geographic information should be a two-way routing data. Moreover, city officials realized partnership that includes both use of VGI that by releasing governmental data to the data and support for further improvement of public, the OSM community would contribute those datasets. to data improvement and will enable TriMet to provide better services to the public. Many • Continuous funding is needed for applications, transit agencies also share their networks and services and datasets to remain up to date and schedules as public datasets, using the General in line with the latest trends of application Transit Feed Specification (GTFS) format created development, both in terms of usability by Portland TriMet and Google. Combining the and functionality. geographic information of OSM and the temporal information of GTFS feeds provides many possibilities for TriMet. TriMet also led efforts to bring OSM to a high level of accuracy for their area of interest and is the first government agency in the US to adopt OSM and commit to continued improvement of the map data as the city changes or new data elements are tracked. More recently, TriMet was awarded a $678,000 grant from the Federal Transit Administration to develop a multimodal, integrated, automated, accessible and connected APPENDIX 37. The Base Adresse National (BAN) Project Interaction type Government → Public Trigger event Adoption of new government policies. Domain Topographic data (national address database). Organizations IGN, La Poste, the OpenStreetMap France association, local authorities, government agencies and the General Secretariat for Modernization of Public Action (SGMAP) through Etalab Actors IGN, La Poste, the OpenStreetMap France association, local authorities, government agencies and the General Secretariat for Modernization of Public Action (SGMAP) through Etalab Datasets Addresses Process All involved partners will contribute their existing data and, with the help of citizen contributors, the consolidated database will be kept up to date. Feedback A national address database which functions as a common reference dataset. Goal Create a complete and up-to-date address database for France which will be free and open for everyone. Side effects The creation of novel synergies between government and citizens. Impact of the project National Temporal pattern Ongoing Funding Public Contact point - APPENDIX: CASE STUDIES 121 122 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT The Base Adresse National (BAN) providing a reliable tool for their mailing lists; and emergency services and first responders, where it project can improve their efficiency. The Base Adresse National (BAN) project aims BAN offers a number of tools and services such to provide all necessary data about French as a free downloader for the address files, a addresses, facilitate access for all interested geocoder, APIs for programmable integration parties including citizens, provide online tools and management, a Local Address Counter to and services to use the recorded data and enable support municipalities in creating, identifying everyone to contribute to the improvement of and numbering the road network and an its content. The project is available from adresse. interactive map for exploring the data. Moreover, data.gouv.fr, developed by Etalab. Etalab is error reporting is under development and until responsible for the French government’s policy its completion errors can be reported to the of opening and sharing public data. The French managing bodies. government sees BAN as a project that highlights the will of the administration to create and The project’s data is available under a dual license maintain large collaborative common goods, such system: either a share-alike license (which can as addresses, to build them with open source be either the dedicated license developed by La information systems and to serve their economic Poste and the IGN or an ODbL license via the dynamics, the efficiency of public services and the diffusion made by OpenStreetMap France) and a autonomy of citizens. paid-for license for traditional customers who not wish to contribute. The BAN project solves a number of practical problems like the fact that authorities had many, Main lessons: yet different, incomplete and incompatible address databases, making it complicated, costly • High level government support can quickly and time-consuming to keep them all up to date. establish the basis for solving long-standing BAN was built from the address databases of the bureaucratic problems. participating government agencies and citizen • A collaborative mentality can lead to contributions. The data contributed have been successful projects. harmonized, reconciled and redundant names have been eliminated. Through the project’s • Public participation is crucial for mass data website, municipalities, enterprises and citizens collection and curation. can contribute and freely use this database on the principles of collaboration (share, improve and reuse a common good). The database is designed to quickly identify all addresses on French territory and also manage the 200,000 to 300,000 addresses created each year. The BAN project, apart from data curation, offers benefits to all stakeholders including municipalities, who can use it to handle their daily address management; the private sector, by APPENDIX 38. Citizen participation in urban planning, Kirtipur, Nepal Interaction type Public → Government → Public Trigger event Lack of public spaces. Domain Urban planning Organizations Kirtipur Municipality, UN-Habitat and Mojang Actors Kirtipur Municipality, UN-Habitat, Mojang, local NGOs and local communities Datasets Minecraft digital models of public spaces. Process A series of workshops for engaging young people in the geo-design of public spaces. Feedback – Goal Implement a collaborative approach to improve the functionality of public parks. Side effects Social impact on several levels.. Impact of the project Local Temporal pattern Ongoing (in other cases) Funding NGOs Contact point - APPENDIX: CASE STUDIES 123 124 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Citizen participation in urban of young people and considerably shortens the planning process. The use of the online game planning, Kirtipur, Nepal also increased youth interest in urban design in Kirtipur is a city in the Kathmandu Valley, Nepal, general, provided an efficient tool for overcoming south-west of country’s capital Kathmandu. the participation barrier to decision making, and The city faces an intense and urgent issue of helped young people to develop important skill public space availability. The public spaces of and networks. Kirtipur are under pressure because of unplanned However, it was shown that there was a urbanization and poor land administration by gender bias towards males and that digital the authorities. Public spaces play a vital role in literacy influenced individuals’ representation, the well-being of citizens, especially when they participation and voice, and thus participants lack alternative places, a common observation in with better digital skills had more influence in the poor and deprived neighborhoods. Once a city has design process. been developed, it is difficult, if not impossible, to create new public spaces or considerably alter the Main lessons: functionality of existing ones. • New technology can enable the In this context, the Kirtipur Municipality, implementation of innovative approaches supported by UN-Habitat, along with local NGOs to solve long-lasting problems, both and local communities, worked on a participatory spatial (e.g. geo-design) and social (e.g. method to revitalize public spaces. The main marginalized citizens). tool for achieving this goal was the computer • Biases due to digital literacy and gender game Minecraft, aiming to encourage youth inequality should be expected and proactive participation in urban design, as this demographic measures should be taken to address them. group is usually marginalized or excluded from decision making. The project (Block by Block) • Projects that run under the umbrella of bigger is co-managed with Mojang, the developer of efforts and initiatives (e.g. Block by Block) Minecraft. The objective of the overall project is gain from the management competence of the to develop public spaces in low income countries personnel involved. by using Minecraft as an urban planning tool. • Crowdsourcing spatial local knowledge can Minecraft was used in a series of workshops to improve government practices in terms of allow young people to produce a new design time needed and outcome produced. for an existing public space. The workshops, • Innovative participatory methods can help apart from designing a new version of the park, authoritative decision making processes to included observations of, and a questionnaire and become more inclusive. interviews with, the participants and the different stakeholders. The results showed that Minecraft enabled participants to design a more functional and useful version of the public space. This alone is important as new tools are added to existing participatory methods of geo-design. Moreover, it was shown that this method attracts the interest APPENDIX 39. Land tenure in Tanzania Interaction type Public → Government → Public Trigger event Protect human rights. Domain Land tenure Organization USAID, Cloudburst Group Actors USAID, Tanzania’s government, local authorities and citizens Datasets Parcel data and land ownership documentation. Process Trained mappers collect and document land tenure rights using mobile mapping devices in mutual agreement with owners to support government authorities. Feedback – Goal Improve the land administration system. Side effects Social improvement and promotion of gender equality. Impact of the project National Temporal pattern One-off Funding USAID Contact point - APPENDIX: CASE STUDIES 125 126 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Land tenure in Tanzania administration processes, and considerably increase the speed of data collection. By lowering The Land tenure project in Tanzania ran from the cost of land certification programs the process 2014 to 2016 as a USAID project funded with $1 becomes more inclusive and accessible, and thus million. The aim of the project was to enhance more transparent. In turn, this could minimize, or economic growth in the country by securing even eliminate, land disputes, which can improve land tenure rights since there is no properly investment opportunities. In parallel, this process functioning land administration system in could help raise women’s awareness of their right place. The existing processes are costly, poorly to own and inherit land, and help minimize social managed, and corruption and lack of transparency disputes and turmoil around land ownership. are common. This means rural dwellers and marginalized social groups, like women and However, this process requires citizen children, find it hard to secure and document contributors to have basic technical competency their rights to their land. This creates social and be comfortable using applications in tensions, poor asset exploitation and little, if any, handheld devices (e.g. smartphones and tablets) investment in land. Due to customary norms and to map and document land parcels. Otherwise, social prejudice, it is hard for women to obtain citizens need to be trained to use the technology land rights as these go through male family and applications developed for this purpose. To members or spouses. USAID launched a project increase acceptance and minimize interpersonal to map geographic and demographic data using disagreements, the local mappers also have to mobile phone technology. The Mobile Application ensure that landowners or their representatives to Secure Tenure (MAST) project enables and neighbors are present during the process. villagers to identify property boundaries and Main lessons: gather the information officials need to issue land ownership documents. • Deep-rooted social problems can be addressed through participatory and The project builds on a participatory approach crowdsourced approaches. for capturing land rights information and helps to create a collaborative and participatory • Familiarization with technology, even mentality by crowdsourcing the recording and at a basic level, is the cornerstone for documentation process, thus securing land crowdsourced efforts in the digital era. rights without excluding anyone. The land • Land tenure rights and transparent land rights recorded will be linked to databases that management processes are fundamental for Tanzania’s government can use to issue formal social development and economic growth. documentation in a more cost-effective and time Thus, up-to-date and accurate geographic sensitive manner, or serve as an independent information is a key enabler. registry of claims, thus increasing land tenure security and transparency. • A positive impact and optimum results need well-designed, managed and funded projects. This project offers multiple benefits to government and citizens alike. The participatory crowdsourced approach can help the government to change traditional and more expensive land APPENDIX 40. Open for Business, UK Interaction type Public → Government → Public Trigger event 2015 UK flooding events. Domain Mapping of business locations after a flooding event. Organization Department for Communities and Local Government (DCLG) Actors DCLG Communications and GIS staff, local businesses in the North of England Datasets Locations of businesses that were “Open for Business” after 2015 flooding. Process A simple ArcGIS web map was set up to capture the data. This was then embedded with the ArcGIS online (AGOL) crowdsourcing web map application template and a simple form created to capture the relevant data. Feedback From the outset there was excellent take up and the map provided a useful communication tool for highlighting which areas were getting back to normal after the catastrophic flooding events. Goal To inexpensively and efficiently obtain data from businesses in the north of England about which ones were Open for Business. Side effects The data will no doubt be skewed to the more digitally literate (and social media engaged) businesses and individuals. However, there is good geographical coverage. Impact of the project National Temporal pattern One-off (seasonal) Funding Public funding Contact point Simon Roberts APPENDIX: CASE STUDIES 127 128 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Open for Business, UK Only businesses comfortable with digital means and social networks could take advantage of the During December 2015 heavy rainfall led to project, but no geographic bias was observed. widespread flooding in many areas of northern There was no quality assurance of the data Britain causing disruption in businesses services. being captured but DCLG GIS staff did have the The Department for Communities and Local capability to delete fake or misleading data. Government (DCLG) aimed to inexpensively and efficiently obtain data from businesses in the Main lessons: north of England about which ones were “Open • In disaster situations authorities can be for Business”. The GIS and communications staff overwhelmed but crowdsourcing and at DCLG held several discussions regarding the participatory approaches might fill the gap. design of the mapping application that could support this aim. Given the fact that DCLG was • Crowdsourcing data can provide an accurate openly embracing ArcGIS web mapping at the image of the reality on the ground even in time of the national floods, they decided to trial cases of disaster. their crowdsourced data collection web mapping • The ability to quickly set up mapping application to capture this data. A simple ArcGIS applications that will gather all the necessary online (AGOL) map was quickly set up to capture information is fundamental to the success of the data, which played a key role in the success of the effort. the project given the urgent situation. Although the technology was not specifically designed for • Poor familiarisation with technology might a crowdsourced activity, it served its purpose cause bias in the data. perfectly. The application was then embedded with AGOL’s crowdsourcing web map application template and a simple form created to capture the relevant data. The application helped businesses to inform local communities and visitors that they were open, customers to support local businesses recovering from the flood, and tourists to see what was open in any particular area. The mapping application was supported by a communication strategy and social media feeds, and was quickly embraced by local businesses. Businesses could upload updated details regarding their re-opening or the services and products available using the online mapping application and complete a simple form including their location and opening times. In parallel, a Twitter hashtag (#OpenforBusiness) further increased visibility. APPENDIX 41. International Hydrographic Organization’s Crowdsourced Bathymetry Database (CBD) Interaction type Public → Government → Public Trigger event Safety requirements and need to improve services. Domain Hydrography, ocean mapping, navigation, marine spatial planning, physical oceanography. Organization International Hydrographic Organization (IHO) Actors IHO, National Oceanic and Atmospheric Administration (NOAA), ocean scientists, Rose Point Navigation Systems Datasets The IHO’s Data Centre for Digital Bathymetry (DCDB) Crowdsourced Bathymetry Database (CBD). Process Volunteers contribute anonymous GPS position and soundings data to a new international database managed by NOAA’s National Centers for Environmental Information (NCEI). Feedback A more informed and up-to-date bathymetric database. Goal Cover with bathymetric data as many sea areas as possible. Side effects – Impact of the project Global Temporal pattern Ongoing Funding Public Contact point LT Anthony Klemm APPENDIX: CASE STUDIES 129 130 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT International Hydrographic and effort for mariners. In the future, the DCDB plans to support other models, including individual Organization’s Crowdsourced mariner contributions. Bathymetry Database (CBD) While CBD data may not meet accuracy According to IHO, less than fifteen percent of the requirements for charting areas of critical under- world’s ocean depths have been measured; the keel clearance, it holds limitless potential for rest of the data used to compile seafloor maps myriad other uses. If vessels collect and supply is estimated depths. These estimated depths are depth information while on passage, the data can largely derived from satellite gravity measurements, be used to identify uncharted features, to assist which can miss significant features and provide in verifying charted information, and to help only coarse-resolution depictions of the largest confirm that existing charts are appropriate for seamounts, ridges and canyons. About fifty percent the latest traffic patterns. CBD can also provide of the world’s coastal waters shallower than 200 vital information to support national and regional meters remain unsurveyed. In this context, the development activities, and scientific studies in shortfall in bathymetric data is important as areas where little or no other data exists. Providing knowledge of the depth and shape of the seafloor crowdsourced bathymetry data allows hydrographic underpins the safe, sustainable and cost-effective organizations to access valuable reconnaissance execution of almost every human activity on, or data that can be used to rapidly assess the adequacy beneath, the sea. Recognizing the relevance of of nautical charts, which can lead to more targeted bathymetry to international maritime policy and and faster nautical chart updates. Finally, for the the blue economy, and noting that crowdsourced success of the project a number of factors need to bathymetry may be useful for many potential be considered. First, as in all crowdsourced projects, users of the world’s seas, oceans and waterways, the recruitment of volunteers is a challenging task. the IHO has developed a guidance document to Maintaining public engagement and interest is state its policy towards, and provide best practices another important issue. for collecting, crowdsourced bathymetry dealing with subjects such as data contribution, data Main lessons: collection, data and metadata, uncertainty and • The adoption of a high-level participatory legal considerations. strategy can considerably facilitate the All data, hosted by IHO’s Data Centre for Digital realization of crowdsourced projects. Bathymetry (DCDB), is accessible online via • Technological developments allow massive interactive web map services (https://www.ngdc. in-situ data gathering. noaa.gov/iho/). The DCDB currently accepts • In-situ data gathering and data update at the crowdsourced bathymetry contributions through a global scale cannot be achieved efficiently network of Trusted Nodes, which are organizations without citizen participation. or individuals that serve as data liaisons between mariners (data collectors) and the DCDB. Trusted • Crowdsourcing projects prove a valuable Nodes may assist the mariner by supplying data tool even in more demanding cases in logging equipment, providing technical support terms of accuracy and overall quality of the to vessels, downloading data from data loggers, datasets produced. and providing the information to the DCDB. The DCDB works with each Trusted Node to standardize metadata and data formats and define data delivery requirements. This model standardizes data contributions and minimizes the requirements APPENDIX 42. Disaster Management, Early Warning and Decision Support Capacity Enhancement Project in Indonesia Interaction type Public → Government → Public Trigger event - Domain Disaster management Organizations Humanitarian OpenStreetMap Team (HOT), United States Agency for International Development (USAID), Office of US Foreign Disaster Assistance (OFDA), Pacific Disaster Centre (PDC), Massachusetts Institute of Technology (MIT), Badan Nasional Penanggulangan Bencana (BNPB) Actors Local communities and Indonesian government Datasets Infrastructure data Process - Feedback Datasets that can be used for disaster management. Goal The project aims to enhance the capacity of national and provincial disaster management agencies to access automated international, regional, national and local hazard information; support infrastructure data; share information between agencies, and disseminate alert warnings to communities and populations at risk. Side effects – Impact of the project National Temporal pattern One-off (year-long project) Funding NGOs Contact point - APPENDIX: CASE STUDIES 131 132 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Disaster Management, Early up the process of data collection are mapathons and mapping parties. Both of these have been Warning and Decision Support used by the project’s partners to gather as much Capacity Enhancement Project information as possible. Once the information in Indonesia is mapped by the data entry specialists, it will be reviewed and validated by a team of dedicated The Disaster Management, Early Warning data quality specialists. and Decision Support Capacity Enhancement Project focused on the two major cities of During the lifespan of the project, particular Indonesia, Surabaya and Jakarta. Due to local attention was paid to user engagement, training particularities, issues and difficulties in mapping and capacity building of local communities might stem from dense settlements, surveying and volunteers. For example, mapathons were permits, low-resolution imagery, inaccurate GPS open not only to university students but to all points, weather conditions, ambiguity in road volunteers and citizens. A collaboration with a classifications and overlapping boundaries. The local radio station helped to drum up support project’s overall aim is to enhance the capacity and urge citizens to participate in collaborative of national and provincial disaster management mapping, and socialisation kick-off workshops agencies to access automated international, with BNPB and BPBD took place in Jakarta and regional, national and local hazard information Surabaya. The HOT team was actively involved and support infrastructure data, as well as share in training local BPBD staff and aspiring youth information between agencies, and disseminate and university students. A structured university alert warnings to communities and populations curriculum had been developed from HOT at risk. Indonesia’s previous experience. The program is based and focuses on The OSM data collected will be used by the development of InAWARE, a disaster Indonesia’s Disaster Management Agency, BNPB, management tool, to improve overall risk as well as its sub-national agencies BPBD DKI assessment, early warning, and disaster- Jakarta and East Java to enhance its real-time management decision making in Indonesia early warning and decision making support with (http://inaware.bnpb.go.id/inaware/). The project the use of InAWARE. In addition, OpenStreetMap needs to develop a geospatial database for and its exposure data provides the basemap for disaster risk, which will include administrative Peta Bencana, a project initiated by the Urban boundaries, building footprints, road networks Risk Lab at MIT as a free platform for emergency and disaster vulnerability characteristics. Data response and disaster management in megacities collection processes include remote mapping in South and Southeast Asia. As part of the using OSM’s Tasking Manager for buildings project, the team will also distribute print-outs of footprints; in-situ raw data gathering and the areas mapped as a token of appreciation of the feature attribute enrichment with the use of collaborative efforts of volunteers. the OpenMapKit and GeoDataCollect (GDC) applications’ close collaboration and consultation with local authorities; importing existing datasets; and training and collaboration with local universities. Common approaches to speed APPENDIX Main lessons: • The existence of efficient tools can considerably enhance collaborative efforts. • The use of OSM and up-to-date satellite imagery can provide basemaps for the collection of specific data. • The collaboration and active involvement of experienced partners considerably helps the success of a project. • Volunteer engagement, training and capacity building are fundamental for the success and continuation of a project. APPENDIX: CASE STUDIES 133 134 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 43. Summer Camp Guide Interaction type Public → Government → Public Trigger event County GIS staff members became involved with web mapping development. Domain Public and private services Organization Tompkins County, NY Actors Ithaca Times newspaper, Tompkins County Youth Department, Tompkins County ITS – GIS Division Datasets GIS point data with attributes, links to websites and photos. Process Summer camp managers upload details and locations of their facilities, which citizens can search to find what suits them best. Feedback Summer Guide Web Application enables young people, parents and guardians to easily search and find concise information with links to additional information by age group and keyword. Goal Create a user-friendly web interface with searchable web applications and maps to benefit summer camp agencies, the county’s youth services, and community. Side effects Transformed the paper copy of the county’s Summer Camp Guide into an all-encompassing user-friendly web interface featuring searchable web applications and maps. Impact of the project Local Temporal pattern One-off (seasonal) Funding Public Contact point Svetla Borovska APPENDIX Summer Camp Guide However, a small number of summer camp managers did not have internet access, so a staff A collaborative, crowdsourced-driven effort has member from the youth services department had been supported by Tompkins County, NY, US, to complete the GeoForm for them. Moreover, around the county’s Summer Camp Guide. The a number of possible improvements have been aim was to create a user-friendly web interface identified in the first version of the application, with searchable web applications and maps to and it only works with point-based layers that benefit summer camp agencies, the county’s contain the locations of available summer camps youth services, and the community by providing with their attributes, links to websites and photos. updated information for all summer programs and camps for 2017 in and around Tompkins Main lessons: County. The project has been supported by local • Simple yet handy web mapping applications actors and stakeholders. The application consists can considerably improve the level of public of a data collection GeoForm that enables services provided. summer camp managers to describe and record their facilities on a map; an informative website • “Professional” crowdsourcing can generate with all the necessary details for the public and a critical mass of data to fuel a new way of a web mapping application that enables the disseminating information. public to search, find and locate the most suitable • A handful of champions inside a public summer camp for their needs. authority can change and improve its culture The application transformed the way citizens and methodologies. are informed as it provides an alternative to the • Given the availability of technological paper copy of county’s Summer Camp Guide solutions, a small investment of time and through an all-encompassing user-friendly web effort in crowdsourced mapping applications interface featuring searchable web applications can create the basis for innovative solutions and maps. Moreover, the application reduced the for public services. workload and turnaround time for the county’s youth services staff; gave summer camp agencies more control (and responsibility) in presenting their business and enabled young people, parents and guardians to easily search and find concise information with links to additional information (webpages belonging to agencies providing camps and programs) by age group and keyword. While developing the web application, various old and longstanding issues were solved using GIS technology and crowdsourcing all of the mapping and background information about the summer camps, which at the same time upgraded public sector services. APPENDIX: CASE STUDIES 135 136 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 44. Ramani Huria Interaction type Public → Government → Public Trigger event The fastest growing city in Tanzania has a high level of flood risk due to rapid, unplanned urban development. Domain Community mapping Organization World Bank Actors World Bank, Red Cross, Tanzanian Commission of Science and Technology, University of Dar es Salaam, Ardhi University, Buni Innovation Hub, Tanzania Data Lab, City Council of Dar es Salaam Datasets Infrastructure data contributed to OpenStreetMap; aerial imagery from UAVs added to OpenAerialMap. Process University students and community members are trained to provide necessary data for resilience and disaster reduction. The content generated is freely available and is used in software that produces realistic natural hazard impact scenarios to inform better planning against natural hazards. Selected community members also become Red Cross disaster risk responders/volunteers, supporting their wider community with risk mitigation measures and support. Feedback Better planning, preparedness and response for disaster. Goal Collect data and generate an understanding of flood-prone areas for flood resilience across Dar es Salaam. Side effects Launch of the project to other areas: Bukoba in Tanzania, Mozambique, and Zanzibar. Impact of the project National Temporal pattern Ongoing Funding NGOs Contact point Mark Iliffe APPENDIX Ramani Huria resource-constrained environment. Dar es Salaam is Africa’s fastest growing city However, data coverage was not as full as and Tanzania’s primary city with a population initially required. Reasons for this include poor of around 5.5 million and is projected to be a quality assurance and control coupled with the megacity by 2030. The city’s urbanization process complexity of tools (and potentially language is largely unplanned and its pace challenges challenges for non-native English speakers). traditional methods of planning and public Further iterations of Ramani Huria will include service provision. Every year during the rainy research components to understand and mitigate season, Dar es Salaam suffers from devastating these challenges, while expanding the geographic floods. In response to this, Ramani Huria was scope of data collection across Tanzania’s created as a community-based mapping project urban environments. in Dar es Salaam, training university students Main lessons: and local community members to create accurate maps of the most flood-prone areas of the city. • The development of a project as part of The project is now being transferred to other a larger effort and support from strong parts of East Africa. institutions play a major role in success. There are two key factors contributing to the • Project management oriented to capacity success of Ramani Huria. The first is how the building facilitates project sustainability project has been administered and pushed for by as it can be passed from the initiators to stakeholders in the local government. It began in local communities. 2011 as a World Bank project, which has slowly • The use of open source software frees projects moved towards a collaborative governance model, from unnecessary costs that could hamper with all stakeholders currently in the process of their development and allows for diffusion in forming committees to execute and administer local communities. the project. This has changed the dynamic away from a World Bank-driven project, towards a • Interconnection among similar community-driven one. This has engendered a projects can spread best practice and deeper collaborative mentality between outsiders effective methodologies. (WB, Red Cross), government, universities, and community members, as they all feel that this is their project and that they have the capacity and ability to shape and make decisions. This negates issues of acceptance and supports a wider re-use of the data. The second key factor is the nature of the open platforms and tools used by Ramani Huria (OSM, InaSAFE, QGIS, etc.), which allows for benefits to be achieved at scale and shared throughout a global community. Improvements to software made in Indonesia can and have been applied in Tanzania, and vice versa. This allows for more intelligent use of time and money in a APPENDIX: CASE STUDIES 137 138 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 45. Natural Resources Canada-OpenStreetMap Synergy Interaction type Government → Public → Government Trigger event The proven inability of the NMA to keep datasets up to date along with the familiarisation of its personnel with OSM data, quality and processes. Domain Update of national topographic database. Organization Mapping Information Branch (MIB) at Natural Resources Canada (NRCan) Actors OSM Community, MIB, NRCan Datasets Canvec (mainly the road network). Process NRCan releases its database into .osm format. The data is imported into OSM and updated/modified by the OSM community. NRCan regularly compares OSM datasets with its own database as a change detection mechanism to keep its database up to date. Feedback Change detection datasets that were verified in the field by NMA employees. Goal Keep national databases up-to-date. Side effects OSM data cannot directly be used by the authorities due to IPRs and licensing issues. Impact of the project National Temporal pattern One-off Funding Public Contact point – APPENDIX Natural Resources Canada- Canadian authorities that needs to be solved. OpenStreetMap Synergy The project stopped when the person championing the project retired from NRCan. The role of the Mapping Information Branch The team was gradually reallocated to other (MIB) at Natural Resources Canada (NRCan) projects deemed of higher priority and was is to provide accurate geographic information never restarted, despite the fact that most map of landmass at the scale of 1:50000. This task content provided by NRCan (buildings and their translates into the need to update 13,200 functions, land use, power lines, tracks and trails, map-sheets. Taking into account the results etc.) has not been updated since the mid 1990s. of ongoing research regarding VGI quality, Today, NRCan uses other authoritative data to Canadian authorities experimented with the update some of its own data (road, railway and OSM community. The first step was made by water networks). the Centre for Topographic Information in Sherbrooke (CTIS) which released the Canvec Main lessons: product (digital topographic map of Canada) in .osm format. This move further enabled • VGI projects can be triggered by government the integration of Canadian authoritative data need to provide up-to-date and accurate into OSM and gave the OSM community the topographic maps. opportunity to interact with it. The next step was • Updates from OSM could have been used for to regularly compare the OSM database with the change detection to update authoritative data. Canvec data to pinpoint differences, which were treated as potential changes and verified in the • Differences in data structure and licensing field. hindered progress. • The role of championing personnel inside On the positive side, the work of keeping the government agencies is paramount, especially datasets up to date has been facilitated by the in the first stages of a project. OSM community. Leveraging the OSM crowd- sourcing mechanism the Canadian authorities • Poor support for long-term innovative ideas have developed a change detection process from higher level managers can damage the which helps concentrate resources and effort on development of successful projects. potential changes. This is important when one takes into account that not all original spatial entities in the authoritative database have been updated. Challenges of the project included the compatibility of the two datasets (in terms of semantic and attribution differences), the lack of metadata for OSM data and the differences in coverage (OSM mainly covers urban areas compared to uniform authoritative coverage), which needed to be addressed. Moreover, there is a conflict between the license and use terms of OSM and the intellectual property rights of APPENDIX: CASE STUDIES 139 140 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 46. Participatory mapping and decision support tools for disaster risk reduction, the Philippines Interaction type Government → Public → Government Trigger event The Department of Interior and Local Government (DILG) in the Philippines wanted to better support local government units (LGUs) to prepare risk- sensitive land-use plans, structural audits of public infrastructure and disaster contingency plans. Detailed data to be used for planning was not available. Domain Generic mapping by local government. Organization Department of Interior and Local Government (DILG) Actors World Bank - East Asia Pacific (EAP), Environment Science for Social Change (ESSC), DILG, LGUs of Pampanga, Project NOAH, OpenStreetMap (OSM) Philippines Community, GeoRepublic Japan, Humanitarian OSM Team (HOT) Datasets OSM data describing standard features, land use and administrative boundaries. Process New training materials were created specifically aimed at LGUs in the Philippines. Next, training and mapping activities were conducted in the selected LGUs. Further community building activities to support the growth of the OSM community in the Philippines were also conducted. Key to this was inviting community members to all the training sessions, either to learn or assist in delivering the training. Feedback Success in the data collection in OSM varied by participating municipality. In two municipalities, little data collection occurred after the initial training and mapping phase, but one continued to map afterwards. Goal Collect detailed data to better support LGUs in developing disaster plans. Side effects - Impact of the project - Temporal pattern - Funding - Contact point Kate Chapman, HOT APPENDIX Participatory mapping and Philippines to provide support in OSM. Creating and discovering local champions in LGUs will decision support tools for disaster help to ensure greater support for mapping risk reduction, the Philippines after training. Using these methodologies in other places is certainly possible, but it would The goal of this project was to use participatory be important to take into account the context mapping and InaSAFE impact modeling software and government structure to ensure support for to support Philippines LGUs in disaster risk mapping outside of initial training. reduction activities. DILG needed to better support such activities and determined that Main lessons: participatory mapping and use of InaSAFE was the best way to accomplish this. • It is important to highlight the benefit of mapping outside of the immediate context, to The project provided training and technical encourage continued work and data collection assistance to LGUs to create basemap after the project period concludes. information and then perform impact analysis using InaSAFE. Three LGUs participated in • Efforts to build broader VGI communities as the program from the province of Pampanga part of focused data collection activities can be (Candaba, Lubao and Guagua). Eighty-five local valuable and should be encouraged. people participated alongside six OSM volunteers • With care, VGI data can be an important who also assisted the training team. Initial complement to official data sources and used training was about the collection of infrastructure for scientific modeling. data using OSM. Later InaSAFE workshops were held to show how the collected data could be analyzed. The intention was for LGUs to continue mapping in OSM to improve the analysis in InaSAFE and create a more detailed basemap for other uses. Freely available data of the areas are now available in OSM. During training, the OSM Philippines community assisted with the goal of further strengthening and supporting the OSM community as a whole. Teaching LGUs how to map in OSM and then analyze the data in InaSAFE was of benefit, though it was difficult for LGUs to continue mapping after their initial workshops. Little mapping occurred outside of the scheduled mapping activities during the training. Only one LGU continued to map afterwards. New methods for better supporting mapping are being explored, including creating a Training of Trainers (ToT) program to expand the community in the APPENDIX: CASE STUDIES 141 142 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 47. Government open data usage in Lithuania Interaction type Government → Public → Government Trigger event - Domain Generic mapping Organization OpenStreetMap Lithuania Actors OSM, GIS-Centras, National Land Agency, Road Administration, Vilnius city municipality Datasets Road network, places, water bodies, public transport, Vilnius addresses. Process Government openly publishes data which is downloaded and compared to OSM data. Problems are fixed in OSM or reported to appointed government communication point. Feedback Improved data and update speed on all sides. Goal To improve data quality, data completeness and update speed. Side effects - Impact of the project National Temporal pattern Ongoing Funding - Contact point Tomas Straupis (association Atvirasis žemėlapis) APPENDIX Government open data usage Official conferences with the participation of both government institutions and open data (OSM) in Lithuania presenters were important in establishing the Lack of open data in generic mapping and of project. They boosted knowledge about open government data on some topics (e.g. tourism) data and created personal and institutional links. was the principle reason for the OSM team taking Without personal meetings, communication was action in Lithuania. In Lithuania, although there close to zero. It was also important to find at least was some publicly accessible data, almost none of one person in each agency, who would champion it was open and there was little or no knowledge the open data cause and explain how both sides about open data and open data projects in the would benefit. Lithuania is a small country and in government sector. The aim was to update this context the interaction between OSM and the authoritative spatial datasets by combining government worked well. conventional and open source data. Public sector Main lessons: and NGO/international organizations worked together to fill this gap, and successfully achieved • Collaboration between open data communities appropriate cooperation between the public and government may result in the mutual and government. improvement of datasets. The project started in 2015 and Lithuanian OSM • Conferences can raise awareness about open members first approached different government data and develop networks. agencies with requests for open data to start the • Institutional open data champions play a vital project. Once legal and knowledge issues were role in government agencies. solved, data began to be released. Government contact points were assigned and feedback allowed the corresponding government datasets to be updated. The OSM community corrects OSM errors, and GIS-Centras corrects the government data. Thus, the integration between OSM and government is done by synchronisation – checking for differences and taking mostly manual action on both datasets. Among the positive outcomes were that knowledge about open data increased and data quality and update speed of both OSM and government datasets improved. However, one institution (Registrų Centras), which holds the national dataset of addresses, was unwilling not only to share their data, but even to communicate with the OSM community. APPENDIX: CASE STUDIES 143 144 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 48. Xalapa collaborative transport mapping Interaction type Public → Government Trigger event Absence of information regarding public transportation routes. Domain Transportation data Organization Codeando Xalapa Actors Codeando México, “Buen Gobierno” Office of Municipal Government of Xalapa, WRI Mexico, OpenStreetMap, Telenav-OpenStreetCam, Mapillary, Computer Science and Administrative Computer Systems School of the University of Veracruz Datasets General Transit Feed Specification (GTFS) feeds, passenger counts, public transit routes. Process Meetings attended by all stakeholders, including representatives of government, transport companies, citizens and specialists in mobility, to document existing routes; bus schedules, costs and areas covered analyzed; data collected, and participants trained to use the mobile application developed; publicity campaigns disseminate the project through the media; datasets collected are verified and published. Feedback Detailed transit routes for further use and transportation planning. Goal To collect public transportation datasets. Side effects The realization that spatial data requires regular updates and thus established processes are needed to keep them up to date. Impact of the project National Temporal pattern Ongoing Funding Funding was provided by the Office of Municipal Government of Xalapa and the OpenStreetView/ Telenav while support in equipment was provided by Mapillary. Contact point mapaton.org APPENDIX Xalapa collaborative the information was verified, participants were informed, route data was cleared and the datasets transport mapping were published. More than 200 routes were In Xalapa, as in most Mexican cities, public collected (almost 100km), which can be used on transportation is governed by private companies. official platforms by the federal government as However, the transport itineraries are not fully well as by open civic platforms. known, and thus the 480,000 inhabitants, 45 The project received monetary support from the percent of whom are workers and 30 percent are Office of Municipal Government of Xalapa, which students, cannot effectively program their daily was used to sponsor users’ travel costs when mobility. In this context, the public sector, NGOs tracing and collecting information to create a feed and international organizations, universities of the route. A contribution was received from and the local government joined forces to OpenStreetView/Telenav, which was used to buy overcome low resources and build the data fuel and map routes from a private car. Finally, infrastructure needed to upgrade public sector Mapillary provided basic equipment (e.g. selfie transportation services. The main objective is to sticks) for taking photographs. The project gained collect a public transportation dataset through support from more than 300 citizens who helped citizen participation. to trace their routes from start to finish while the The project was led by Codeando Xalapa Mapillary application helped to upload 18,400 (codeandoxalapa.org) in collaboration with photographs of the routes. the local government. Together they began a One negative aspect of the project is the difficulty process to collect public transportation routes of involving private companies active in the called Mapatón Ciudadano (mapaton.org). The public transportation business. The absence of work was strengthened with the participation, federation between these companies multiplies support and input of entities such as WRI Mexico the number of partners and makes collaboration (wrimexico.org), OpenStreetMap and Telenav. difficult. Another challenge is the lack of An open source mobile application (transitwand. continuity, as authoritative support is diminishing com) was customized and improved to perform and there is no clear actor to take on the role of dataset collection. However, before tracking promoting this process. Finally, datasets need routes, several meetings were organized with constant updating, particularly in an unregulated all stakeholders including representatives of the transport system, which represents a crucial government, transport companies, citizens and challenge that must be addressed effectively. specialists in mobility. The aim was to identify where routes start and end, the schedules of Main lessons: the buses, the cost and the geographical areas covered. Then, the data collecting process • Constant funding support is needed started with parallel training of participants on throughout a collaborative project. how to use the mobile application. At the same • Leadership of the project is important both time, part of the team participated in publicity during the first – enthusiastic – phases and for campaigns to disseminate the project in media future support and development. such as radio and television. During the mapping process, teams of citizens were created to gather • Data updates should be considered as information on routes close to their home, work important as data capture. Continuity of or school. All information was stored in a central participation and public support is crucial to repository. After the data collection process, keep data up to date. APPENDIX: CASE STUDIES 145 146 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 49. FloodTags Interaction type Public → Government → Public Trigger event Witnessing an Indonesian NGO manually transcribe tweeted data onto a map to inform disaster response actions. Domain Disaster management Organizations Deltares, Radboud University, VU University Amsterdam, Universitas Padjadjaran Actors Citizens, ICT research community COMMIT and Topconsortia Kennis en Innovatie (TKI), the Red Cross Climate Centre, Deltares, Radboud University, VU University Amsterdam, World Bank, United Nations Datasets Data from social media (Twitter, Instagram, blogs) plus SMS and online news media combined with hydrologic models. Process Messages shared on social media during a developing flood event are used to map flood risk. Feedback Flood risk maps that can be lifesaving. Goal Create an efficient solution to capture risk information generated on social media to help improve preparedness and response actions. Side effects - Impact of the project National Temporal pattern Ongoing Funding Social enterprise Contact point - APPENDIX FloodTags be used to speed up recovery work after a disaster or to identify flood-prone areas that FloodTags is a social enterprise located in The would benefit from protection. The project can Hague (Netherlands) which brings together a improve reporting during disasters and, in turn, large number of researchers from universities support timelier and better informed response and other institutions to develop and implement and preparedness actions. Ultimately this will data analysis scripts for water management. help to improve outcomes for those vulnerable The project invests in the power of citizens and to disasters. The next steps would be to link communities as key resources in the solution to the existing social media analysis to national many water management problems. forecasting tools. Indonesia was selected as a case study for the However, since crowd-based observations tend to project due to its disaster-prone location but be rough estimates, the data need to be filtered, also because the local population is familiar enriched, validated and transformed into easily with, and active in, the use of social media. interpretable maps that can be used by disaster The project developed a process that combines managers. When the researchers compared their natural language processing and flood modeling results with photographs of the Jakarta floods to improve risk information. The project was at more than 100 points, they found they had inspired after witnessing an Indonesian NGO modeled the floods correctly in around two thirds manually transcribe tweeted data onto a map of them and in three quarters of the districts. The to inform disaster response actions public to project aims to minimise false positives while various social media outlets. When a disaster allowing for a fair amount of false negatives. (e.g. a flood) happens, people increasingly find This is the reverse approach to what FloodTags out about it from social media, as individuals and had been doing so far (i.e. using all observations, groups use the internet and social media to spread ending up with a fair amount of false positives). the word. For example, after the Indonesian capital Jakarta was hit by floods in February 2017, Main lessons: related Tweets peaked at almost 900 a minute, • Crowdsourced geographic information can with a significant number including information come in many formats and from various about location and water depth, according to a sources. The development of appropriate joint study by two Dutch organizations, Deltares methodologies can turn such unstructured and Floodtags. Spontaneous Tweets about major data into valuable geographic products. floods are being turned into a mapping tool that could be used by emergency services and disaster • Technical limitations can occur during the response teams to save lives and provide aid, first steps of a project but these can be Dutch researchers said. addressed once the proof of concept has been established. The project presents the information in standardized maps, tables and graphs to effectively support response and preparedness procedures. Thus, anyone using the “event server” can obtain timing and location of new floods as observed via Twitter and take appropriate and effective measures. The maps could also APPENDIX: CASE STUDIES 147 148 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT 50. Open Cities, Sri Lanka Interaction type Public → Government Trigger event Disaster-prone area Domain Disaster preparedness Organization OpenDRI Actors World Bank/GFDRR, United Nations Development Programme (UNDP), local authorities, local universities, HOT Datasets Building footprints, road network, key infrastructure, street names, road conditions, land uses and other crucial data for disaster management. Process An OSM expert trained 30 students in initial data collection and the 10 most promising ones were further employed to complete the data capture and help train district secretariat staff. Feedback A rich dataset with all the necessary elements that could be used for risk preparedness. Goal The goal of the project was to create useful data for risk preparedness and equip the government to use and maintain it. Side effects - Impact of the project Local, but also served as a paradigm for other cases as part of the Open Cities project (international) Temporal pattern Ongoing Funding World Bank/GFDRR, OpenDRI Contact point - APPENDIX Open Cities, Sri Lanka maintain the data. The effort is part of the Open Cities project, which aims to learn from these Since 2000, flood and drought events have experiences to create a scalable approach to cumulatively affected more than 13 million understanding urban challenges and disaster risk people across Sri Lanka. Nearly $500 million in in South Asian cities. unplanned expenditure resulting from flooding in 2010 and 2011 has strained government At the project outset, the OSM expert recruited budgets and required reallocation from other and trained a team of 30 students from the planned development priorities. In this context, Eastern University to lead an intensive, 30-day in Batticaloa city successive disaster exposure- remote (satellite-based) data collection exercise mapping exercises gradually built a constituency across the six priority DSs and then the rest of for larger institutional adoption of crowdsourced the district. Visible buildings, roads and land VGI data and tools in the provincial government. uses were all digitized in this way. After the OpenDRI first supported a tsunami exposure initial push, the 10 most promising students survey in the Manmunai North Divisional were retained for a further two months to create Secretariat (DS) in 2013. Forty-eight recent the remaining data, validate results and assist in university graduates collected building footprints training DS staff. and basic characteristics for all 30,000 buildings These staff were trained on Field Papers and the in the area, including number of floors, usage and iD editor, then deployed across the six priority construction materials of walls and roofs. DSs to collect key infrastructure, street names, Mappers used what would become the baseline road conditions, land uses and other crucial data toolkit for exposure mapping: heads-up, remote for disaster management, transport planning and digitization into OpenStreetMap via freely normal business operations. By the end of the available Bing imagery; follow-up field surveys project, over 160,000 buildings and 500km of using Field Papers; and finally, office-based data roads has been digitized. validation. A series of maps was developed from The goal of the project was not merely to create the data and DS staff were trained to download, useful data but equip the government to use and update and analyze the resulting OpenStreetMap maintain it. To do this, it collaborated with an data. Since that time DS government staff have ongoing, multi-year UNDP project to upgrade used the OSM data to maintain building addresses the GIS capacity of the Batticaloa district and and update local roadways. DS governments. The UNDP saw an opportunity Inspired by the success of the original project and to create more and better data for newly trained its scaled-up successor in Gampaha province, the officers to use with their new skills; the project Government Agent (GA) of Batticaloa District team saw an opportunity to make their work asked the World Bank/GFDRR to support field- sustainable. Staff were trained to collect, use based data collection for a further six disaster- and update OSM data. Training covered OSM vulnerable DSs and satellite-based data collection editing software, field data collection methods, for the entire district. This was supported by downloading OSM data and using it with the an OSM expert working in partnership with industry-standard ArcGIS software provided by the United Nations Development Programme the UNDP. At the project’s conclusion, a series of (UNDP), preparing local universities and the thematic maps was prepared by the newly trained DS staff to collect, validate and ultimately DS staff for the six priority DSs. The data is now APPENDIX: CASE STUDIES 149 150 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT freely available on OSM and in the government geospatial data sharing platform RiskInfo (www. riskinfo.lk). Main lessons: • Previous experience on crowdsourced projects can greatly facilitate the adoption and development of new efforts, both at community and authority level. • Funding and active support by international (and experienced) organizations considerably contributes to the success of the project. • Being part of a greater project means smaller projects can benefit from the experience developed so far. References and sources Aronoff, S. Geographic Information Systems: a management perspective. WDL businessofgovernment.org/report/using-crowdsourcing-government. Publications: Ontario, 1989. Brain Off, 2010. “Haiti OpenStreetMap Response.” Accessed 8 December Atkinson, G. M. and Wald, D. J. “‘Did You Feel It?’ intensity data: 2013. http://brainoff.com/weblog/2010/01/14/1518. A surprisingly good measure of earthquake ground motion.” Brown, M. and Heaton, L. 2011. “New Country in the Making: Building Seismological Research Letters, 78 (2007): 362-368. a Map for South Sudan.” Accessed 28 April 2014. http://www. Ball, M. 2013. “New York City Opens More of City’s Data, Including enoughproject.org/blogs/new-country-making-building-map-south- OpenStreetMap Collaboration.” Accessed 22 December 2013. http:// sudan. informedinfrastructure.com/6088/new-york-city-open-more-of-the- Castellote, J., Huerta, J., Pescador, J. and Brown, M. “Towns Conquer: A citys-data-including-openstreetmap-collaboration/. Gamified Application to Collect Geographical Names (vernacular Barron, C., Neis, P. and Zipf, A. “A comprehensive framework for intrinsic names/toponyms).” Proceedings of AGILE 2013 in Leuven, Belgium, OpenStreetMap quality analysis.” Transactions in GIS (2014). 2013. Accessed 16 April 2014. http://eprints.nottingham.ac.uk/2131/. doi:10.1111/tgis.12073. Chapman, K., Wibowo, A., Nurwadjedi. 2013. “Filling the Data Gap with Barth, A. 2013. “New York City and OpenStreetMap Collaborating Through Participatory Mapping for Effective Disaster Preparedness.” Accessed Open Data.” Accessed 22 December 2013. https://www.mapbox.com/ 10 December 2013. blog/nyc-and-openstreetmap-cooperating-through-open-data/. City of New York. n.d.a. ”About NYC Open Data.” https://data. Beaulieu, A., Begin, D. and Genest, D. “Community Mapping and cityofnewyork.us/stories/s/About-This-Site/9v8v-e23e/. Government Mapping: Potential Collaboration?” Symposium of ISPRS ———. n.d.b. ”Open Data Law - DoITT.” https://www1.nyc.gov/site/doitt/ Commission I, Calgary, Canada, June 16-18, 2010. initiatives/open-data-law.page. Begin, D. “Towards Integrating VGI and National Mapping Agency Cohn, J. P. “Citizen science: Can volunteers do real research?” Bioscience Operations - A Canadian Case Study.” The Seventh International 58, no. 3 (2008): 192–197 . Conference on Geographic Information Science, 2012. Accessed 15 February 2014. http://web.ornl.gov/sci/gist/workshops/2012/ Deffner, R. 2013. “Mongolia, mapping Ulaanbaatar.” Accessed 12 April documents/Begin,%20Daniel%20-%20Paper.pdf. 2014. http://hot.openstreetmap.org/projects/mongolia_mapping_ ulaanbaatar. Berdou, E. 2011. “Mediating Voices and Communicating Realities. Using information crowdsourcing tools, open data initiatives and digital EIONET. 2012. “Corine Land Cover 2006.” Accessed 18 December 2013. media to support and protect the vulnerable and marginalised.” http://sia.eionet.europa.eu/CLC2006. Accessed 16 April 2014. http://mapkibera.org/wiki/images/7/78/ Fowler, A., Whyatt, J.D., Davies, G. and Ellis, R. “How Reliable Are IDS_MediatingVoices_FinalReport.pdf. Citizen-Derived Scientific Data? Assessing the Quality of Contrail Brabham, D. 2013. “Using Crowdsourcing in Government. Observations Made by the General Public.” Transactions in GIS, 17, no. Collaboration Across Boundaries”. IBM Center for The Business of 4 (2013): 488–506, doi:10.1111/tgis.12034. Government: Washington, DC. Accessed 16 April 2014. http://www. GFDRR. 2013. “Preparing Communities through Understanding Risk.” REFERENCES AND SOURCES 151 152 CROWDSOURCED GEOGRAPHIC INFORMATION USE IN GOVERNMENT Accessed 10 December 2013. http://www.gfdrr.org/sites/gfdrr.org/files/ Maron, M., 2007. OpenStreetMap. “A disaster waiting to happen.” Pillar_1_Preparing_Communities_through_Understanding_Risk_0. Accessed 8 December 2013. http://www.slideshare.net/mikel_maron/ pdf. openstreetmap-a-disaster-waiting-to-happen. Gollan, J., Lobry de Bruyn, L., Reid, N. and Wilkie, L. “Can Volunteers Marras, S. 2012. “Kibera, Mapping the Unmapped.” Accessed 17 April 2014. Collect Data That Are Comparable to Professional Scientists? A http://www.djemme.com/docs/docs-schede/KiberaEN.pdf. Study of Variables Used in Monitoring the Outcomes of Ecosystem McLaren, R. A. Crowdsourcing Support of Land Administration. A New Rehabilitation.” Environmental Management 50, no. 5 (2012): 969–78, Collaborative Partnership between Citizens and Land Professionals. doi:10.1007/s00267-012-9924-4. London: Royal Institute of Chartered Surveyors (RICS), 2011. Goodchild, M.F. “Citizens as sensors: the world of volunteered geography.” Meyer, R. 2013. “How Online Mapmakers Are Helping the Red Cross Save GeoJournal 69, no. 4 (2007): 211-221. Lives in the Philippines.” Accessed 22 December 2013. http://www. Hagen, E. 2010. “Putting Nairobi’s Slums on the Map. Final Report.” theatlantic.com/technology/archive/2013/11/how-online-mapmakers- Accessed 18 April 2014. https://openknowledge.worldbank.org/ are-helping-the-red-cross-save-lives-in-the-philippines/281366/. bitstream/handle/10986/6072/deor_12_1_41.pdf?sequence=1. The National Map. 2014. “Home.” Accessed 27 April 2014. http:// ———. 2012. “Workspaces: The changing environment of infomediaries/ nationalmap.gov/. Map Kibera.” Accessed 19 April 2014. http://wiki.ikmemergent.net/ The National Map, US Topo. 2014. “US Topo Quadrangles — Maps for index.php/Workspaces:The_changing_environment_of_infomediaries/ America.” Accessed 27 April 2014. http://nationalmap.gov/ustopo. Map_Kibera. OpenSource.com. 2010. “OpenStreetMap Haiti.” Accessed 8 December Haklay, M. 2010. “Usability of VGI in Haiti earthquake response.” Accessed 2013. https://opensource.com/osm. 8 December 2013. http://www.slideshare.net/mukih/usability-of-vgi-in- haiti-earthquake-response-preliminary-thoughts. OpenStreetMap. 2014. “WikiProject Mongolia.” Accessed 6 June 2014. http://wiki.openstreetmap.org/wiki/WikiProject_Mongolia. ———. “Citizen science and volunteered geographic information: Overview and typology of participation.” In Crowdsourcing geographic ———. 2013. “WikiProject Indonesia.” Accessed 16 December 2013. http:// knowledge. Springer Netherlands, 2013. wiki.openstreetmap.org/wiki/WikiProject_Indonesia. Heymann, Y., Steenmans, Ch., Croissille, G. and Bossard. M. 1994. “Corine ———. 2012. “WikiProject Corine Land Cover.” Accessed 18 December Land Cover. Technical Guide.” Luxembourg: Office for Official 2013. http://wiki.openstreetmap.org/wiki/WikiProject_Corine_Land_ Publications of the European Communities. Cover. Humanitarian OpenStreetMap Team. 2013. “Haiti.” Accessed 8 December OpenStreetMap Indonesia. 2014. Home. http://en.openstreetmap.or.id/. 2013. http://hot.openstreetmap.org/projects/haiti-2. Accessed 4 January, 2014. ———. 2013. “Indonesia.” Accessed 15 December 2013. http://hot. ———. 2012. “Community Mapping for Exposure in Indonesia. Project openstreetmap.org/projects/indonesia-0. Report.” Accessed 7 December 2013. https://openstreetmap.id/docs/ Community_Mapping_for_Exposure_in_Indonesia_EN.pdf. Huxhold, W. An introduction to Urban GIS. Oxford and New York: Oxford University Press, 1991. ———. 2012. “Evaluation of OpenStreetMap Data in Indonesia. Final Report.” Accessed 10 December 2013. https://openstreetmap.id/docs/ iRevolution. 2011. “Crowdsourcing Satellite Imagery Analysis for Somalia: Final_Report-OSM_Evaluation_in_Indonesia_2012.pdf. Results of Trial Run”. Accessed 25 May 2014. http://irevolution. net/2011/11/01/crowdsourcing-unhcr-somalia/. Roberts, S. 2014. “Open Cities and Crisis Mapping.” Accessed 8 December 2013. https://blog.usaid.gov/2014/01/open-cities-and-crisis-mapping/. ———. 2011. “Crowdsourcing Satellite Imagery Analysis for UNHCR- Somalia: Latest Results.” Accessed 25 May 2014. http://irevolution. Roberts, S. 2011. “UNHCR and Crowdsourcing, A partnership with the net/2011/11/09/crowdsourcing-unhcr-somalia-latest-results/. Stand By Task Force.” Accessed 1 June 2014. http://swfound.org/ media/54239/s2-02%20roberts.pdf. IT News Africa. 2011. “Google maps out Southern Sudan.” Accessed 25 April 2014. http://www.itnewsafrica.com/2011/07/google-maps-out- Seeger, C. J. “The role of facilitated volunteered geographic information in south-sudan/. the landscape planning and site design process.” GeoJournal 72, no. 3-4 (2008): 199-213. King, S.F. and Brown, P. “Fix my street or else: using the Internet to voice local public service concerns.” In Proceedings of the 1st international Shelter Associates. 2014. “Research and Poverty Mapping (city slum conference on Theory and practice of electronic governance, 2007. profiles).” Accessed 9 June 2014. http://shelter-associates.org/ ICEGOV: 72–80. mapping-research/. Lamy, F. 2011a. “South Sudan is now official on Google Maps.” Accessed 25 ———. 2014. “Home Page.” Accessed 9 June 2014. http://shelter- April 2014. http://blog.google.org/2011/09/south-sudan-is-now-official- associates.org. on-google.html. ———. 2014. “GIS and Remote Sensing.” Accessed 9 June 2014. http:// ———. 2011b. “South Sudanese sing and map their way to independence.” shelter-associates.org/gis-remote-sensing. Accessed 27 April 2014. http://google-africa.blogspot.gr/2011/07/south- Silvertown, J. “A new dawn for Citizen Science. “Trends in Ecology & sudanese-sing-and-map-their-way.html. Evolution 24, no. 9 (2009): 467-471. ———. 2011c. “Using the power of mapping to support South Sudan.” South Sudan Mapping. 2014. Accessed 27 April 2014. https://sites.google. Accessed 26 April 2014. http://googleblog.blogspot.gr/2011/05/using- com/site/mapsudan/. power-of-mapping-to-support-south.html. Sui, Z., D. “The wikification of GIS and its consequences: Or Angelina Longley, P. A., Goodchild, M. F., Maguire, D. J., and Rhind, D. W. Geographic Jolie’s new tattoo and the future of GIS.” In Computers, Environment Information System and Science. Fourth Edition. Oxford: John Wiley & and Urban Systems 32 (2008) : 1–5. Sons, Ltd, 2015. Thörnelöf, E. EPA Network Workshop – Citizen Science 12 -13 March Map Kibera. 2014. Accessed 15 April 2014. http://mapkibera.org/. 2013 - Increasing involvement of the public in large-scale monitoring of large carnivores across Scandinavia. Sweden, EPA, 2013. Accessed 1 June 2014. http://epanet.ew.eea.europa.eu/ad-hoc-meetings/epa- network-workshop-citizen-science-12-13-march-2013/eva-thornelof_- sepa-sweden.pdf. Tomlinson, R. Thinking about GIS: Geographic information system planning for managers. Fifth edition. Redlands, CA: ESRI Press: 2013. Touya, G. and Brando-Escobar, C. “Detecting Level-of-Detail Inconsistencies in Volunteered Geographic Information Datasets.” Cartographica 48, no. 2 (2013): 134–143. United States Agency for International Development. 2017. “Our work.” Accessed June 2017. https://www.usaid.gov/bangladesh/our-work. United States Department of Agriculture, Farm Service Agency. 2014. “Naip Imagery.” Accessed August 2017. https://www.fsa.usda.gov/programs- and-services/aerial-photography/imagery-programs/naip-imagery/. U.S. Geological Survey. 2014. “Home.” Accessed 1 June 2014. http:// navigator.er.usgs.gov/. Wald, D.J., Dewey, J.W., Atkinson, G.M., Worden, C.B. and Quitoriano, V.P.R. “Science Resulting from US Geological Survey’s ‘Did You Feel It?’ Citizen Science Portal.” In AGU Fall Meeting Abstracts, 2016, February. Waters, T. 2010. “The OpenStreetMap project and Haiti Earthquake Case study.” Accessed 8 December 2013. http://www.slideshare.net/chippy/ openstreetmap-case-study-haiti-crisis-response. Zoljargal, M. 2013. “Ulaanbaatar to become a ‘smart city’.” Accessed 12 April 2014. http://ubpost.mongolnews.mn/?p=3238. REFERENCES AND SOURCES 153 The Global Facility for Disaster Reduction and Recovery (GFDRR) is a global partnership that helps developing countries better understand and reduce their vulnerabilities to natural hazards and adapt to climate change. Working with over 400 local, national, regional, and international partners, GFDRR provides grant financing, technical assistance, training and knowledge sharing activities to mainstream disaster and climate risk management in policies and strategies. Managed by the World Bank, GFDRR is supported by 34 countries and 9 international organizations. WWW.GFDRR.ORG