An environmental impact assessment (EIA) is a decision-making process that systematically evaluates the possible significant (negative or positive) effects that a proposed project action may exert on the natural, social and human environment of a particular geographic area. The assessment results are often included in a document known as an environmental impact statement (EIS). An EIS usually includes, but is not limited to, the following topics: the environmental objectives and regulations, the baseline conditions of the existing environment, the proposed project and alternatives, impact analysis on the affected environment and consequences, public comments and comment analysis, mitigations and recommendations and other monitoring measures. The EIA process consists of multiple stages, and citizens normally participate in EIA at the stage of scoping and public reviewing. The scoping stage provides an opportunity for the pubic to express what they would like the EIS to address and the reviewing stage allows the public to comment on the draft EIS before the final EIS is generated. Other name variations, such as environmental impact report (EIR), have been used; for consistency, this study uses the term “EIA” to represent similar processes and “EIS” to represent similar documents hereafter.
Although the importance of involving the public to participate in the EIA process has long been recognized [1
], it is often regarded as ineffective, due to material, technical and process complexity, as well as the nature of environmental information [2
]. How to improve the effectiveness of public participation in the EIA process continues to draw the attention of researchers [1
]. Traditional methods for the public to participate in EIA include public meetings, telephone interviews, emails and surveys [3
]. Information technology and geographic information system (GIS) are considered two major vehicles that have brought innovation to the conventional approaches [4
]. The term “public participation geographic information systems (PPGIS)” first appeared in a GIS workshop organized by the National Center for Geographic Information and Analysis (NCGIA). The primary goal was to identify GIS solutions to empower the less privileged groups and incorporate socially differentiated local knowledge [5
]. Since then, PPGIS has evolved towards the direction of leveraging GIS to engage the public in policy-making, support non-governmental organizations (NGOs), grassroots groups and community-based organizations (CBOs) [6
]. Many PPGIS environmental information systems have been developed [7
]. However, researchers pointed out that previous work often attached importance to the aspects of environmental modeling, analysis or implementation techniques and rarely discussed the design and system usability with regard to involving the public more actively and effectively in EIA [10
]. According to Sieber, one of the major challenges for PPGIS is how to effectively and accurately present the available data and the results retrieved from the underlying modeling and quantitative analysis in a GIS-based environmental system such that the public would understand them correctly and be willing to participate [6
Recently, the advancement of Internet technology and the proliferation of Web 2.0 has brought new behaviors to a geographically-enabled information society and has infused public participation with new energy. The phenomena fall into six major yet overlapping categories [11
]: (1) user-generated content, which refers to digital information, such as videos, photos, blogs, news feeds and podcasts, created by the web users and normally shared with other users [12
]; (2) leveraging the collective wisdom of the crowd to identify a design or solution (e.g., crowdsourcing) [13
]; (3) large-scale data of multi-dimensional complexity that often require massive storage space and powerful database searching, processing and analysis capabilities; (4) participatory frameworks, which associate citizens’ local knowledge and observations with geospatial footprints (e.g., an address) [11
]; (5) prevalent virtual community and social media involvement that encourage user interaction through online platforms, such as message boards, chat rooms or social networking sites; and lastly, (6) harnessing geospatial technology to benefit non-expert users or communities [16
]. With the empowerment of these evolving categories, users are more likely and willing to be engaged in generating and consuming web-based geospatial contents [17
]. This, in turn, leads to another new phenomenon which has been termed “volunteered geographic information” (VGI) [18
]. Researchers have also used “neogeography” [16
], “cybercartography” [22
] or “asserted geographic information” (in the sense that its content is asserted by its creator without citation, reference or screening of another authority) [18
] to describe similar activities. For the sake of consistency, this paper will use the term “VGI” hereafter.
The growth of VGI contributes to the gathering, sharing and visualization of geographic information. Unlike the traditional top-down approach that generates or distributes paper-based or web-based geospatial contents from authoritative or professional sources (e.g., governmental agencies), VGI adopts a bottom-up approach that engages a large number of citizens and shifts the role of users from mere “data consumers to active participants and providers” of user-generated content [23
]. This shift can leverage local intelligence to benefit local government, state agencies and CBOs in improving decision quality, reducing process time, promoting cost-effectiveness, maintaining transparency and legitimacy and ultimately contributing to the development of a civil society [24
The importance of VGI in environmental areas has also been recognized by scholars. Research suggests that VGI can contribute to reflecting public environmental concerns [25
], promoting environmental awareness [26
], providing environmental policy makers with local and timely data [26
] and expediting new public policy changes on small scales in environmental monitoring activities [27
]. Particularly relevant to GIS and public participation, citizens can now access variations of VGI to contribute information located on a map through web-based mapping application interfaces [28
]. For example, Werts developed a WebGIS framework to increase public participation in soil and water conservation [26
]. Another example is OakMapper (http://www.oakmapper.org/
) that allows people to report tree disease information and obtain an overview of the tree health status across California using an online mapping tool [29
]. With the help of VGI, agencies and organizations are now able to gather geographic information with local intelligence that may not be available otherwise.
As mentioned previously, public comments are often submitted to the EIA lead agency via mail, email, phone or fax in many EIA projects. These comments consist of text mostly and the volume can grow rapidly during the draft EIS phase. After receiving the comments, agency staff need to screen, consolidate and analyze them based on the National Environmental Policy Act of 1969 (NEPA) and/or the EIA lead agency’s approach and respond to each of the individual comments (NEPA §1503.4(a), http://ceq.hss.doe.gov/nepa/regs/ceq/1503.htm
). In many cases, there is no effective automated mechanism for decision makers to explore and extract meaningful information from raw comments. This problem can be alleviated by data mining. Data mining is a technique that aims at automatically extracting data and discovering implicit patterns and trends from large data sets or databases [30
]. Those patterns and trends often represent intelligent information related to relevance, innovativeness and interestingness. Data mining usually consists of three kinds of tasks: information retrieval, named-entity recognition and information extraction [31
]. Spatial data mining often takes place when data mining algorithms are applied to geospatial information. Geospatial information retrieval and extraction has recently been identified by the National Geospatial-Intelligence Agency (NGA) as one of the key research areas in the next 15 years [15
]. Benefits of data mining include improving data collection efficiency, reducing efforts and costs, enhancing data quality and coherence of assessments [32
] and supporting good strategic identification in decision-making [33
The primary objective of this study was to develop a spatially intelligent solution to improve the effectiveness of public participation in the EIA process. Criteria to define effectiveness are presented later in Section 5.1
. The secondary objective was to leverage the solution to raise the environmental awareness of the public and contribute to the knowledge base regarding public participation in policy making in general. The research integrated GIS, VGI, social media tools, data mining and mobile technology to design a spatially enabled framework that presented and shared EIA information effectively to the public. A prototype spatially intelligent public participative system (SIPPS) was also developed as a proof-of-concept of the framework.
7. Limitations and Remarks
The study has received not only positive feedback, but also raised interesting questions. A few participants expressed concerns about the quality of comments and social media contents collected from the SIPPS prototype system. Comments submitted electronically can be redundant, irrelevant or even malicious. Similarly, the social media contents displayed on the application map viewer are dynamically retrieved from the corresponding social media networks and can contain inappropriate or irrelevant information. A more fundamental problem related to this is the credibility of VGI. VGI content generally consists of three forms: personal opinions (e.g., user comments), objective or factual information (e.g., a photo or video pointing to a particular location), and measurements (e.g., GPS readings) [41
]. Although VGI has the strength of collecting local information from a large number of citizens, the content generated from collective mapping approaches is often produced by non-experts, and the process usually lacks standardization and coordination of authorities and professionals [21
]. The strength is, thus, a double-edged sword that also becomes its major challenge that causes data quality problems [42
]. Data quality is compromised when the user-generated content is subjective, unstructured, poorly documented, insufficiently referenced or validated without scientific screening [43
]. Data quality can also be impaired due to missing, inconsistent or conflicting values [44
]. Therefore, how to integrate participatory mapping techniques especially collective mapping approaches into more formal governance adds another dimension of complexity to the VGI credibility problem. Since collective mapping approaches normally take place when users participate in the data contribution process, strategies are suggested to tackle the complexity challenge from three angles: contributor, data and the contribution process. First, local users or community volunteers can be recruited to screen VGI content and serve as gate-keepers to improve the data quality [45
]. Since these contributors are often familiar with the surrounding environment or local events, they are more likely to collaboratively validate the VGI content and identify missing or conflicting data. Moreover, literature suggests that integrating the capability of commenting on VGI submission may improve public understanding, data quality and the knowledge base [21
]. The design of the SIPPS framework is aligned with this suggestion. The comment component in SIPPS provides citizens with the ability to submit new comments, review comments, comment on existing posts and rate comments. The survey results also indicated that by providing this interactive functionality, the comment component increased user confidence in the EIA process transparency and willingness to access and share information. It is therefore believed that SIPPS can help overcome the VGI credibility issue and formalize the data collection framework. Nevertheless, it should be noted that this strategy assumes that the recruited contributors are from the local community or have knowledge about the local environments [28
]. To validate that they possess the local intelligence, a brief test is suggested to evaluate user knowledge of the related topic or environmental condition [28
]. For instance, if the goal of an EIA project is to determine a new power station site in a park, a possible evaluation might involve participants to identify locations of major park features on a map to confirm their familiarity with the area. Note that this research also recognizes that the self-evaluation step could add extra overhead to the existing process and may lessen a user’s enthusiasm to participate.
Second, data quality issues in collective mapping may be alleviated by leveraging the power of social media networks and crowdsourcing. Social media networks generate a large number of live feeds on a continuous basis and within a short turnaround time. Although individual feeds can have uncertain quality, data abundance provides an opportunity for grass roots participation to perform cross-validation with considerable VGI content [43
]. With this strategy, inputs are aggregated from numerous individuals or sources, and the relevance of the output VGI clusters are evaluated by the contributors in a given context [46
]. As a result, more reliable information can be extracted for better data quality. The SIPPS framework responds to this strategy by integrating social media tools in the comment component. The social media tools retrieve live feeds with predefined keywords from social media networks (YouTube, Twitter and Flickr) and display them on a map. By allowing the crowd to review multiple feeds from various sources, SIPPS provides participants with an opportunity to corroborate the information using crowd wisdom. The content retrieved from social media tools can still be irrelevant, and that is why data mining was proposed as part of the SIPPS framework with the goal to identify interesting patterns and filter out unwanted or undesirable entries in an automatic fashion. It is also believed that content filtering should be part of the monitoring process of the EIA responsible agencies if they choose to employ a production web system similar to SIPPS to support the EIA process. It is therefore recommended that the EIA agencies delegate an administrator to steward the data contents and manually remove any inappropriate information. Furthermore, a rating mechanism is considered one type of new arbiter of VGI credibility for bottom-up approaches implemented through collaborative efforts [21
]. Since VGI content in many cases represents personal opinion or observation rather than scientific data, an objective notion of data quality or accuracy by itself does not seem entirely suitable to assess this type of information [21
]. Instead, methods that can reflect people’s perception and response to specific scenarios or cultural, political and economic conditions are more appropriate [11
]. With regard to the SIPPS prototype, it introduced rating functionality in the comment component for users to examine the more subjective VGI contents. The rating mechanism makes it possible for comments with higher ratings to draw close attention. Collaborative rating on a large scale may also contribute to elimination of inaccurate values. One of the survey respondents suggested that SIPPS include the “Like” functionality (as can be found on Facebook) in the comment component as another dimension for users to examine VGI contents. This suggestion is actually being considered as an area for future work.
In addition to improving the management of contributors and data, the contribution process should be regulated such that the data collection and production chain can operate in a more controllable manner [47
]. A series of standards, methods and techniques need to be established to streamline VGI content generation, specify boundaries for data requirements, data ownership, data management and distribution, participant recruitment and address possible responsibility and ethical concerns [21
]. The establishment of this discipline may progress gradually and formalization of governance may limit the freedom of contributors to some extent. To achieve the best outcome, a balance must be maintained between allowing contributor freedom and compliance with necessary specifications [49
]. Similarly, a more comprehensive way to deal with VGI content is to assess each given scenario (e.g., an EIA alternative) with data from both authoritative origins and non-expert VGI sources [50
Another interesting finding that emerged from the survey is the adoption of collaborative mapping techniques into the decision-making process. Evidence from the survey results show that collaborative mapping techniques have the following advantages over other traditional participatory methods in the EIA process: First, the spatial element provided by collaborative mapping has the strength of visualizing geographic features that may not be effectively demonstrated through other interfaces or information sources. Spatial analysis also facilitates early identification of environmental alternatives and problems. Second, findings from the survey (Question 30—What is your opinion about the EIA project before and after visiting the website?) suggested that spatial presentation along with the content sharing and interacting capability can disclose environmental facts and change perceptions on activities and problems. Not only is this valuable to the public in promoting understanding and social learning, but also to EIA decision makers in managing public expectation and mitigating potential conflicts that may happen later in the EIA process. This study identified three major factors that impact the adoption of collaborative mapping techniques in the EIA process: cost, time and outcome. A number of respondents indicated that they would like to know the budget for each EIA alternative and the expense to implement the SIPPS solution. With the current situation of budget cuts and personnel reduction in many environmental agencies and public services, it is important that a solution fits into the EIA process with reasonable cost. SIPPS meets this requirement by providing a reusable application template that can be configured and deployed with manageable cost and effort to support similar environmental projects. Although a few commercial-off-the-shelf software products were used in its development, data and map services used in the framework can alternatively be published and hosted through cost-reduced packages designed for organizations, especially environmental and humanitarian institutions.
The time factor is about when to participate and whether a solution will improve time efficiency in EIA. Most survey respondents recognized that SIPPS reduced the time required to participate in the EIA process. From the decision-making side, concern was raised about the timeframe constraint in the EIA public reviewing phase. One respondent acknowledged the benefits of SIPPS in increasing environmental awareness and improving public participation in the EIA scoping phase, but was concerned that the application might make the public reviewing process unnecessarily elongated and delay the final decision process, due to the extra volume of comments that need to be analyzed and additional back-and-forth conversation and negotiation. The study recognized this concern as a possibility and an important factor for the acceptance of collaborative mapping approaches in decision-making, but since the main goal of the study is to improve the effectiveness of public participation in EIA, how to handle the situation is beyond the scope of this research.
Outcome is also considered as a significant factor that influences the adoption of collaborative mapping approaches in EIA decision-making. User feedback in the survey suggested three aspects to cover outcome: uncertainty, diversity and conflict. As previously mentioned, VGI credibility is a major challenge that causes uncertainty. Aggregating crowd wisdom is able to help verify general information and enhance information provided by authorities. Reducing data fluctuation between good and bad will minimize uncertainty and increase the confidence of decision makers in undertaking a VGI solution. Furthermore, one of the fundamental goals of involving the public in EIA is to empower marginalized groups and establish a more equitable decision structure across the distribution of political power [51
]. Ideally, the EIA participatory process should account for common views shared by individuals or representatives of various groups [4
]. Therefore, diversity becomes a consideration for decision makers to accept collaborative approaches. Collaborative mapping approaches, on one hand, provide more flexibility by releasing people from spatial and temporal constraints of attending physical meetings (Feedback in Questions 23 and 24 confirmed this finding); on the other hand, VGI contents are often presented through a web-based interface, which can create a participation barrier for people who lack computer literacy and Internet access. This limitation is known as the “digital divide” [52
]. In terms of conflict, a desirable participatory method should support decision makers to prioritize concerns, communicate outcomes and manage conflicts. Before decisions are made on EIA projects, risk and impact analysis of each alternative needs to be conducted. VGI content that reflects local intelligence and creativity not only assists decision makers in identifying alternatives, but also allows them to explore various options and discover better solutions. In SIPPS, this is realized by the spatial feature editing function and comment responding capability. Once a decision is made or project activities are under development, the progress and outcome must be communicated to the public. Collaborative mapping approaches are capable of distributing the information in a timely manner to a large population via the web and the mobile platform. In SIPPS, a map viewer on both the web and the mobile interface is provided for decision makers to inform and collect feedback from the public. The social and political interaction enabled by collaborative mapping approaches between decision makers and the public also represents a more proactive way to avoid and resolve conflicts, because it demonstrates a sense of ownership on issues, a traceable sequence to address concerns and mutual understanding of legal requirements, values, perceptions and interest of involved parties.
Besides the above challenges, it is recognized that the study sample size is limited and some of the inferences discovered from the research sample may not be generalized to a larger population. The results can be skewed towards younger, more educated people who are computer literate. Responses indicated that the older generation, who may not have computer skills and access to social media tools, were usually more concerned about the impacts of EIA activities (especially the impact on health and safety) and, thus, more likely to participate in EIAs.
Lastly, while the study produced a conceptual framework and a prototype system to improve the effectiveness of public participation in EIA, it does not recommend completely abandoning the traditional methods. One respondent pointed out that she would use both the traditional methods and a website, like the one designed in this research, to raise her voice if an EIA was really concerning her. Therefore, a combination of both instruments may generate a better outcome and benefit the citizens to the greatest extent.