Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 23
A Research Agenda for Geographic Information Science at the United States Geological Survey 2 USGS Needs for GIScience Capabilities This chapter addresses the first of the committee’s tasks—to identify current and future U.S. Geological Survey (USGS) needs for geographic information science (GIScience) capabilities. The chapter begins with USGS-wide needs, then describes National Geospatial Program Office (NGPO) needs, and ends with the needs of The National Map. Given the strategic importance of The National Map to NGPO and its potential value across USGS and to the nation, the majority of this chapter is devoted to understanding the drivers of potential priorities among the many GIScience research needs of The National Map. Following USGS’s suggestion that the committee expend the majority of its effort on the other two items in its task, the committee relies heavily on summarizing existing information in this chapter. BUREAU-WIDE GISCIENCE RESEARCH NEEDS Multidisciplinary research is a central goal of the USGS science strategy (NRC, 2001; USGS, 2007). GIScience research underpins this goal by addressing such challenges as data integration across the disciplines. As the Center of Excellence for Geospatial Information Sciences (CEGIS)’s USGS-wide FY 2007 call for proposals states, “The role of the [Geographic Information Office] in providing a geospatial framework for integrating information among all the USGS science disciplines is an important element of GIScience at the USGS.” Access to and mining of geographic base data are also critical components of the USGS research infrastructure (McMahon et al., 2005), and GIScience research supports these capabilities.
OCR for page 24
A Research Agenda for Geographic Information Science at the United States Geological Survey The USGS’s science strategy outlines an inspirational science agenda for the agency. Every one of the six interdisciplinary “science directions”1 discussed in the report has a profound reliance on geospatial databases and methodologies. This broad and deep reliance on geospatial content and solutions points to a rich opportunity for CEGIS to contribute to the other USGS disciplines, especially through an improved National Map. Examples are cited in the following paragraphs. For the ecosystem direction, the report states that “USGS and partners will develop new products, including standardized national maps of ecosystems in the United States and regularly updated status and trends assessments” The human health direction underscores its dependence on geospatial technology with the statement that “USGS capabilities in environmental monitoring and mapping are important components for understanding environmental relations to human health and evaluating probable outcomes of future human health risk factors.” However this initiative goes further in defining a strategic action for geospatial technology to “develop an online atlas of potential environmental health threats … [and] develop and implement a national-scale, real-time, environmental health threat warning system that combines biological, water-quality, and geologic information with GIS [geographic information system] decision-support tools.” Indications that there is more work to be done in GIScience also appear in relation to the water census direction: “The Water Census will also require improvements in the mapping and characterization of the geologic and geomorphic framework of the Nation’s principle aquifers and watershed systems.” These objectives will necessitate enhancements to current databases, platforms, and tools—in particular, advances in scale handling, temporal analysis, real-time access to data, standards, and multidisciplinary analysis. Requirements for the first three are captured in the vision of the ecosystem direction that “comprehensive, multi-scaled, online digital maps of the Nation’s ecosystems and their physical and biological components are routinely used for management, education, and portrayal of change over time. Real-time ecological data, images, maps, and research findings are available to the public on interactive USGS websites.” Scale Adjustment The ecosystem direction aims to “coordinate, develop, and regularly update a standardized national map of ecosystems and their physical and biological components, at scales appropriate to land-manager needs, to facilitate the ability to assess, monitor, manage, and restore ecosystems.” This range of scales is unique to USGS: “The position of USGS as a non-regulatory agency, with capabilities in environmental 1 Water census, human health, hazards, ecosystems, climate, and energy.
OCR for page 25
A Research Agenda for Geographic Information Science at the United States Geological Survey monitoring and mapping at all scales from national to local and the ability to understand environmental and ecological processes, is found nowhere else in the Federal Government.” Temporal Analysis Temporal analysis and real-time access expand the roles for a geospatial database. The temporal analysis requirement also underpins the key question in the climate direction: What links between climate, land use, and hydrology influence the temporal and spatial characteristics of water resources? Access to Real-Time Data Access to real-time geospatial and other data is a rapidly growing requirement for managing the nation’s natural resources. The human health direction identifies a strategic action requiring real-time analysis and challenges the USGS to “develop an online atlas of potential environmental health threats, which consolidates USGS data and information and provides real-time data for researchers and public-health agencies to enhance the Nation’s ability to respond quickly to current threats and anticipate potential future threats.” Management of natural hazards is also highly dependent on timely data dissemination. The USGS’s vision in terms of hazards expects that “by 2017, the USGS will … significantly expand urban hazard mapping throughout the Nation … and we will have the models, metrics, decision-support tools, and portals that provide intelligent access to remotely sensed data and geospatial information for cost-effective risk-reduction, response, and recovery efforts … [G]eographic methods and tools need to be developed … and intelligent access [provided].” In addition, the energy direction anticipates the need to share information as well as “maintain and update the geological and geophysical databases and geochemical baselines used to develop national and global resource assessments … [and] … assure the data are accessible both internally and externally.” Lastly, the vision for the hazard direction is to have all seismically active areas served whereby “associated maps of shaking level, population density, and susceptibility to landslides will be posted on the Internet within minutes of the determination of the earthquake location and magnitude.” The implications of information dissemination are global and multinational—as summarized in USGS’s final vision statement: “An international consensus is developing on the need to leverage recent advances in computer science and related technologies to create a next-generation, integrated science computing and collaboration platform that will be as transformational as the Internet.”
OCR for page 26
A Research Agenda for Geographic Information Science at the United States Geological Survey Standards Standards are a catalyst for the success of GIScience technology at USGS. The final vision of the science strategy refers to the importance of standards: “The use of open standards and of tools has minimized the difficulty of merging or comparing data sets; searches on a location or topic of interest quickly yield comprehensive research data.” Multidisciplinary Analysis The USGS’s multidisciplinary composition is unique among federal agencies. “The USGS is the only Federal agency that combines scientific expertise in biology, hydrology, geology and geography.” This creates unique capabilities such as in hazard forecasting, for example: “These accomplishments [hazard forecasting, monitoring, diagnosis] are all possible because the USGS is able to bring a unique combination of disciplines—biology, geology, hydrology, geography, and geospatial information technology—to bear on all these hazards.” The key ingredient for success is data integration, captured in the strategy for the climate direction: “Our breadth of multidisciplinary scientific expertise … enables us to deliver uniquely integrated information.” In fact, this capability has the potential to be world class, because, as the human health direction states, “USGS databases constitute one of the most comprehensive and high-quality arrays of national, regional, and local biologic, organic and inorganic analyses available from any single source.” Yet the integration of diverse data sets is still a great challenge, as discussed in the next section. Challenges Although the USGS has a unique set of databases and disciplines, there is substantial work ahead because, as discussed in the health direction, “the many data sources are scattered across the USGS and not easily available to most users. If [catalogued,] this array of environmental data and information could provide our partners and customers with unified spatially and temporally referenced sources of information. An important step in the overall goal of protecting public health is to integrate existing USGS databases [to identify] potential environmental health threats and provide the underlying framework for USGS environmental health studies.” The solution to these and other data integration needs is a geospatial platform comprising spatial and semantic reference systems (Kuhn, 2005) that will support spatial and semantic integration. The National Map is such a geospatial platform, but significant research and development is needed for it to meet
OCR for page 27
A Research Agenda for Geographic Information Science at the United States Geological Survey USGS’s science needs over the next decade. In addition, it is both conspicuous and of concern to the committee that throughout USGS’s inspirational science strategy—with its numerous references to mapping databases, capabilities, and tools—The National Map is never mentioned. However, this can be considered an opportunity for NGPO and CEGIS. As shown above, for the USGS disciplines to successfully implement their science strategy, there are numerous challenges that are inherent to GIScience, including issues with scale, temporal analysis of geospatial data, improving access to real-time data, and developing the standards and data integration techniques that will allow true multidisciplinary analysis. All of these are challenges that must be resolved for successful implementation of The National Map, therefore it became clear to the committee that the GIScience needs of the USGS as a whole and those of The National Map are highly complementary. GISCIENCE RESEARCH SUPPORTING NGPO’S COORDINATION RESPONSIBILITIES FOR THE DEPARTMENT OF THE INTERIOR The Department of the Interior leads federal activities to improve the use of geospatial data through the development of the National Spatial Data Infrastructure (NSDI) and the coordination of geospatial activities across the federal government. NGPO’s four federal coordination activities are the Federal Geographic Data Committee (FGDC), Geospatial One-Stop, Geospatial Line of Business, and the National Geospatial Advisory Committee. These activities have not established research needs, which raises the question of whether these coordination activities might generate a need for specific research. Early work in relation to federal coordination focused on technology to reduce barriers to data sharing, but emphasis has now shifted to identifying common agency needs and aligning business practices and investments. The remaining technology component focuses on improved interoperability or the ability of dissimilar networked computer systems to exchange data and instructions to provide computing services. Of particular interest are approaches that improve interoperability but minimize the need to change or retrofit participating organizations’ existing technology implementations. Nonetheless, while the success of NGPO’s coordination activities has a technical component, participants in NGPO “listening sessions” have noted that the main challenges are organizational, not technical, in nature. Therefore, the committee has assumed that any research needs of these geospatial activities would be adequately met by the research agenda developed to meet the criteria developed for this report.
OCR for page 28
A Research Agenda for Geographic Information Science at the United States Geological Survey NATIONAL MAPPING GISCIENCE RESEARCH NEEDS The list of potential topics in a CEGIS research agenda is daunting. An allencompassing list would have to address the needs of the NGPO, the USGS disciplines, other federal agencies, state and local agencies, and the public (commercial firms and private citizens). In addition, a wealth of potential research topics has already been identified in a number of documents (e.g., USGS, 2001; NRC, 2002; NRC, 2003; McMahon et al., 2005; CEGIS, 2006; and DiBiase, et al., 2006). In light of CEGIS’s lean start with two full-time researchers and a director there is no doubt that the center needs to prioritize among all these demands. As discussed in Chapter 1, there is wide consensus on the importance of The National Map. This consensus began to build with the conclusion by the National Research Council (NRC, 2002) that developing The National Map was the single most important initiative for the geography discipline. McMahon et al. (2005) reiterated this sentiment. Also in an interview for this study, Karen Siderelis, Associate Director for Geospatial Information at the USGS, said that NGPO must get The National Map right as its first priority (Karen Siderelis, USGS, personal communication, 2006). An effective National Map is the most critical success factor to CEGIS and the NGPO. If The National Map is to be the primary source of USGS geospatial products to customers outside of the USGS, then ultimately focusing on the research needs of The National Map will address the needs of these external users. Yet what about the additional research needs of the NGPO and the USGS disciplines? As shown earlier in this chapter, the committee consensus is that topics relevant to National Map implementation cover a wide range of the GIScience research agenda that is broadly relevant to USGS (e.g., integration of data from diverse sources, effective display and processing, temporal processing of spatial data). Given the time constraints of this study and the sponsor’s request to focus on tasks 2 and 3, it was difficult for the committee to assess the unique GIScience needs of the USGS disciplines above and beyond those that meet the overall USGS science strategy. However, the needs of the USGS disciplines of water, geology, biology, and geography can be directly assessed by CEGIS and included in the research agenda to ensure that The National Map will meet their important needs with appropriate priority. Consequently, a dedicated initial focus by CEGIS on The National Map will achieve three broad USGS goals: The National Map is primed for success, its use is increased, and USGS regains a leadership role in GIScience. The National Map becomes an essential platform serving the geospatial needs of the USGS disciplines and supports execution of the 2007 USGS science strategy.
OCR for page 29
A Research Agenda for Geographic Information Science at the United States Geological Survey The initial focus provides a concrete and necessary early target for CEGIS that yields a visible and measurable outcome of the research. It is important to note that The National Map is not just a data viewer. It includes the data, the processes for obtaining and maintaining the data, the national standard-certified nature of those data, the data models and relationships, and the data knowledge systems and ontologies that make up the vision of The National Map. In fact, there may be multiple viewers for The National Map, government, academic and commercial, as these organizations seek to exploit the richness and value of the data in The National Map. Success with The National Map is the prerequisite for any additional GIScience research at CEGIS. Since CEGIS must prioritize its research agenda to optimize the effectiveness of limited resources, the committee feels that an initial focus on the GIScience needs of The National Map would provide the most visible benefits to CEGIS, NGPO, and the USGS disciplines while tackling some of the most significant GIScience topics confronting the geospatial community. As the resources of CEGIS grow and early successes with The National Map are achieved, the research agenda can and should broaden to more diverse areas of research. RECOMMENDATION 1. CEGIS should initially focus on research that will improve the capabilities of The National Map. Within The National Map focus, the delivery of topographic maps deserves special mention. Topographic maps were the virtual brand of the USGS during its formative years. The National Map is often thought of as the electronic evolution of the topographic map.2 However, the current National Map does not deliver fully on the concept of the traditional topographic map nor has it fully embraced a broader concept of topographic information representation and delivery in other than map form. Despite the impressive data access and manipulation features of The National Map, the topographic map “brand” has been diluted. Also, because paper maps are still favored by many users (particularly in the emergency management community), USGS will have to consider bringing its branded product back into play in The National Map. USGS’s focus would need to remain on topographic information but with an expanded range of products and services. One thrust involves revitalization of the traditional topographic map as an effective cartographic product in light of new digital design research as well as research on user interaction and effective interactive cartographic products. Another thrust involves ontology development and new data models for The National Map that move it beyond traditional and current forms of topographic information delivery. Investigation of “intelligent” 2 See http://nationalmap.gov/report/national_map_report_final.pdf.
OCR for page 30
A Research Agenda for Geographic Information Science at the United States Geological Survey topographic features and access to fundamental topographic information in other than map form open new research directions and enhanced capabilities for The National Map that support queries for information not easily or fully served by current methods. The remainder of this section describes critical details of The National Map (including its possible evolution, mentioned in the previous paragraph) upon which the committee builds its discussion of research priorities in the next chapter. The Current National Map The National Map is “a database of continuously maintained base geographic information for the United States and its territories that will serve as the Nation’s topographic map for the 21st century” (USGS, 2001). This new topographic map is an electronic realization of the original paper maps, and with that structure comes greatly increased flexibility and opportunity. The term “national map” encompasses the underlying data, data management approach, data access methods, supporting partnerships, and models of data use (Box 2.1). USGS describes the relationships among the The National Map, Geospatial One-Stop, and the NSDI in the following way: “NSDI is a concept defined as including all aspects of geospatial data, Geospatial One-Stop is a communications portal for geospatial information content and related information, and The National Map is geospatial information content in the form of data and applications” (USGS, 2003). The relationships among The National Map, Geospatial One-Stop, and the FGDC can be visually expressed in the architecture of The National Map (Figure 2.1). In addition to the architectural elements, the figure is divided into three levels that represent general topics under which research can be organized: data integration, data access, and data use (Chapter 3). Data Integration. In the data integration level, all eight national data themes defined by the USGS are integrated into The National Map database (as the data management component). The National Map databases are accessed and used by The National Map web server, the National Atlas database, and the Geospatial One-Stop portal. Data Access. At the data access level, The National Map web server provides the graphics and the data streams for The National Map viewers (web browsers) to display maps. The end users can query and combine multiple layers provided by the map server and customize their final mapping design and layout by using The National Map viewers. The Geospatial One-Stop portal provides data catalog services and web mapping catalog services to the end users (who might be GIS professionals or scientists). End users can search, request, and download the data they need. In addition, the National Atlas server provides mapping functions for the end users through the National Atlas viewers.
OCR for page 31
A Research Agenda for Geographic Information Science at the United States Geological Survey Data Use. At the data use level, users create end products of The National Map, including paper maps or graphic maps displayed on a computer screen or embedded in a program. Two major groups of users are USGS scientists in the various disciplines and the general public. Scientists can download data from Geospatial One-Stop or directly from The National Map databases for their domain-specific research or spatial analysis. These applications will create new GIS data sets with added value. These new data sets are added back to the Geospatial One-Stop Portal or The National Map web server for future applications or use. Public users generate maps for activities such as hiking, school projects, and map reading. BOX 2.1 What Is The National Map? The USGS’s vision of The National Map targets improvements in data characteristics such as currentness, seamlessness, consistent classification, variable resolution, completeness, consistency, variable positional accuracy, spatial reference systems, standardized content, metadata, and temporal dimensions (USGS, 2001). The goals of The National Map are ambitious in attempting to integrate data from a large number of sources, many with widely varying specifications. Here, the committee describes the components of The National Map and then illustrates how the components fit together in the cases of two layers within The National Map: the National Hydrography Dataset and high-resolution imagery over urban areas. Data: The data themes in The National Map are orthoimagery, elevation, hydrography, geographic names, land cover, transportation, structures, and boundaries of governmental units (e.g., states, counties) and publicly owned lands (e.g., national forests, state parks). These themes were selected in large part because USGS is authorized to provide them if no other source is available, and they typically comprise the information portrayed on USGS topographic maps. USGS’s original intent was that the data would retain the characteristics appreciated by users of USGS topographic maps, including consistent feature identification and classification and comprehensive national coverage. In addition, the data would be seamless (not interrupted by arbitrary edges introduced by the data production process), would be more current, and would have positional accuracy equivalent to or better than that of USGS topographic maps. Data management: USGS’s data management approaches support two methods through which partners participate in The National Map. These methods generally parallel the “blanket” and “quilt” approaches described by the National Research Council (2003). In the blanket method, the USGS maintains databases of national data coverage. Partners help maintain these databases by sending data snapshots or transactions that are incorporated into a national coverage. In the quilt method, USGS maintains a web-accessible database that catalogs web services maintained and hosted by partners. The former approach offers improved data integration and utility to customers, especially for advanced geocoding and modeling applications; the latter offers the potential for faster availability of more current data. Data access: USGS offers several methods to access data in The National Map. For users seeking to view a map of the data, a map viewer is available (see http://nmviewogc.cr.usgs.gov/viewer.htm) and provides a “print” function to create a page-size hardcopy. For those developing their own viewer or other applications, the USGS offers an application programming interface (API) and related web service through which users can use the catalog to access inventoried services. Metadata entries in the catalog are “harvested” into Geospatial One-Stop and so can be discovered through that portal. For those interested
OCR for page 32
A Research Agenda for Geographic Information Science at the United States Geological Survey in retrieving copies of data, especially those in national coverages, USGS offers interactive and preprocessed methods to select and retrieve data through the web or on media. Agencies have negotiated service-level agreements with USGS for more advanced web-based access to national databases. Partnerships: Perhaps the most fundamental change represented in The National Map approach is the transition from USGS relying on internal resources to collect new data to relying on partners to provide new data. These partnerships are based on an exchange of value between USGS and partners. In exchange for partners’ data, USGS has provided value in the forms of funding, data, data models, data collection software tools, and access to contracts and related management and quality assurance processes, information technology, web and other data management services, and expertise. Data use: The “build once, use many” approach of The National Map supports up to three models of data use. The first is visualization of the data in the form of a map graphic, in which the data encode the location and a basic description of a feature. The second is support for geocoding, in which the data support the assignment of a position based on another reference system (such as street addresses or stream reach codes). The third supports more advanced modeling, in which the data encode spatial information in addition to position, such as the direction of flow along streams. The three use models are a progression from lower value but higher volumes of use (and lower costs of development) to higher value but lower volumes of use (and higher costs of development). The National Hydrography Dataset, the hydrography component of The National Map, provides an example of the partnership activities mentioned above. Federal agencies, including USGS, the U.S. Environmental Protection Agency, and the U.S. Forest Service, provided the initial data model, supplied software tools and related training to implement the model and maintain the data, and identified common business needs among the agencies and with state partners. The agencies and more than 35 state and other partners pooled funding and data, and used the software tools to develop the data within their organizations or through contractors. USGS reviewed the resulting data, incorporated them into the national coverage, and made them available. USGS has begun to accept updates from partners in the form of transactions from partner-maintained databases. These data provide the “blue lines” for portrayal of hydrography on graphics, the basis for geocoding observations of water quality, water quantity, habitat, and other characteristics relative to the hydrographic network, and the network for modeling the flow of water. In addition to uses in USGS, the data provide a geospatial basis for the U.S. Environmental Protection Agency’s Watershed Assessment, Tracking, and Environmental Results project and for the exchange of geospatially referenced environmental information between the agency and its state partners, the U.S. Forest Service’s Natural Resource Information System water module, the Census Bureau’s Topologically Integrated Geographic Encoding and Referencing (TIGER) system modernization program, and the Federal Emergency Management Agency’s flood map modernization program. In the example of high-resolution imagery over urban areas, USGS pools its funds with those of intelligence and homeland security agencies, and reaches out to state and local governments to identify opportunities to share costs of data collection. In cases where partners are available, the federal and other organizations’ funds are pooled and the USGS or partners let contracts to collect data. In cases where partners already have data that meet or exceed federal requirements, the organizations can provide the data to The National Map through web mapping services registered in the catalog. They also provide the data for incorporation into USGS national data holdings, although the distribution of such data might be restricted to selected federal agencies. This approach spreads the costs of data development over more uses and users than would the initial “contingency” application of disaster response, in which data would likely be used only in case of an emergency. It also helps ensure that local and federal responders would be using the same base for a “common operating picture” when there is an emergency.
OCR for page 33
A Research Agenda for Geographic Information Science at the United States Geological Survey FIGURE 2.1 The current framework of The National Map. NOTE: API = application programming interfaces. EPA = Environmental Protection Agency; NASA = National Aeronautics and Space Administration; NOAA = National Oceanic and Atmospheric Administration. It was beyond the scope of this committee to do a detailed analysis or critique of The National Map; however, based on testimony from invited experts (Appendix B), USGS science strategy documents and other published material, and informal testing by the committee, a general consensus about its current state was reached. While the concept of The National Map, drawing its content from local, reliable sources, and integrating those data for consistent access by all, is an excellent process for serving the nation’s geospatial needs, implementation of The National Map requires improvement in three key areas. First, access to and presentation of the data in the current system make usage
OCR for page 34
A Research Agenda for Geographic Information Science at the United States Geological Survey cumbersome and could benefit from user interface design and display methodologies. Secondly, the enormous benefit realized by combining data from reliable local sources brings with it difficult challenges in integrating and fusing those data with respect to scale, resolution, and quality. The current system has not automated these functions for the most part because, the technology needed is not available today. To deal with these issues effectively, the committee believes specific GIScience research is needed that will contribute not only to The National Map but to the field in general. Finally, with this foundation of superb data and integration, the committee believes that The National Map can take mapping to a new level of knowledge access and understanding, even beyond simple mapping. This requires significant restructuring of the data models in The National Map but it could be the key to being a world-class resource that integrates data not only over broad geographies but across varying semantics and time as well. Based on these and an analysis of the future requirements for The National Map described in the next section, the committee could begin to develop a list of priority research needs. Requirements for The National Map Requirements for National Map products should drive the design characteristics and content (Jack Dangermond, ESRI, personal communication, 2006). From the desired characteristics and content flow the practical challenges of delivering on these requirements. The challenges, in turn, determine the areas in which GIScience research could be most helpful. Based on discussions with USGS participants at the committee’s first meeting and a range of non-USGS participants at its second meeting (Appendix B), and through the committee’s general experience, it is clear that there is still great diversity in perceptions of the requirements, as well as the characteristics and content, of The National Map five years since it was first envisioned by USGS. To the committee’s knowledge, the principal requirements assessment for The National Map was conducted during the development of The National Map vision (USGS, 2001).3 Several high-level (i.e., nonspecific) needs emerged from this assessment (USGS, 2003): 3 USGS interviewed “key customers and individuals familiar with the development and use of geospatial data from five federal, two state, and two regional government agencies; nine private sector organizations; one educational organization; four professional organizations; and fifteen leaders and scientists in the USGS. A draft version of the report was posted for public comment, and this elicited 122 responses. A second, directed review was also conducted” (USGS, 2003).
OCR for page 35
A Research Agenda for Geographic Information Science at the United States Geological Survey There continues to be an unmet need for a common set of basic spatial data. For some places, large amounts of data are available; for others, very little. A standing collection of basic spatial data is needed. These data need to be current and useful for any arbitrarily defined geographic area. Both digital and paper forms of basic spatial data are needed. Federal leadership and commitment are needed to ensure that basic spatial data are available to support federal agencies in accomplishing their missions. As the nation’s civilian mapping agency, the USGS mission is to lead in the development and maintenance of this common set of basic spatial data. Requirements feature heavily in the initial phase of the USGS’s implementation plan for The National Map (USGS, 2003) (which was projected to end in FY 2005). This implementation plan proposes that requirements be collected by and fed through an advisory board. On a USGS web page dated March 20054 that describes the needs for The National Map, this entity is mentioned in the future tense. To the committee’s knowledge, specific requirements have not been published. In 2004, USGS published a cost-benefit analysis for The National Map (Halsing et al., 2004). The framework for estimating the benefits was “based on expected improvements in processing information to perform any of the possible applications of spatial data.” Consequently, the results are, at best, only an indirect indicator of requirements.5 Taken at face value (since the committee is not aware of an independent evaluation of this analysis), results from the “most likely” estimates of model parameters and data inputs indicate that over its 30-year projected life span, The National Map would bring a net present value of benefits of $2.05 billion in 2001 dollars, with a 95 percent confidence interval of approximately ± $1 billion (Halsing et al., 2004). In 4 http://nationalmap.gov/nmabout.html (accessed August 29, 2007). 5 The analysis did not attempt to determine the benefits and costs of performing spatial data-driven applications. Rather, “it estimates the change in the differences between those benefits and costs with The National Map and the current  situation without it. The estimates of total costs and benefits of The National Map were based on the projected implementation time, development and maintenance costs, rates of data inclusion and integration, expected usage levels over time, and a benefits estimation model.”
OCR for page 36
A Research Agenda for Geographic Information Science at the United States Geological Survey this scenario, the average time until the initial investments are recovered is fourteen years.6 Given the general ways in which requirements for The National Map have been expressed to date, the committee explored additional ways of exposing drivers for prioritization of CEGIS’s research. One such approach is to consider what differentiates (or could differentiate) The National Map from the many other geospatial resources now available. Unique Features of The National Map The National Map resides in an environment with a large number of other electronic mapping products and services from government agencies, academia, and private industry. In particular, the emergence of Google Earth and Microsoft Virtual Earth has captured the interest of the public as well as professional users. To develop a clear understanding of the added value and niche of The National Map, the committee considered this crowded environment and focused on what differentiates (or could differentiate) The National Map from other electronic map services. Although the government does not compete with other entities in a classic business sense, it has a responsibility to provide value from its investments and to serve its customers with the highest-quality products designed precisely to meet their needs. Who are USGS’s customers? They span many disciplines, organizations, levels of geospatial knowledge, individual experiences, education, and expectations—and their needs for spatial data vary from recreational to social to professional. The National Map must be the trusted geospatial information source for all of these constituencies and applications. The measure of the success of The National Map will be the extent to which these diverse users embrace and depend on the product. That said, it is impossible to be all things to all users right from the start (NRC, 2003) and identifying what differentiates The National Map in the crowded geospatial product field can point toward the 6 Because of many uncertainties in model assumptions and input data, the analysis considered 60 different scenarios to obtain an indication of the effects of these uncertainties. This sensitivity analysis shows that the baseline results are robust to large changes in one or more of the input values (several model parameters could be doubled, halved, or eliminated without causing substantial changes). However, the results were quite sensitive to changes in average change in the net benefit of an application implementation as a result of data from The National Map, rate of innovation of new applications, and amount of cumulative diffusion of data from The National Map and of the new and existing applications those data can inform. A few scenarios resulted in slightly negative outcomes, whereas the majority clustered around the “most likely” value.
OCR for page 37
A Research Agenda for Geographic Information Science at the United States Geological Survey high-impact initial research areas that help The National Map serve many different users. The USGS mission and the purpose of The National Map establish five requirements that in combination would make The National Map the most trusted source of quality data and knowledge about the geography of the United States: High quality: Data incorporated in The National Map need to be reliable and accurate. Critical applications including disaster response and homeland security must be able to depend on the quality of the data offered by The National Map. National coverage: The National Map will have to represent geospatial data across the United States to serve federal, state, local, and tribal needs in as consistent a manner as possible. Accessible: The National Map will need to be readily accessible to a wide constituency, from private citizens to professionals within government, academia, and industry. Continuously updated: While the original goal of The National Map was to have all data current to seven days, it is more practical to implement a continuously updatable database with processes to automate data ingestion, validation, and integration. The frequency of updates should reflect data latency, which may vary according to type of data, sources, and complexity of validation. Previous versions should be retained under a temporal model from which the history of changes can be obtained. Standardized: The National Map will need to reflect standards in classification schemes, metadata documentation, naming, and other characteristics so that data are easily shareable among different uses and users. Achieving this vision for The National Map is partially a matter of research and partially a matter of resolve and decision making. NGPO has an opportunity to define the personality and structure behind The National Map. With thoughtful but clear choices, this definition will establish a nationally recognized geospatial resource. After initial decisions to position The National Map, it will undoubtedly evolve, perhaps into different configurations and with specific access tools for varying constituencies. This is a benefit of an electronic product and will need to be exploited. For the important purpose of foreseeing what new areas of research would have to be conducted by CEGIS to support the evolution of The National Map, imagine what The National Map might become—a next generation that evolves from USGS’s initial vision from 2001 and reflects changes in geospatial capabilities since that time (Box 2.2). The committee’s vision is a means of highlighting what could be the priority areas for CEGIS research, since
OCR for page 38
A Research Agenda for Geographic Information Science at the United States Geological Survey it will need to support a version of The National Map that reflects today’s capabilities and their evolution. This vision is intended as a point of departure for an ongoing discussion led by USGS that should be linked to requirements for The National Map when these are defined. BOX 2.2 An Updated Vision for The National Map It is now 2015 and The National Map is one of the most visible products of the USGS— within the agency, to other federal agencies, to state and local geospatial offices, to commercial industry, and to the public. Users and organizations build their systems and models on The National Map. Because it is the only national trusted source of high-quality and comprehensive geospatial data, most organizations mandate the use of The National Map for all geography-based activities. At the public level, hikers, homebuilders, and others depend on The National Map for comprehensive topographic information just as in the mid-twentieth century, but now with the benefit of interactive geospatial services as a front end to a powerful and semantically rich database. Just as the paper topographic map has been distinguished by the quality of the data it represents, The National Map is distinguished by the quality of its database. The quality of a database is not only a function of the quality of the data it contains but of the underlying data models and the capabilities of the database management system. Thus, these have become focal points for setting a new National Map brand of excellence. Capabilities The National Map has become an intelligent topographic knowledge base capable of responding to wide ranging public needs for geographic information. This knowledge base is Knowledgeable about geographic features as they exist in the world, not just their cartographic representations; Beginning to accumulate process knowledge about how these features interact and how they change over time; Capable of responding to a range of different queries that might be posed by users seeking geographic information; histories of features; Capable of delivering not just current information but information on past states and In certain cases capable of making projections of future states; and Aware of the quality of its information and capable of communicating variation in quality and uncertainty to users. Topographic maps and their digital counterparts play a key role in the spatial integration of information by providing a common spatial reference framework. The National Map has expanded its role as an integrating mechanism by establishing itself as a semantic reference system (Kuhn, 2005). The explicit semantics of geographic features imparted through ontologies associated with The National Map provide this mechanism. An ontology formally defines the concepts of a domain and relationships among these concepts such that the concepts can be understood and shared by both humans and computer systems. The National Map is distinguished by a supporting geographic feature ontology that provides an authoritative specification of topographic features for the geospatial community. The National Map is further supported by a National Map gazetteer that is the most comprehensive source for geographic features names, place names, and feature footprints in the nation and by a geographic feature thesaurus. The gazetteer and feature thesaurus have been embedded in The National Map as critical pieces for its information integration role. The gazetteer provides the essential information to translate between the
OCR for page 39
A Research Agenda for Geographic Information Science at the United States Geological Survey heterogeneous location representations (place names, addresses, geocodes, spatial footprints) used by different agencies and information providers. The geographic feature thesaurus provides the essential information to translate between equivalent or synonymous terms—in this case equivalent or similar feature classes (e.g., “cove” as equivalent to “bay”). These knowledge sources not only assist human interactions with The National Map but provide the basis for the expanding machine to machine and agent interactions with The National Map. Information Demands Information demands on the new National Map range from simple fact-based queries (what is the source of the Ohio River?), data requests across USGS data assets (streamflow data from stations on first- and second-order streams in the Potomac River watershed), complex collections of data for process model input (temporally specific fuel loadings by slope and aspect face for fire models), requests for previous states or the history of a feature (e.g., flood stages of the Missouri River over the last two years), user-defined maps (e.g., a map of the hiking trail for Mt. Katahdin, including streams, points of interest, and elevation profile), to requests for traditional topographic maps. To respond nimbly to such a range of information requests, The National Map has an intelligent spatiotemporal data model based on formal specification of geographic feature concepts in the form of ontologies. The National Map has an individual feature database as well as the more traditional layer databases associated with the original version. The set of features (streams, lakes, canyons, watersheds, mountain ranges) and the feature database schema derive from specification of geographic feature ontologies (e.g., hydrographic feature ontology, terrain feature ontology). The central role of the individual feature database is to manage multiple spatial scale and temporal versions of the same geographic feature that result from observations of the feature at different levels of detail or resolution, from different sensor sources (aerial photography, light detection and ranging [LIDAR], ground based, or human), and from different times. The feature database is complimented by a layer database that organizes features seamlessly into their expected connectivity with other features. Differentiation USGS topographic maps have always maintained recognized standards of quality, and the updated National Map maintains the highest-quality data as embodied in standards for classification schemes (e.g., land use classifications, stream orders), official place names, relevant web services, (e.g., Open Geospatial Consortium standards), and cartographic display templates. In addition, the updated National Map is newly distinguished by quality-aware features that rely on the underlying feature ontologies and on the accumulation of several observations (versions) of the same feature over time. The management of multiple spatial versions of features and multiple temporal or time-stamped values for feature properties in The National Map database creates empirical distributions for a feature’s attributes, shapes, and positions. Such distributions provide a basis for identifying outliers and change, supporting temporal synchronization of feature versions and properties as well as other quality attributes. New versions of features being submitted to The National Map database can be compared against these distributions for quality assessment and recommended inclusion in or exclusion from The National Map database. The updated National Map also offers new supporting functions to respond to user needs. It employs local validation tests in which citizens can examine National Map data to verify the correctness of place names, feature classes, feature attributes, features shapes and locations, and expected relationships among features (e.g., “the courthouse is north of Main Street”). By encouraging citizen participation in local validation, the reliability of The National Map has grown along with user satisfaction. Also in the spirit of responsiveness to user needs, The National Map accommodates various map display settings from high-resolution computer screens (e.g., 5000 ×3000
OCR for page 40
A Research Agenda for Geographic Information Science at the United States Geological Survey resolution) to small portable devices such as pocket PCs or cell phone displays (with 300 ×240 resolution) and supports various mapping tasks. Hard-Copy Products The topographic map is a very visible feature of the updated National Map. Although there are an enormous number of functions and data inclusion possibilities for the power user, The National Map viewer includes a prominent “Topo Map” button that provides a wizard to guide the user through construction of a topographic map in the area of interest, at several selectable scales, and with a few key data choices for enhanced utility. The power user can manipulate the final product many more ways, and citizens and organizational users have ventured into these additional functions to produce customized topographic maps for their specific needs. Performance Evaluation The National Map is now designed for better performance evaluation. National Map web services track usage statistics including which features, features types, geographic regions, and information products (maps on demand, topographic maps, fact queries) are requested most frequently. The resulting statistics are available to evaluate usage patterns and help in the allocation of resources to higher-use, higher-demand needs. In addition, The National Map tracks demands for new data. For example, if users in a search for data (e.g., orthophotoquads for northwestern Tennessee) find them unavailable, they can log data requests. Accumulation and evaluation of these user logged data requests highlight potentially overlapping data demands and help USGS target and prioritize new data collection initiatives. Local Ownership To build a productive sense of local ownership, The National Map now encourages local submissions of geographic information (e.g., from local names and places and feature of interest to spatial locations and alignments). Every individual or autonomous sensor with a Global Positioning System receiver or other tracking device is a potential data collector; thus, a human and sensor data collection “workforce” is being tapped to contribute generally to The National Map feature database (in the vein The National Map corps suggested for the original National Map). For example, a National Map hiking trail program has solicited and supported submission of hiking trail alignments, sights, wash-outs, and other trail features captured in real time by hikers as they are hiking. Furthermore, The National Map is now rich with detailed, accurate and very fresh data from every state and virtually every county in the United States that collects and manages such data. The USGS has been successful in demonstrating to these local authorities that regular submission and updating of their data, in fact management of their data, in The National Map has enormous benefit to them in terms of standardized tools with available training; standardized formats and symbology, joint management of national emergencies such as earthquakes, tornadoes, and fires with national emergency response organizations; and access to seamless data in neighboring counties and states. With local authorities and organizations depending on The National Map for their operations, they now have a stake in its accuracy and freshness that makes The National Map the single most authoritative and useful source of geospatial data in the nation. The vision for the next generation of The National Map outlined in Box 2.2 includes a number of additional differentiators to those listed earlier for the current National Map (quality, accuracy, national coverage, standardization, and continuous updates), including the following:
OCR for page 41
A Research Agenda for Geographic Information Science at the United States Geological Survey Authoritative geographic knowledge base of topographic features based on a geographic feature ontology; Comprehensive database of official geographic feature names, and local, regional, and historic variants in The National Map gazetteer; Enhanced spatial-temporal integration framework for organization and synchronization with other USGS data collections; Geographic semantic reference system; Multiple levels of spatial detail; Feature histories (for spatial locations, attributes, names); User-supported local validation; Flexible product generation (e.g., response to fact queries, process model data packages, maps on demand, traditional topographic maps); and Smart adjustment of maps or other visual display settings for different devices. This is a challenging but necessary and achievable vision for CEGIS, NGPO, and USGS. Its realization will require focused work and resources, but a dedicated effort has the potential to ensure a continued role for USGS as the preeminent distributor of topographic information. The new knowledge base model for The National Map, enabling widely diverse queries over time as well as over geography, along with highly flexible product generation means that The National Map has gone far beyond being just a map. SUMMARY CEGIS is not lacking in potential GIScience research topics. The hard decisions facing CEGIS leadership are, (1) Which among the myriad potentially useful research topics will provide the most benefit in advancing the goals of The National Map, NGPO, and the other disciplines? and (2) What is the right research portfolio with respect to the balance among serving The National Map, NGPO, and the other disciplines? On the second question, the committee favors an initial emphasis on serving The National Map because of its strategic importance to NGPO and USGS as a whole. On the first question, and in the absence of detailed published requirements for The National Map, the committee explored the distinguishing traits of The National Map—present and future—as a guide to what traits are most worth acquiring or developing through research.
OCR for page 42
A Research Agenda for Geographic Information Science at the United States Geological Survey This page intentionally left blank.
Representative terms from entire chapter: