3
TRANSFORMATION AND TRANSITION

THE DIGITAL REVOLUTION IN SPATIAL DATA HANDLING

In the second half of the twentieth century, American society has been fundamentally and irreversibly transformed by the digital computer. As a result of its introduction and widespread use, ours has increasingly become an information-based economy. The physical resource base, so essential to prior economic development in the nineteenth and early twentieth centuries, has become relatively less important as this century draws to a close, and information as a national resource now occupies a dominant position. The implications of this for the federal establishment and for the USGS and NMD, in particular, and the traditional cartographic enterprise as a whole are significant.

Prior to the late 1950s the analog map was the primary device or tool for storing, organizing, and displaying the locations of features on the earth’s surface. While geographic representation could certainly be achieved by means of drawings, text, and numerical arrays, the printed map was generally the medium of choice.

The digital computer suddenly changed all this in two distinctly different ways. On the one hand, cartographers saw computer technology as a means for producing traditional analog maps. This was called “automated” (or “computer” or “digital”) cartography. On the other hand, researchers, government administrators, business managers, earth scientists, military planners and developers, and others) saw computer technology as a tool providing alternatives to printed



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



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 14
Spatial Data Needs: The Future of the National Mapping Program 3 TRANSFORMATION AND TRANSITION THE DIGITAL REVOLUTION IN SPATIAL DATA HANDLING In the second half of the twentieth century, American society has been fundamentally and irreversibly transformed by the digital computer. As a result of its introduction and widespread use, ours has increasingly become an information-based economy. The physical resource base, so essential to prior economic development in the nineteenth and early twentieth centuries, has become relatively less important as this century draws to a close, and information as a national resource now occupies a dominant position. The implications of this for the federal establishment and for the USGS and NMD, in particular, and the traditional cartographic enterprise as a whole are significant. Prior to the late 1950s the analog map was the primary device or tool for storing, organizing, and displaying the locations of features on the earth’s surface. While geographic representation could certainly be achieved by means of drawings, text, and numerical arrays, the printed map was generally the medium of choice. The digital computer suddenly changed all this in two distinctly different ways. On the one hand, cartographers saw computer technology as a means for producing traditional analog maps. This was called “automated” (or “computer” or “digital”) cartography. On the other hand, researchers, government administrators, business managers, earth scientists, military planners and developers, and others) saw computer technology as a tool providing alternatives to printed

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program maps. Computer technology5 would allow them to store, process, analyze, and display spatially distributed data in various forms, including but not limited to maps. This use of computer technology came to be called geographic information systems (GIS). What eventually linked these two rather disparate endeavors of automated cartography and GIS was the shared need for digital data to “fuel” the machines used to capture, store, manipulate, analyze, process, and output data as information for a given user or user community. It was fortuitous that at the time this need for digital data products was increasing, the technology of remote sensing was beginning to produce large quantities of digital data that could be employed in the cartographic process and GIS. This synergism between automated cartography, GIS, and remote sensing has had a significant impact in accelerating the pace of technological innovation and expanding the market for GIS systems. Researchers have found that map-derived digital data in a GIS can improve the accuracy of interpretation of remotely sensed data while remotely sensed data provides the ability to update map products in a more expeditious manner. The data produced by all these systems have a spatially referenced component (“earth location-referenced” might be a more accurate term). For consistency such data will be referred to throughout this report as spatially referenced digital data (srdd). This is not to say that srdd sets appropriate for creating a map are always necessarily identical to those used in a GIS; but they do share, at a minimum, coordinates that allow some “interoperability” where shared locations, expressed digitally, act as a medium for data exchange. In 1989 the distinction between the two original endeavors, automated cartography and GIS, has become blurred. The argument could be made that any distinction is no longer accurate or even appropriate and might, indeed, be creating troublesome “pseudo-issues.” It would be easy to dismiss this matter as mere word play, but the fact is that a rigorous, useful analysis of NMD’s role in an era of technological transformation and transition is impossible unless some agreement is achieved concerning the nature of NMD’s current and future role in what was formerly the “cartographic enterprise.” To do this we must continue to compare and contrast GIS with automated or digital cartography. Geographic Information Systems: Background Prior to the diffusion of computer technology into so many fields of human endeavor, discrete disciplines evolved around distinct categories of problems and tools to deal with them. Geography, cartography, geology, and related fields shared a fundamental interest in the arrangement of objects, features, and phenomena on the earth’s surface, and in the use of maps for representation (Figure 1). Many of these fields, however, can now be regarded as subsets of the superordinate field of spatial information processing. Specialty tools like the

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program FIGURE 1. Concept of separation of categories within a GIS (figure from A Process for Evaluating Geographic Information Systems, U.S. Geological Survey Open-File Report 88–105, 1988, p. 6). map or the microscope have been augmented by computer technology. It is therefore useful in this presentation to describe briefly the characteristics of what is currently called GIS. The Nature of GIS To anticipate the future nature and use of geographic data, it is important to understand the means by which these data are processed. A geographic information system has been defined as an information system that has as its primary source a base of data referenced by geographic or spatial location. Basically then a GIS is a structural approach to collecting, archiving, analyzing, manipulating, and displaying data having one or more spatial components, using a combination of personnel, equipment, computer software, and organizational procedures. Here, the major characteristics of this technology are outlined, and that outline is used to briefly consider current and future trends.

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program The field of digital geographic data processing is now over two decades old. Over the past several years, it has grown dramatically, not only in terms of the numbers of individuals involved and the variety of applications addressed, but also in terms of the sophistication of the applications addressed. Though the field has not yet reached its maturity, it probably is approaching adolescence. It is discovering new capabilities it never realized it possessed and exploring them with both awkwardness and self-confidence. It is also drawing much new attention and facing heightened expectations. A good example of the maturing status of the field is a feature article (“Managing with Electronic Maps”) in the May 1989 issue of Fortune magazine. Perhaps most importantly, the field is now developing a certain self awareness, a realization that it has a future and one that can and must be directed. It has taken GIS 20 years to become an overnight success. Much of this recent advancement in the field of geographic data processing can be attributed to more general trends in the broader field of computing. As we move from the processing of numbers to words to pictures (such as maps) and sound, certain patterns in this evolution are perceptible, perhaps most notably a shift from highly centralized computing to more decentralized but still highly integrated networks. Developments associated with civil geographic data processing in particular have been driven to a large extent by practitioners in fields such as land-use planning, hazardous waste routing and disposal, and scientific discipline with spatial analysis needs. Initially many of these developments were associated with academia, but in recent years the field has moved out of academia into an active commercial setting where integrated GIS “toolboxes” for general applications are now offered. One way to characterize these developments is in terms of the three major components of a GIS. Like those of any information-processing system, these include a body of data, a means of processing those data, and a mechanism to control that processing. The data that tend to be processed by geographic information systems describe phenomena not only in terms of what and when, but also where. The magnitude of this locational component may be measured in units that range from centimeters to thousands of kilometers, and the ability to transform data from one scale to another is an important part of the geographic data processing. This is not simply a matter of changing the size of a particular graphic product, but a matter of moving accurately between detailed and generalized data bases with efficiency and consistency. A problem of accuracy is often encountered in moving from one scale of data, where certain generalizations might have been made, to other scales more appropriate for the application being pursued. The way in which geographic data are organized in a GIS can generally be expressed in terms of the way in which facts pertaining to what (theme), when

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program (time), and where (location) are respectively either held constant, allowed to vary in a controlled manner, or measured. A railroad timetable, for example, records time (of departure) as a function of location (station) for a constant theme (train). Most traditional maps, on the other hand, record theme as a function of location at a constant time. Data used in GIS have also traditionally been organized in this manner. Recently, however, new organizational schemes have begun to emerge. A number of systems have moved from location-oriented schemes to feature-oriented structures that record location as a function of theme at a constant point in time. Some systems are also beginning to address the temporal dimension of data as more than a constant. In short, the traditional “map” format is becoming just one of many alternative ways to organize geographic data. The ability to easily reconfigure data and to flexibly associate any one piece of information with others is coming to be regarded as a standard feature of any data management system. The way in which geographic data are actually represented (stored, manipulated, and displayed) in a GIS may also vary from one system to another. Most of this variation relates not so much to the representation of what or when as where. Some encoding schemes associate theme and location on an atomistic basis. They refer to elemental pieces or “atoms” of cartographic space. Other data encoding schemes are more holistic in nature. They associate theme and location by way of cartographic “wholes.” Raster (gridded image) and vector (line drawn) data structures epitomize this distinction with atomic grid cells and holistic sets of points, lines, and polygons. The distinction is beginning to fade, however, as raster resolution improves and as the ability to convert from one form to the other becomes routine. In terms of the ways in which these data are processed, GIS technology is still at a point where methods are modeled after traditional techniques. Though our ability to do more work more rapidly and more economically is something we have come to expect, the new technology is most often used not to do new and different things but merely to do familiar things better. This process, however, is also changing. We are rapidly approaching a time when the primary processor of geographic data is no longer a human being looking at a piece of paper but a machine assimilating data in ways for which there is no human analogue. The data-processing capabilities of a GIS can nonetheless still be characterized in terms of four major types of traditional activity. They include programming, data preparation, data interpretation, and data presentation. The programming capabilities of a GIS are those that affect the way in which a system is operated in general. These are capabilities associated with user interaction, program execution, error handling, and so on. Development work in this area involves concerns that are common to all types of computing

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program such as security, large-volume data handling, multiple and concurrent access to common data bases, updating, and data management in general. The data preparation capabilities of a GIS are those that provide for the acquisition, encoding, storage, and routine maintenance of geographic data. These may range from field investigations to digitizing to the reformatting of data. While capabilities in this area have improved remarkably within the past several years, this remains a “traffic jam on the way to the airport.” The data interpretation capabilities of a GIS are those that provide for the transformation of data into information. This transformation generally involves a process in which facts of a general nature and potential utility (i.e., data) are translated into facts of a more specialized nature and actual utility (i.e., information) in answering questions, making decisions, or otherwise solving problems (Figure 2). For many applications, map reading has been replaced by interactive measurement, queries, and display functions. The interpretative process typically involves both objective measurement and subjective judgment to transform facts, relationships, and/or meanings from an implicit to an explicit form. Interpretation of vegetation types, for example, might bring out facts pertaining to ecological characteristics. Interpretation of a topographic surface might rely on geometric relationships to infer a drainage pattern. The data presentation capabilities of a GIS are those that provide for communication of facts to a general audience. This may involve maps, charts, reports, statistics, animations, and so on. Here, too, recent developments have been remarkable, and most current needs are generally being met. Advances in this area, however, continue to be made with an increasing emphasis on visualization and the use of sound, motion, and whole-environment simulation. It is the orientation toward data interpretation that distinguishes geographic information systems from other types of automated mapping systems. While mapping systems are primarily concerned with data preparation and presentation, the GIS is more concerned with data interpretation associated with on-the-ground applications. As the GIS user community has become more and more sophisticated in its ability to prepare and present its digital data, attention has now begun to focus on data interpretation and the techniques by which environmental phenomena can be modeled. Modeling applications may range from land-use planning and environmental science to navigation and marketing. One way to characterize their range is to draw a broad distinction between descriptive and prescriptive models. While the former deal with the realm of what is (e.g., landscape ecology or economic analysis), the latter deal with the various arts that focus on what should be (e.g., natural resource management or transportation planning). Modeling capabilities in both of these generic areas are becoming more sophisticated and more accessible. They are becoming established tools for determining the capability

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program FIGURE 2. A decision support system using GIS (figure courtesy of D.Cowen). or suitability of land for different purposes. Perhaps the most notable trends are toward increasing dynamic visualization (i.e., simulated movement over time and/or space) and artificial intelligence (e.g., natural language interfaces, learning, and a mechanized ability to draw inferences for later use). A third component of a GIS is its capability for user interaction. The introduction of computer technology has, for this reason, triggered a good many changes in institutional relationships in the cartographic/spatial data handling communities. Many such alterations in relationships, including shifts between private and public sector responsibilities and opportunities, have yet to be fully evaluated. As Kling and Scacchi6 write: “Many of the difficulties users face in exploiting computer-based systems lie in the way in which the technology is embedded in a complex set of social relationships.” GIS is not simply a neutral technology change; it creates an imperative for and may indeed require social and institutional change as well.

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program Computer Technology and Cartography: Background By the late 1960s, cartographers began to think of applying computer technology to the problem of geographic representation but in a much more restricted way than did the researchers and government administrators who were simultaneously creating the phenomenon of GIS. The U.S. military mapping and charting organizations used computers for most of their photogrammetric data extraction operations starting in the late 1950s. By the 1960s computer-driven hardware such as UNAMACES and AS-11s were the primary tools for photogrammetric compilation supporting both military operational needs and some of NASA’s manned spacecraft operations. However, many mapmakers saw the computer as a tool that would substitute for human labor, thereby “automating” the production or drafting of analog maps. As originally envisioned, digital representations of manuscript maps would be created by the process of line-following digitization, or scanning, and computers would then carry out the labor-intensive processes of scribing, typing, and color-separating to produce plate-ready printing film. It has taken quite a long time to work out all the details of this approach in a cost-effective manner, and indeed the process cannot really said to be complete in many map production applications to this day. For example, issues such as label placement, line generalization, and scale change still need additional work. Meanwhile, several interesting, significant and complex changes occurred that are creating a need for fairly dramatic restructuring of the cartographic enterprise in the broadest sense and of the USGS’s National Mapping Program more specifically. By the 1980s it became apparent to many researchers and operational map producers and users alike that the spatially referenced digital data streams that were essential for running geographic information systems and the srdd streams that were needed to run computer-based mapping systems could be one and the same thing. In most cases data sets for GIS were initially generated from existing maps, but given the relative ease with which all digital data (if it is coordinate-based) can be combined, data sets could now be assembled from many diverse sources, including photography, global positioning systems, electronic imaging (often from a remote sensing platform), newly created maps, and so on. NMD AND DIGITAL CARTOGRAPHY To be more specific, NMD now finds itself in the position of creating digital data sets to produce and update print maps. But meanwhile many of the traditional users of these maps have become GIS users and are aggressively transforming USGS maps into digital data; they now want to augment their data sets

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program directly from NMD digital data, not from maps. A major and complex transformation of a significant segment of the user community has thus occurred, and NMD must come to terms with it. Many of the significant USGS map users of earlier decades are now GIS users; they want srdd, not maps. Furthermore, many of them now produce srdd that “registers” to the traditional USGS/NMD base data and could conceivably be added to the storehouse of map-making digital data that NMD is mandated to produce. It is now possible to imagine geographic information flows inward from data donors to NMD’s data sets as well as the traditional outward flow from the National Mapping Program on a routine basis. Spatial Information and the Economy The two-way flow of spatial data discussed above is a revolutionary concept and will require significantly altered conceptual and technical structures as well as revised thinking on institutional and national economic issues. The challenge to NMD, to the USGS, and to the administration and Congress is this: If ours is to be an information-based economy that is competitive on a global basis, there is a critical need for a coordinated and efficient national information “infrastructure” to facilitate the sharing and communication of information resources. One component of this information must deal with where things are, that is, it must be a geographic information infrastructure, to support all manner of resource, transportation, planning, administration, marketing, and communication activities. In the future, map-making will be just one aspect of a larger enterprise—one focused on acquiring and manipulating srdd in various forms to solve problems and meet spatial information needs. Producing a hard-copy map from digital data will still be useful, but there will be times when an ephemeral map on a video display will provide an adequate or a better means of visualization. And still other needs will best be met with numbers or text. Shifts in User Requirements The committee finds that traditional map-using constituencies for the National Mapping Program are changing in ways that go beyond the shift from analog (print maps) to digital data, ways that relate to computer capabilities for handling geographic information. In the past, society “made do” with the shortcomings of maps because there was no better alternative technology. For example, because of the time involved to produce high-quality topographic maps, users became accustomed to getting maps that often were significantly out of date by the time they were published.

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program Furthermore, since only a limited amount of detail could be printed legibly on maps at particular scales, this detail was limited and rather rigidly formalized. Because a national printed map series must be reasonably uniform in content and detail across all areas, feature “textures” (which differ significantly among regions as well as between rural and urban landscapes) were often compromised. Tradition, too, has played a role in making the print products of the National Mapping Program less useful in 1990 than they were several decades earlier. The topographic maps of the 1980s are remarkably similar to those made a century ago. Terrain is represented in considerable detail, with contours often dominating the print map image. Yet all streets are not labeled. Perhaps most inflexible of all is the fixed areal unit of a map sheet or quadrangle. Human activities focus on quite different units (political, administrative, and property units—cities, counties, school districts, and so on). Piecing map sheets of different dates together in a mosaic to accommodate different areal units has been a major inconvenience and a source of frustration for these users. Such printed map limitations (and this is only a sample) can now be effectively overcome where digital data sets can be created to drive computer-based systems. It becomes quite clear that the national need for srdd is far larger than has ever existed before. To some degree it can be met by suppliers in the private sector. But because the demand for geographic data and base data consistency is so vast it appears to the committee that the most important function of the USGS/NMD in the future might be not to produce maps or even digital data, but to act as the interdepartmental administrator of the national geographic data infrastructure. Given the flexibility of transferring, accumulating, and processing data in the digital mode, the concept of a one-directional information flow out from NMD to the world has become outmoded. A major role could now be played by NMD in defining and creating (or causing to be created) the essential geographic base data to which much subsequent observation about spatially ordered features and phenomena can be referenced. It is apparent from the various briefings and presentations made to the committee that the USGS/NMD leadership is keenly aware of the transformation in its user community and the need for the USGS/NMD to transform its programs and even its organizational structure to respond more effectively. Just how adequate its current efforts are is a more difficult but essential question. To respond to its charge, the committee attempted to determine just how NMD is evaluated by its users in regard to its products, services, and programs during this critical transitional era in geographic or spatial data handling. There are several ways to approach the problem. First, the committee documents briefly the changes in requirements that have developed in the public sector, federal and elsewhere, among traditional map users. Second, it looks at what is happening to geographic or spatial information requirements in the

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program private sector, both Type A and Type B businesses. Third, it looks at certain national spatial data needs that are only just emerging in our economy as a result of technological, computer-related developments. These are needs for which there were no strictly comparable precedents in the traditional map era, and which the National Mapping Program as currently structured does not meet. In a sense these are “nonuser” map requirements, and we are witnessing the potential or real transformation of nonusers to users, related to spatial digital data requirements. Finally, the committee examines some of the steps NMD has already taken to adjust to the transformations in technology and user communities that have been described. TRANSFORMATION OF PUBLIC SECTOR REQUIREMENTS Much significant technological innovation in this country is driven largely by two basic categories or domains of need. One is defense needs and the other is a need for businesses to sell products and services to large consumer markets. As it happens, both of these have the potential to have a significant impact on the cartographic enterprise in general and on the USGS/NMD in particular. Military Requirements The Department of Defense has been the single largest consumer of NMD maps. But its systems and weapons are increasingly computer based and often require srdd to function. The Defense Mapping Agency (DMA) made early and significant commitments to both digital cartography and GIS. At this time, over two thirds of its resources are devoted to products that are digital in nature. It seems logical to assume that for many domestic applications DMA would rather have digital files from NMD than printed maps. In any case it is essential that digital data standards be established and maintained to allow the data to be transferred between and used in both srdd producing and user organizations (both military and civilian). DMA is developing sophisticated and expensive computer systems to meet both map production needs and digital data requirements for GIS and weapon systems. NMD is cooperating with DMA on this technology to acquire similar benefits at much lower costs, with a major new system development effort called Mark II that will implement advanced technologies and production procedures to satisfy National Mapping Program requirements through the year 2000.7 Mark II is designed to collect, populate, and maintain the National Digital

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program FIGURE 4. Concept of a national digital spatial data base (1:24,000).

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program standards. Nationwide, we can expect to see substantial redundancy and excessive cost if such uncoordinated, unstandardized, localized data base building continues unchecked. TRANSFORMATION OF PRIVATE SECTOR REQUIREMENTS As noted, the National Mapping Program was not created to meet the spatial information needs of firms and individuals in the private sector. It has traditionally been assumed that pricing mechanisms would evolve in a free market economy such that many, if not most map/srdd needs could be met by profit-making firms. The current pricing policy tends to affect commercial firms somewhat differentially, however, with certain businesses (e.g., the petroleum industry) taking significantly greater advantage of the 1:24,000 coverage simply because their scale/content/coverage needs most closely resemble those of federal resource management agencies. Firms in transportation and marketing, at the other extreme, require spatial information about categories of landscape features that even now are not conceived of as “base” information to be shown on topographic maps (e.g., complete street networks with names and block addresses and ZIP codes). Ironically, this same level of detail is required by multiple federal agencies, many whom are nonresponders to the A-16 process. Somewhere between these extremes are businesses that require far more detail (scale) about both man-made features and the physical surface than is shown on 1:24,000 topographic maps, similar in many respects to the needs of local governments. These include utility companies, land-use planners, construction firms, and so on. In many of these situations the geodetic control and feature representations on the 1:24,000 quadrangles serves as an anchoring network to which more detailed mapping is referenced. This large-scale mapping creates finer textured data that fills in the spaces that are “empty” or characterized by features at a higher level of abstraction on the 1:24,000 printed maps, such as land use. Additional Base Data Needs There are two important issues that must be recognized in connection with user requirements for NMD maps and spatial information, realities not yet dealt with in an explicit fashion. One is the historical distinction made at the federal level between “base” information and street/highway information on maps. The other (and somewhat related matter) is the traditional distinction made between

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program what spatial information was to be gathered and mapped at public expense, and what by commercial firms. Despite the fact that in the late twentieth century the data on 1:24,000 topographic maps that is probably most used as base data (i.e., data to which other locations can be referenced) is the transportation network of streets, roads, and highways, NMD is not mandated to give the labeling and updating of this feature class a high priority. As this nation has evolved from an economy that is largely physical resource based to one that is transportation, information, and service based, certain aspects of traditional topographic mapping have become less important while others (streets) have become increasingly desirable. This situation may originate from the fact that the Department of Transportation (DOT) is a separate entity from Interior, and its Federal Highway Administration has responsibility for mapping the nation’s road surfaces. Much of the mapping it funds is actually carried out by state and local government agencies. In the past such localization of activity made sense, because cities, county, states, and so on were relatively discrete entities. There was no particular need for a national street map, and one was not created. This is in contrast to the 1:24,000 series, which, in effect, creates a national map of terrain and hydrography, among other things. But as more and more geographic information is stored and integrated in GIS for all manner of public and private sector entities across the country, this lack of a detailed, current national street map to use as base information is becoming a serious problem. When user groups were queried about satisfaction with USGS products (described above), the chief complaint had to do with outdated information. It is unlikely that they were referring to contours and hydrography; in most situations they were referring to the urban detail, including streets and highways and built-up areas. It is not only the increased use of GIS that makes the need for upgrading information about the nation’s transportation network as a whole more urgent. Rather, more and more key businesses are operating nationally (even globally), and localized, nonstandard geographic information is not adequate as informational infrastructure for these components of the economy. As an editorial in American Demographics10 puts it, firms like Federal Express and UPS are the “rivers and railroads of our age.” And a DOT-issued “Discussion Paper on Intelligent Vehicle-Highway Systems”11 reinforces this point: “There is a direct relationship between efficient commercial transportation and the nation’s economic vitality.” This is not to suggest, necessarily, that the USGS take responsibility for all aspects of representing the nation’s transportation surfaces. But it points up the need for more comprehensive planning and coordination among agencies with responsibility for spatial feature information in the digital era than was ever conceived of as essential or even useful in the printed map era.

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program Somewhat related to this is the second issue that must be considered in this era of transformation and transition: the relationships between public and private sector participants in the traditional cartographic enterprise. Private/Public Sector Relationships in Mapping In the past a rather thin, sharp line (at least in theory) separated mapping that was appropriately done at public expense and that which was carried out in the private sector. It was acknowledged that governmental units at all levels needed to make maps to keep track of their territories. If these maps could be used for other purposes in the private sector, they were. If not, and if firms needed spatial information to operate their businesses or to create other maps for publication, it was assumed they would pay to make their own and retain the information as proprietary. In summary, private sector mapping has usually been done where more detailed information is needed or where value-adding processes can transform public sector data into proprietary products. Commercial mapping firms have been traditionally well organized politically to protect their “right” to do any mapping at scales larger than 1:24,000. At the same time and continuing today, many of these private enterprises are involved in the mapping programs of federal and nonfederal agencies. The expanding use of GIS technology is creating demands for more detailed uniform information over ever-larger areas. Such data needs will not be met economically if many individual GIS users are obtaining the same information, independent of one another. The need is now national in scope, and it seems reasonable to consider the possibility of meeting these emerging needs at the federal level. Direct federal responsibility is probably not the solution, but cooperative, directed funding of some sort may be in order to develop shared data bases of widespread utility with national standards. It is already the case that several inconsistent digital street data bases exist for the major metropolitan areas. The institutional structures for widespread cooperation between the public and private sectors are not in place, but in a society whose economic well-being and ability to compete globally depend on effective, high-quality information, it may be that traditional arrangements should be restructured. It is time for the administration to require greater interdepartmental cooperation and to designate a lead agency for GIS. Here, as in many other applications, we find at the heart of the matter increasingly complex systems in place in our society. But the information to support these systems (an informational infrastructure) is fragmented by tradition and by earlier technological limitations. The technology has changed to allow complex systems to develop: this is an appropriate time to acknowledge that institutions and organizations involved in this informational infrastructure will also have to change.

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program Other User Requirements Individuals in this country buy a good many topographic maps each year, and form a supportive constituency for NMD’s products. It is likely that many of these purchases are made for recreational purposes: for boating, hiking, fishing trips, or others where map usage is too low to support commercial mapping. In a sense these users are the beneficiaries of the “goodies over the fence” principle. NMD does not track map sales in adequate detail. In 1987 it sold more than $8 million worth of printed maps, of which $1.64 million of purchase were made by federal buyers, and $6.63 million were made by the public. It is, as the NMD Marketing Plan for FY 198812 notes,“…ironic that by mission, NMD cannot be completely responsive to its greatest source of revenue.” It is unlikely that individual map user requirements will be much affected by the computer at this time. The portability and flexibility of paper maps to meet fairly simple informational needs is hard to beat, at least with the electronic technology available today. This may be changing quickly, however, with new programs under way in the Bureau of the Census and the British Ordnance Survey. Both of these organizations are beginning to use electrostatic plotters to generate custom maps. Even the casual map user would like a customized map centered on the area of interest, at an appropriate scale. Finally, the committee conducted interviews (Appendix B) with seven mapping firms in the private sector, three of which are Type B (creating maps as tools for clients) and four Type A (publishing maps as end products to sell to mass markets). It determined that such firms do not rely heavily on NMD products as source material, especially the 1:24,000 quadrangles. The amounts of money these firms spend annually at the USGS is rather small, ranging from $300 to $1200 per firm. While respondents praised NMD maps for accuracy, consistency, and availability (no copyright), they often find the maps too out-of-date (one assumes they refer here to built-up areas, street names, and points of interest). The firms queried indicated that they expect to move eventually to digital operations, but are doing so slowly. As might be expected, all respondents want the federal government to limit its mapping activities to producing “general” maps for use by the general public, staying away from any mapping that would place it in competition with the private sector. This competition may be ambiguous in practice, as is the private/public distinction in mapping generally. On a less structured basis, committee members obtained some information about requirements for spatial information from firms in the utility and transportation industries who want very detailed information about large (multistate or national) areas. Clearly, current NMD products do not meet their needs, needs that are now generally digital as well. Such firms are, of necessity, embarking on the development of extremely expensive proprietary data bases. There is consid-

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program erable evidence that they would welcome cooperative or data exchange programs with the federal government,13 but, as has been observed above, institutional structures to facilitate such cooperation are not yet in place. The lack of a data donor program requires urgent, imaginative attention from Congress and the administration. USGS/NMD RESPONSES TO TECHNOLOGICAL TRANSFORMATION The transformation of user requirements from analog to digital spatial data precipitated by the widespread adoption of GIS technology throughout the public and private sectors has not gone unnoticed by the USGS. Nor has the potential of the computer to provide more current printed maps been overlooked. Many aspects of NMD’s operations and structure have undergone significant changes in the past decade because it has recognized shifts in all type of user requirements. It is essential to review the progress NMD has made to date in order to offer relevant guidance as requested in the initial charges to the committee. A brief listing of NMD products, services, programs, and structure that reflect or appear to be direct responses to the computer-based transformation in spatial data handling follows: NMD has been collecting digital data from the cartographic source materials produced by the division since the mid-1970s. The cartographic community respects the leadership the USGS/NMD has provided in the early experimental stages of digital cartography and GIS development. The National Mapping Program now includes digital cartographic data (that is, information digitized from, or as a by-product from, making its printed maps) at various scales. Probably the 1:2,000,000 and the 1:100,000 (TIGER) linework files are the most used digital data sets produced to date. The NMD has created Geographic Names Information Systems products, available digitally. The NMD has been active in collecting, processing, and disseminating a variety of nonmap imagery and digital elevation models to a diverse user community. The NMD information dissemination services include, to some extent, data that are digital. The monthly periodical, New Publications of the U.S. Geological Survey, for example, describes national inventories of digital spatial data and cartographic applications software that are available from the USGS’s Earth Science Information Centers (ESIC). The NMD has been assigned lead responsibility for GIS research and development activities that include the Federal Land Information System and

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program cooperative GIS applications and projects. The NMD has begun a major new system development effort called Mark II that will implement advanced technologies and production procedures to satisfy National Mapping Program requirements. More specifically, NMD is creating the National Digital Cartographic Data Base (NDCDB) as the central focus of its activities. It is planned that Mark II will be fully implemented during the 1990s, to allow fully digital revision and production of printed 1:24,000 maps by the year 2000. The NDCDB, as currently envisioned, will include topologically structured digital representations of the printed maps now in the National Mapping Program at the scales of 1:24,000, 1:100,000, and 1:2,000,000. From these digital files NMD intends to support the production of special-purpose maps at scales from 1:10,000 to national map atlas scales. It should be noted that the NDCDB is therefore actually three distinct data bases, with individual features represented differently in each, depending on the scale of the source map, though feature correspondences among the data bases will be incorporated “if feasible.” As Olson and Callahan7 comment: “A scale-independent data base, with data from any number of sources including maps of varying scales, capable of making any product is considered to be beyond the Mark II timeframe.” The NDCDB is to serve two major functions:7 (1) a working data base for production of standard USGS graphic products, and (2) a central archives for the dissemination of digital data to the user community for information systems analysis. An OMB director’s memorandum of April 4,1983, created the mandate for a Federal Interagency Coordinating Committee on Digital Cartography (FICCDC). FICCDC consists of a steering committee and five working groups dealing with requirements, standards, user applications, technology exchange, and reports, all with multiagency representation. FICCDC is chaired by the USGS. The USGS also chairs the Interior Digital Cartography Coordinating Committee (IDCCC) and actively supports the work of the National Committee for Digital Cartographic Data Standards (NCDCDS), a public/private sector group. The 1987 FICCDC report, Summary of GIS Activities in the Federal Government14 recognizes the need for federal coordination to, in effect, create a data base that transcends digital cartography needs to include thematic data sets that will be useful to a wide variety of GIS users. That report states: The desire to exchange digital cartographic and geographic data and use the data in geographic information systems is increasing the benefits for federal agencies to create data bases that are compatible and can be integrated to support decision making processes. The FICCDC has begun working with agencies to establish national lead-

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program ership for coordinating the collection, storage and distribution of cartographic and geographic data that are national in scope and have multi-agency utility. In addition to the NDCDB managed by the USGS, other major digital data bases are being developed to standards that will enable multiple use of an agency’s data. Clearly, NMD is charged with major responsibility for providing advisory leadership among all the federal agencies that generate and use digital spatial data. It does not, however, have the authority to enforce recommended standards or procedures, and this limits the effectiveness of its recommendations. Evidence of increased USGS/NMD concern with user requirements in relation to its internal production policies and procedures is contained in the Primary Mapping Economic Analysis,2 a study conducted in 1987 and 1988 to estimate benefits and costs associated with different USGS map revision production levels. This study was conducted, according to the director of the U.S. Geological Survey, as “…part of a continuing effort to improve our understanding of the many ways Geological Survey maps are used and to gain better insights into the benefits of these products” (Dallas L.Peck, Foreword, Phase II report). The standard A-16 process, already discussed, is one way in which NMD evaluates federal user requirements, both conventional and digital. But Appendix C, prepared by NMD, provides information about other related activities that go beyond A-16. “In addition, requirements are gathered through other coordination mechanisms, such as FICCDC and IDCCC topical meetings, technical exchange meetings, bl-lateral coordination committee meetings, in writing, over the telephone, through congressional inquiries, legislation, and regulatory documents” (Appendix C). An early 1989 reorganization of NMD also demonstrates an increasing formalized attention to user requirements. Coordination and requirement functions have been elevated to the assistant division chief level and are carried out within the Office of External Coordination. Responsibilities falling in this new area include FICCDC and IDCCC support, providing requested focus/planning input to three-year plans, A-16 coordination, international activities liaison coordination, and data bases/management information systems oversight. In 1987 the A-16 requirements solicitation process was extended from approximately 40 federal agencies to include all 50 states. This is not an exhaustive list but does serve to show that in many respects the USGS/NMD has demonstrated considerable awareness of and has some mechanisms in place to respond to the dramatic shifts in user requirements that have occurred in recent decades. This awareness, however, has not yet led to a robust research program—a research program that keeps pace with expanding

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program technological potentials and the increased demand for a diverse array of new products. RESEARCH The USGS has a tradition of in-house research dating back more than a century. Over the years NMD has focused on the more applied, operational aspects of the cartographic process, for good reasons. However, recent developments in digital spatial data production and handling suggest there is a need for a change in NMD’s research priorities. NMD research activities should be given greater priority and higher visibility and should be significantly expanded. NMD research should span the spectrum from applied to fundamental, from improved methods for spatially modeling land cover changes to advanced visualization. Research is also required that will provide staff with expertise allowing adequate evaluation of contractor offerings of advanced hardware and software for managing srdd. The key is balance, that is, balance between the need to keep abreast of ever more sophisticated user requirements for data and the rapidly developing technology used for digital data production and the processing of cartographic data. Through the conduct of in-house research, NMD personnel gain increased insights into those issues that could enhance the operational mandate of the division. These insights range from hardware capabilities to software potential, from science needs to technological alternatives. The committee believes that unless NMD maintains a leadership role and a strong research presence within the national srdd infrastructure, the nation as a whole will suffer. Accurate up-to-date cartographic products (digital or analog) must form the basis upon which physical, cultural, and natural resources management decisions of the future are made. In an era of increased awareness of global change (e.g., climatic warming, tropical deforestation, and reduced biological diversity), the scientific community will need to employ new techniques and methodologies for enhancing sustainable development on national, continental, and global scales—advanced cartographic research is imperative. Mapping is the key. Without accurate maps we cannot hope to understand the dynamic social and environmental changes that are occurring in our own country let alone the global system. While accurate maps of a limited set of features exist for many developed nations today, the timeliness of their overall information content is often suspect. For most areas of the developing world, accurate maps of key land-cover parameters do not exist. Accurate, timely maps facilitate the measurement of important socioeconomic, biophysical, and geopolitical phenomena (see Figure 5). While global mapping is not NMD’s primary mandate, expanding this nation’s mapping research capability will provide substantial benefits not only to the

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program FIGURE 5. Measurement, mapping, monitoring, and modeling of environmental features and processes can be enhanced through the use of a geographic information system (after Star and Estes, 1989, Geographic Information Systems: An Introduction, Prentice Hall, Engelwood Cliffs, N.J.). national mapping infrastructure, but also to the science and resource management communities. To provide such assistance, NMD must not only conduct in-house research but also expand its existing contacts with industry and significantly increase its interactions with the university community. The committee is concerned that in recent years a manpower shortage of trained cartographers in the United States has appeared. Universities are having trouble recruiting qualified faculty in the mapping sciences. The costs of keeping pace with advances in technology are escalating dramatically, making it increasingly difficult for universities to provide advanced training. Few federal agencies are funding the university mapping research community, whereas private industry has become the training ground for a new generation of cartographers. While this may be acceptable to some extent, the need for balance between fundamental inquiry and operationally oriented applied research is probably best achieved if all elements of the national mapping infrastructure (government, private industry, and universities) are healthy.

OCR for page 14
Spatial Data Needs: The Future of the National Mapping Program The USGS/NMD should consider appointing a committee composed of USGS, other federal agencies, private industry, and university personnel to assist in the development of an outline for a program balanced between fundamental and applied research. This program should expand upon current NMD fundamental and applied research and should specifically address needs in areas such as (1) digital cartography; (2) geographic information systems; (3) image processing and analysis; and, (4) remote sensing. Examples of more specific research themes in these areas include digital spatial data modeling, hardware and software development, land-use/land-cover analysis, and the monitoring of global change. While there is some overlap among these themes, they do present some general categories around which an applied and fundamental research program could be structured.