Toward a Coordinated Spatial Data Infrastructure for the Nation (1993)

It is important to conceptualize the NSDI in the broad sense of our definition. However, when the committee received briefings from various agencies and studied the documentation provided, an important area of opportunity concerned data. The issues concerning the reange of spatial data are complex and hence we offer no panaceas, but we have attempted to discuss some of the impediments and improvements to using and sharing data in the NSDI.

The data in the NSDI exist in diverse forms and reside in the analog and digital spatial data bases of various federal, state, local, and private agencies. Parts of these data exist as hard copy maps and charts created and periodically updated to meet a defined need. Other parts are stored as reports and studies that summarize the spatial data standards, needs, uses, and controls of the federal, state, local, and private sectors. Improvements can be made in the underlying structure, procedures, and standards that would allow for easy exchange of these data.

It is important to note that data sharing has occurred. When needed data were found to be unavailable, an organization or agency would set about collecting them. If all or part of the needed data were available from another agency, a data exchange was often set up. When agencies found they were responsible for data collection over the same areas, coproducer agreements were set up where feasible. However, most data collection overlaps were resolved by each agency collecting its own data over the same area, ostensibly to meet needed accuracy and currency requirements. The justifications for this duplication in data collection usually fell into three categories: (1) it would be too expensive to collect the level of data needed by one agency to satisfy the needs of a second; (2) the accuracy or

level of aggregation required by the first agency could not be achieved by the data collection capabilities of the second agency; and (3) the data could not be collected in the time frame required by the other agency.

With the advent of GIS, an extremely versatile, efficient tool has become available for manipulating spatial data. The diverse types of GIS applications have driven the need for more timely and accurate spatial data, particularly in digital form. This also has created the need data base products that are adapatable to rapidly changing user needs.

As federal agencies discovered the value of using GIS for storing and manipulating spatial data, budgets were constructed to purchase such equipment and to collect the needed data in digital form. Early efforts resulted in a myriad of data formats, standards, and processing algorithms. In effect, each federal agency repeated what it had done with paper products, only in digital form.

The cost (in staff hours) of collecting spatial data in digital form is projected to be half the cost of collecting data by manual methods for the generation of individual products (DMA, 1991). Once collected, though, the value of such digital data increases manyfold. The ability to extract subsets, generalize, and thin the data; increase its densification; or merge (fuse) it with other spatial data has created such exciting applications as emergency response (911) location systems, integrated land use/transportation planning models, battlefield management systems, crop rotation forecasts, and environmental impact assessments.

While reviewing the spatial data activities of several federal agencies, the MSC recognized a number of general issues and impediments that need to be resolved to build a more robust NSDI.

ISSUE 1: There is no agreed-upon national vision of the NSDI nor is there an apparatus to implement it. Consequently, there is no national policy covering spatial data nor is there a national organization or agency with the charter, authority, and vision to provide leadership of the nation's spatial data collection, use, and exchange.

Each federal agency with a responsibility for collecting spatial data traces this responsibility to the fulfillment of its primary mission. For example, the spatial data collected by the USGS satisfy its requirement to

produce specific products from the collected data. Agencies must focus their funding toward satisfying their primary mission and mandates, and few, if any, resources are left to support government-wide spatial data oversight activities. In fact, the current cooperative efforts between agencies are more a result of bilateral agreements for data exchange than a concern for a healthy NSDI.

The Federal Geographic Data Committee (FGDC) was established by revised OMB Circular A-16 as an interagency coordinating committee on geographic data matters. However, the present direction and organization of the FGDC have problems that inhibit its effectiveness. The FGDC has no charter to review the spatial data programs of its members and no power to enforce decisions. Because there is always some resistance to change, even FGDC recommendations on spatial data content and format take time to implement. The federal agency members of the FGDC steering committee have varied interest and involvement in collecting and applying spatial data. Some, in fact, seem to have a vested interest in maintaining the status quo. Finally, the agency representatives are not detailed to the FGDC as their primary mission, and their job commitments remain with their representative agencies, not with the FGDC.

ISSUE 2: Because of the lack of central oversight, there appears to be extensive overlap and duplication in spatial data collection at the federal level. Overlap in data collection also appears to occur between federal and state agencies, and among state, local, and private sector organizations, all at a significant cost to the public. These institutions are collecting spatial data at many scales, levels of accuracy, levels of detail, and categories of data, making the sharing of spatial data very difficult (if not impossible).

As shown in Table 4.1, many federal agencies are responsible for collecting spatial data of the same type over the same areas of the country. In the area of wetlands data collection (see Chapter 6 and Appendix A), for example, overlap of data collection is significant.

Additional redundancies in data collection result from the diversity of definitions used for various classes of data. For instance, the Soil Conservation Service (SCS) and the Agricultural Stabilization and Conservation Service (ASCS), using the National Resource Inventory (NRI), collect resource data on a sample basis for all non-federal lands. The Bureau of Land Management (BLM) and the U.S. Forest Service (USFS) collect similar data for the land under their jurisdictions. The Environmen

tal Protection Agency (EPA) has proposed to collect a nationwide sample of environmental conditions. All these agencies and programs use different primary sampling units and slightly different types of data. Data delineating wetlands are being collected by the Fish and Wildlife Service (FWS), the EPA, the SCS, the USFS, the National Oceanic and Atmospheric Administration (NOAA), the USGS, and many state agencies. However, there is still no uniform approach to the definitions and mapping conventions among these programs.

As mentioned, the primary data collection role of most federal agencies is to supply data products to meet their own missions and mandates. These agencies create products of differing scale and content and use different collection techniques and source materials. Often, such data are incompatible with similar data collected for other products without algorithms to thin, match, feather, generalize, densify, remensurate, rescale, or aggregate data categories. This problem is further complicated by the variances in the underlying base data, non-standard feature identifiers, and GIS hardware and software used to store and manipulate the data.

ISSUE 3: There are no current mechanisms that allow identification of what spatial data have been collected, where the data are stored, who controls the access to the data, the content of the data, and the data coverage (e.g., scale, data density).

Each federal agency controls its own spatial data as do state and local governments and private sector organizations. Some agencies, such as the USGS and the Bureau of the Census, have well-publicized mechanisms for obtaining their data products. Others, such as the EPA, make their data available to qualified users through an on-line system. Some agencies have yet to institute any type of formal mechanism for distributing their data. These data often can be obtained only by dealing with the person who controls the data within a specific agency. In summary, there is no single system that provides a catalog or directory of the spatial data holdings of all federal agencies. In January 1993, the Federal Geographic Data Committee printed a Manual of Federal Geographic Data Products, which represents a positive first step in the compiling the various federal spatial data holdings.

Even if a comprehensive spatial data catalog could be developed for the federal agencies, the extension of such a system to state and local government holdings would be very difficult because to its potential size.

Extension to the private sector might further complicate the process by introducing proprietary restrictions.

There is technology available to provide a distributed, on-line system that provides access to spatial data catalogs at various organizations. One such system is a public-domain software program, Wide Area Information Servers (WAIS). WAIS, which runs on the Internet network, provides the capability to scan, search, and often access existing data bases. WAIS could be easily expanded to encompass spatial data catalogs. (See Chapter 8 for additional information on WAIS and spatial data catalogs.)

ISSUE 4: Although a Federal Information Processing Standard (FIPS) for spatial data transfer has been approved, profiles for implementing this standard for the exchange of spatial data between federal agencies have yet to be developed. Moreover, standard activities need to be expanded beyond transfer standards to include more specific measures and standards of content, quality, currency, and performance of various components of the NSDI. As a corollary, there is no agreed-upon representation of ''base data" for small-, medium-, and large-scale spatial data products.

There is general agreement that spatial data exchange standards are a federal responsibility. However, there is a plethora of standards already identified for the exchange of spatial data. Most of these concern special data exchanges between a data producer and its user community or between GIS software systems. Other exchange standards are in the form of specifically defined products such as TIGER/Line, Digital Chart of the World (DCW), Digital Terrain Elevation Data (DTED), Digital Feature Analysis Data (DFAD), or Digital Line Graphs (DLGs). These files of specific spatial data have a particular format that the sending and receiving organizations have agreed to use for data exchange. These product formats, however, are not robust enough to support the ad hoc exchange of digital data.

The Spatial Data Transfer Standard (SDTS; FIPS-173) is an attempt by federal agencies, under the sponsorship of the FGDC, to develop a general family of exchange standards for civilian geographic data. The SDTS should serve as the umbrella for more specific "federal profile" standards, such as DIGEST, VPF, and TIGER.

With regard to base data, they should include, at a minimum, those spatial and primary attribute data that can identify all relevant information for a particular scale product. The base data must meet relevant positional accuracy and unique identifier requirements for inclusion of value-added

data from other organizations. Table 4.2 includes a typical set of base data that are representative for three scale ranges.

Instead of a single base of data, the agencies are using multiple bases that are often incompatible. Some agencies are using the Bureau of the Census' TIGER/Line files as their base, whereas others are using the USGS 1:24,000 quadrangles. Although TIGER contains street names, the DLGs do not. Therefore, it is difficult to transfer TIGER attributes to the more geographically accurate DLGs. Other agencies such as the EPA and the SCS's NRI are using unique sampling units on separate base data at varying scales. Some data collection efforts such as the proposed digital orthophoto program (SCS, ASCS, and USGS) and the Digital Chart of the World [produced by the Defense Mapping Agency (DMA)] collect data to still different base scales and accuracies. These two products could become the base data standards of reference in the future for large-and small-scale products that would complement the USGS 1:24,000 (medium scale) DLG products, which have become the de facto medium-scale standard.

ISSUE 5: There are major impediments to, and few workable incentives for, the sharing of spatial data among the federal, state, and local organizations.

All federal agencies recognize the large expense involved in collecting spatial data in digital form. However, there are no real incentives (in fact there are oftentimes disincentives) to the sharing of these data. To share data, the federal agencies would have to agree in advance on the uses, types, and formats of the collected data. This would require one agency to collect, at its expense, more data than it required to satisfy another agency's needs. At the same time, the agency would become dependent on other agencies to collect data with the needed rigor, accuracy, and urgency. If such agreements could be arranged, one or both agencies might lose personnel and funding as a result of the savings in spatial data collection.

Collection of spatial data for sale to the general public has its own set of disincentives. If state, local, or private organizations are willing to pay for the data, the payments often go to the general U.S. Treasury and are not returned to the agency that dedicated resources to collect and distribute the data to the buyer. The USGS is one of the few agencies that are allowed to collect proceeds from map and data sales and directly reimburse the relevant programs. Without a way to be directly reimbursed for the

resources spent, federal agencies are reluctant to collect spatial data for others.

Even when joint memoranda of agreements are signed between federal agencies, it is usually for equal exchanges of effort. Data collected by one agency in one part of the United States are given to a second in exchange for equivalent data collected elsewhere. Examples of this type of joint data collection include the proposed USGS/ASCS/SCS orthophoto program and the use of USGS 1:24,000 quadrangles by many agencies as the base for their products. Other examples, such as the seemingly redundant resource mapping done by the SCS, the National Agricultural Statistical Service (NASS), the USFS, and the EPA's proposed Environmental Monitoring and Assessment Program (EMAP), show a distinct proclivity for not sharing data, sometimes even within the same department. At a minimum, coordination of sampling activities should occur within and between federal agencies.

The major issues reflect problems that the federal agencies are experiencing with the use of spatial data. Although customer demands for spatial data have increased dramatically and GIS technology has kept pace, the federal "suppliers" of spatial data have had significant difficulty converting their hardcopy production process to flexible production systems capable of supplying spatial data in both digital and printed form.

A review of current federal agency activities (see Table 4.3) illustrates the magnitude of these problems. In general, each agency is trying to overcome the problems within the confines of its own charter, thus erecting additional impediments to establishing a robust NSDI. Most federal agencies are also concentrating on supplying spatial data in forms that match defined hardcopy products rather than in forms that lend themselves to easy manipulation by commercial GIS packages. In this regard, the federal agencies have not had incentives to meet their evolving mission as suppliers of spatial data to their customers.

The MSC paid special attention to federal agencies with major spatial data collection efforts: the Bureau of the Census, the BLM, the DMA, the USFS, the FWS, the SCS, and the USGS. The EPA is discussed because it is one of the largest federal users of spatial data. These agencies will be

discussed in the light of their responsibilities to meet OMB revised Circular A-16.

The OMB has executive oversight responsibility for the budgets of all federal agencies and for establishing policies and procedures for federal agency GIS activities. Part of this oversight includes the review and approval of all large expenditures for spatial data systems. The OMB issued Circular A-16 in 1953 (revised in 1967 and, most recently, in 1990) to facilitate coordination of federal mapping activities. In 1983, an OMB memorandum established the Federal Interagency Coordinating Committee on Digital Cartography (FICCDC) to assist in coordinating data sharing across federal agencies. The FICCDC found many obstacles that prevented data sharing between government agencies and the private sector; barriers were of a technical, institutional, and legal nature. The FICCDC cooperated in developing and publishing the first draft of the SDTS to help mitigate these problems.

One recent concern of the OMB has been the shifting of the nation's map base to electronic media. In 1988, the OMB recognized the need to look for opportunities to share data, standards, and development efforts among the federal agencies. Such sharing should substantially reduce overall costs to the U.S. government.

However, the expected costs continue to rise for developing, implementing, and operating systems to convert existing map graphics from paper or other stable base to electronic form. The results of the OMB Bulletin 88-11 survey in 1988 showed that the estimated costs for "electronic mapping" of the civilian side of the federal government exceeded $100 million for FY1988 and would top $200 million by FY1992. (Briefings by the federal agencies to the MSC indicate that the estimates for current federal mapping initiatives are much greater—by at least a factor of two—than indicated by the 1988 OMB Bulletin.)

The civilian agencies with the largest expenditures in spatial data activities are the EPA, the Bureau of the Census, the BLM, the SCS, the USFS, and the USGS. The EPA indicated that their FY1992 investment in spatial data was around $500 million; the SCS reported that their expenditures were around $220 million. (Both the EPA and the SCS alone exceed the 1988 OMB estimate; however, caution needs to be applied to what is included in these estimates.) In other agencies the MSC found that

modernization efforts accounted for 75 percent of their estimated mapping-related expenditures.

OMB is especially interested in cost savings and return on investment. In particular, new development efforts must show short-term, quantifiable payoffs from investments. Most important are Class I savings, which show direct payoffs for the investment. Cost avoidance approaches (Class II) are also good, but not as meaningful as the direct savings of Class I. Class III savings, the improvement of goods and services for the public sector, are also saleable. [An example of a Class III savings would be the new Federal Aviation Administration (FAA) system that would make flying safer.]

Applying cost savings approaches for spatial data initiatives includes (1) use of standards for storage and transmission of spatial data (such as the SDTS), (2) maximizing the use of existing data bases and encoding schemes, (3) developing excellent benefit-cost ratios for any new GIS or spatial data initiative, and (4) sharing of spatial data collection and new technology developments among federal agencies.

OMB Circular A-16 was revised in 1990 to encourage data sharing among federal agencies. However, no real incentives are in place to ensure that such data sharing occurs.

The FGDC was established in October 1990 by revised OMB Circular A-16. The FGDC is responsible for promoting and coordinating the development, use, sharing, and dissemination of GIS data throughout the government. The revised circular established the FGDC as replacing the FICCDC. It also empowered the FGDC to promote the coordinated use and development of mapping, surveying, and other spatial data.

OMB Circular A-16 specifically states that the FGDC:

"• supports surveying and mapping activities, aids geographic information use, and assists land managers, technical support organizations, and other users in meeting their program objectives through;

"• promoting the development, maintenance, and management of distributed data systems that are national in scope for surveying, mapping, and related spatial data;

"• encouraging the development and implementation of standards, exchange formats, specifications, procedures, and guidelines;

"• promoting technology development, transfer, and exchange;

"• promoting interaction with other existing federal coordinating activities that have interest in the generation, collection, use, and transfer of spatial data;

"• publishing periodic technical and management articles and reports;

"• performing special studies and providing special reports and briefings to OMB on major initiatives to facilitate understanding of the relationship of spatial data technologies with agency programs; and

"• ensuring that activities related to Circular A-16 support national security, national defense, and emergency preparedness programs."

The revised OMB Circular A-16 assigned responsibilities to numerous agencies for various FGDC roles. The USGS was identified as the lead agency for the FGDC. Ten subcommittees were established, each responsible for coordinating activities for a category of spatial data. There are also three working groups (standards, technology, and state and local liaison) that deal with issues common to all spatial data categories. Table 4.4 shows the federal agencies with lead coordination responsibility for the major data categories.

The FGDC (1991) in its report to the OMB refers to a National Geographic Data System (NGDS) as a system of independently held and maintained federal digital geographic data bases. The NGDS would encompass the National Digital Cartographic Data Base (NDCDB) of the USGS plus other spatial data bases that are national in scope. The NGDS would include traditional cartographic data as well as thematic data, such as soils, wetlands, geology, vegetation, and demographic data. The NGDS would be an important component of the NSDI.

To ensure proper populating and use of the NGDS, standards must also be developed for each spatial data layer. Some standards (coordinate systems, code sets, and geographic names) already exist. Standards for many other segments or layers of the NGDS must be developed. Lineage, positional accuracy, attribute accuracy, logical consistency, and completeness (all parts of data quality) are needed to ensure data integrity.

A number of federal agencies have extensive, ongoing, spatial data collection programs, as shown in Tables 4.1, 4.3, and 4.4. At least eight of these—the Bureau of the Census, the BLM, the DMA, the EPA, the FWS, the SCS, the USFS, and the USGS—have national data collections

efforts, which suggest the possibility of data sharing with others. Several of the major programs are briefly described below.

The Bureau of the Census has produced over 1 million different map sheets to support its collection and customer use of census data. However, maps and charts are a supporting activity, not their primary product. The

Bureau of the Census traditionally used three major tools for conducting a census: maps (to tell census takers where they are), address reference files, and geographic reference files. The address reference files are used to match census questionnaires to locations, and the geographic reference files serve as a control inventory of over 7 million census statistical zones.

In 1982 the Geography Division of the Bureau of the Census began developing the Topologically Integrated Geographic Encoding and Referencing (TIGER) System to serve as a single data base incorporating the above three tools for the 1990 decennial census. The TIGER files implemented a new data structure to capture every known street and road in the United States, the name (or names) of each, ranges of address numbers, all the railroads and significant hydrographic features, named landmarks, and other named geographic locations. This information was needed to administer and tabulate the 1990 decennial census of population and housing of the United States and will be used in all future censuses.

The TIGER data base is composed of data from three sources. Scanned 1:100,000 maps from the USGS form the base of 98% of the country. GBF/DIME (geographic base file/dual independent map encoding) files from the 1980 census were used for major metropolitan areas. Data digitized by commercial contractors were used to fill in between the 1:100,000 maps and the DIME files. These data were integrated horizontally and vertically and cut on county boundaries.

The current TIGER files contain the latitude and longitude coordinates of more than 42 million linear features (roads, railroads, water features, and landmarks), address ranges, and political and statistical boundaries. Omitted from the TIGER files are contours, public land survey information, rural route address ranges, and accuracy data. That means the TIGER files must be augmented for most other GIS applications.

TIGER files form one of the most important data sources for states, cities, counties, and utility companies. With proper processing, it may be possible to match these files to USGS terrain data and digitized 1:24,000 quad sheets (as they become available) and eventually to 1:12,000 digital orthophotography to provide very accurate data for all GIS systems. Additional aspects of TIGER appear in the Street Centerline section of Chapter 5.

Some private companies sell improved versions of TIGER; others are using TIGER files as the base for special studies. This usually has been done by converting TIGER files into various layers of data for input to a full-function GIS and then adding data attributes as needed. Some attempts

to use TIGER files, such as Bell South's 911 emergency address data base system, were not successful and other sources are being developed. The fact that this $300 million investment in TIGER could not meet these requirements is an indication of the possible need for better coordination.

The Bureau of the Census and the USGS have recently undertaken a project to convert TIGER files to the SDTS format. The initial results demonstrate that TIGER content can be expressed in the SDTS format, but it remains to be seen whether the SDTS format is preferable to current TIGER file format for data exchange.

The major problem with TIGER files is keeping the data up to date. The Bureau of the Census' charter includes an update every 10 years but not a continual update cycle. The broad use of TIGER in both the public and private sectors underscores the importance of an up-to-date TIGER system. The Bureau of the Census is investigating the establishment of various maintenance agreements with the states so that TIGER data can be updated more frequently.

The BLM administers 272 million acres of public lands primarily in the western United States and Alaska. In addition to the surface the BLM is responsible for the subsurface for an additional 300 million acres of public lands in which the federal government has retained the minerals interest (includes lands under Forest Service jurisdiction, wildlife refuges, and military reservations). The basic mission of the BLM includes three primary responsibilities: (1) the public land survey system (PLSS); (2) maintenance of land and minerals records, including title transfer from federal to state and private ownership and all the land records of land that the BLM currently administers; and (3) natural resource management for multiple use. This mission has led the BLM to invest in a variety of land and geographic information systems. The BLM has been designated the lead agency for the cadastral subcommittee of the FGDC.

The BLM is a decentralized organization, with headquarters and 12 state (often regional) offices. Within the state offices, there are district offices and resource area offices. Most of the daily activities and processing take place at the resource area offices. As a result of the historical land activities, much of the land is in a checkerboard pattern, with BLM land interspersed with state and private lands; this has resulted in the need for

close working relationships with state and local jurisdictions, particularly in sharing digital data.

There are specific types of information, for example, the PLSS (the geographic coordinate data base) and the Automated Land and Minerals Record System (ALMRS), that are critical to the BLM that also have other users. Data layers important to the BLM include transportation, meteorology, demographics, geology, soils, and a variety of layers that are specific to given land management agencies depending on the particular local needs.

The primary mission of the DMA is to provide mapping, charting, and geodetic (MC&G) products and services in support of worldwide military operations. This responsibility also includes military training activities and the production of military specific products to support weapon and system command and control requirements that cannot be fulfilled by civilian MC&G programs. The DMA activities support worldwide statutory responsibilities such as the production of flight information product (FLIP) charts, aids to navigation and pilotage charts, and related public safety products. A significant percent of their product coverage requirements are met with bilateral agreements through coproduction with other U.S. federal agencies and foreign allies.

The DMA supports the use of the SDTS as a national digital data exchange mechanism, but views the SDTS as an umbrella standard rather than one to be used for exchanging data. Therefore, the DMA is working with the FGDC to establish the DMA's Vector Product Format (VPF) as a federal profile for the SDTS. The DMA is also supporting the Digital Geographic Information Exchange Standard (DIGEST), an international standard that will be used for data exchange among NATO nations. DIGEST accommodates object-oriented data formats (in Appendix A of DIGEST) and relational data formats (VPF in Appendix C of DIGEST). Therefore, a logical step in the standards process now under investigation is to make profiles of the SDTS for both DIGEST formats.

The DMA recently completed a two-phase study that addressed its future specifications and standards (DMA, 1991). The DMA concluded from the results of this study that the trends in future Department of Defense systems were for increased capability for merging (or ''fusing") multiple spatial data sets. This trend is the result of supporting users who

use diverse computer systems and need different digital products at various scales. The manipulation of spatial data in GIS to meet these needs will be the norm, not the exception.

The DMA completed the installation of its Digital Production System (DPS) in early 1993. This very large spatial data extraction, processing, and production system attempts to minimize the overlap of data collection within the organization. All spatial data needed to satisfy DMA's mission will be stored within a single set of DPS data bases. These data will then be used for producing products at various scales and levels of detail to meet customer needs.

The major criticism of the DPS is its limited flexibility for fast response to requests for new types of data. Because it was built as a production system, its main focus is to produce standard products that meet defined accuracy, content, and temporal requirements. It does not easily handle requests for ad hoc data and tailored products that are anticipated in the future (this shortcoming is expected to be resolved in the next few years as the DMA ramps-up its production on the DPS).

The DMA maintains spatial data catalogs as hard copy showing the various spatial data holdings, their accuracies and types of data, and their various scales. These catalogs are planned to be put on-line within the next few years.

Although the DMA has expressed interest in supporting the NSDI, their support will be tempered by their primary mission. The needs of the military departments and other federal agencies for spatial data over foreign countries will normally take priority. If the spatial data exchange standards can be accommodated without undue expense, the DMA will likely be a proponent.

The EPA has been involved in a very large information integration initiative since 1999. The EPA is currently investing more than $500 million per year in assembling spatial data (usually collected by others) and disseminating the data. Because of this, the EPA has become the largest government consumer of spatial data in support of GIS applications. Although a member of the FGDC, the EPA is not responsible for the production of any spatial data layer of the NGDS, but it does have lead responsibility with the FGDC activity in state and local liaison.

Protecting the environment requires suitable geographic information. It involves the understanding of the spatial relationships of human population centers, natural resources, and sources of pollution. Early in the information initiative, the EPA Office of Information Resources Management recognized the powerful tools that GIS provides to help analyze these three classes of data.

In 1989, the EPA established the Geographic Resources Information and Data System (GRIDS) as the agency's central repository for national spatial data bases. GRIDS provides a query capability that ties together the spatial data bases under the EPA's control. These include the Bureau of the Census TIGER files, the USGS 1:100,000 DLG hydrography and transportation data, the DMA 1:250,000 Digital Elevation Model (DEM) data, the SCS data files, 1:2,500,000 hydrographic basin files, the FWS's National Wetlands Inventory, 1:2,000,000 political boundary files, 1:1,000,000 ecoregions files, the USGS Geographic Names Information System (GNIS), and the Reach files for all U.S. rivers. Requests for information not in these files are sent to the FGDC.

Integration of these data is needed for the EPA to recognize floodplain hazards, wetlands pollution, surface, and ground-water quality problems, potential unstable foundations, population proximity to environmental buffer zones, and many other types of analyses. The integration of the separate data bases has required the development of a number of standards. These standards are expected to overcome the two major EPA barriers to data integration: inconsistent interfaces and no common link between data bases.

A major objective of the EPA approach has been the potential capability of overlaying various data sets. This capability, however, has been slow to materialize. There currently are no standard "tie points" that allow data of different scales and accuracies to be automatically rescaled,

thinned or densified, and overlaid on each other with relative consistency. This problem must be solved for the EPA data to be used to their potential.

Even with this problem, the EPA is assembling the most extensive multiagency collection of nationwide spatial data. In doing so, it has developed coordinating mechanisms with FGDC, the Bureau fo the Census, the USGS, the Army Corps of Engineers (COE), the National Geodetic Survey, and the National Institute of Standards and Technology (NIST). The EPA also has a data sharing program with all states and Puerto Rico and a working relationship with the National Center for Geographic Information and Analysis (NCGIA), sponsored by the National Science Foundation (NSF).

The EPA is developing an ambitious program, EMAP, "to assess the nationwide distribution of ecological resources and to assess trends in their conditions. A unique aspect of the program is its reliance on probability-based selection of sampling locations for both of those major goals" (BEST/WSTB, 1992). The MSC found that many aspects of the SCS's NRI, EMAP, and other national statistical surveys appear to overlap and be duplicative; however, because these surveys have each selected different probabilistic sampling methodologies, sharing of data is often inhibited.

The goals of the state/EPA Data Management Program are twofold: to build and maintain the infrastructure for data sharing with the state environmental agencies (basically complete), and to integrate this data across multiple media and programs. Because of the aformentioned problems, satisfaction of this second goal may take some time.

The principal mapping activity of the FWS is involved with wetlands mapping. The FWS is also responsible for coordinating the wetlands data layer of the NGDS. (For additional information pertaining to the FWS see Chapter 6 and Appendix A dealing with wetlands and spatial data.)

The flagship program in spatial data of the FWS is the National Wetlands Inventory (NWI). The NWI was mandated by Congress and established in 1974 to standardize terminology and to inventory the nation's wetland habitats by 1998 (conterminous United States). The NWI has currently produced 20,000 wetlands maps covering 60 percent of wetlands areas of the lower 48 states and 16 percent of the Alaskan wetlands areas. The primary map product is the 1:24,000 scale map that shows the location, shape, and characteristics of wetlands and deepwater

habitats. Approximately 3,200 of these maps are produced each year. At this rate the FWS expects to complete the NWI for the contiguous 49 states by 1998. The FWS maps use USGS 7 1/2-or 15-minute topographic maps as their base. Other wetlands maps include 1:100,000 and 1:250,000 small-scale maps, and 1:62,500 or 1:63,360 large-scale maps.

The FWS considers their wetlands data collection to be cost effective and does not want to duplicate data captured by others. However, this is not an easy task. Various federal agencies are involved in wetlands data collection because of their particular mandates and missions.

Wetlands data are supplied to about 50 user organizations by the FWS. Among these, Ducks Unlimited had been paying the FWS to digitize wetlands areas of special interests. The FWS was recently told by the Department of the Interior (DOI) Inspector General (DOI, 1992) to stop this activity and to concentrate on completing the NWI, because the digitization was not included in the original congressional authorization of the NWI. This issue was resolved in 1992 as the NWI authorization was revised (P.L.102-440, §305); the FWS can provide digital products of the NWI providing the requestor pays for 100% of the associated digitizing costs. Because the NWI uses contractors, the flexibility exists to add additional digitizing tasks without affecting the production schedule for the typical NWI products. This limits automation to those that have access to the needed funds for digitizing.

Although the FWS has been digitizing and exchanging digital data for many years, they appear to have taken a more practical approach to exchange standards. Their exchange formats are directly related to the systems they have in house and to the formats of the data they receive from other agencies. Source data sets are acquired from the USGS in the form of 7 1/2-minute quads in DLG-3 optional format, orthophoto quads, high-altitude photography, and Landsat and SPOT imagery. The FWS wetlands inventory data are provided on magnetic tape in multiple formats. (Additional discussion of data and information issues concerning wetlands are in Chapter 6 and Appendix A.)

The USFS, like the BLM, is basically a land management agency responsible for 191 million acres of public lands. The mission of the USFS (much like the BLM's) emphasizes multiple use and sustained yield. The difference between the roles of the USFS and the National Park Service is

that the USFS is multiple land use whereas the National Park Service is usually single use. Also like the BLM, the USFS is geographically dispersed, with 156 national forests and 653 ranger districts. The ranger districts are probably the most important sites that use spatial data. Each district office administers from 150,000 to 500,000 acres and is responsible for virtually everything that occurs in the district.

The USFS obtains its spatial data from in-house activities; other federal, state, and local agencies using cooperative agreements and cost-sharing arrangements; and private companies as a condition of permit or timber sale on USFS lands.

The USFS produces digital spatial data from their base series maps and orthophotography. These base maps use the USGS 1:24,000 base, modified to include land ownership status, administrative boundaries, and transportation networks (at a different definition of the USGS base). About 10,500 quadrangles are required to cover the national forests. The USFS also produces a cartographic feature file in digital form. This uses a simple format and includes coordinates of map corners and point and line features; boundaries are represented as lines (not polygons); the files are not topologically structured but include feature codes; and the nodes are digitized. Although the cartographic feature files are less than optimal for GIS application, it is what the USFS could afford to do; these files can be converted into DLG format.

The USFS has been designated as lead agency for the vegetation subcommittee of the FGDC.

The SCS within the U.S. Department of Agriculture (USDA) has as its mission to help the states, cities, interest groups, and private citizens to better manage land and water resources. Their charter includes soil surveys, snow surveys, watershed studies, flood control, and conservation technical assistance to landowners.

The SCS long ago realized the benefits of using a GIS for tracking and manipulating spatial data. The SCS uses the public-domain, UNIX-based GRASS system as their primary GIS, and the SCS has tailored interfaces for their personnel.

Currently, the SCS spends about $220 million annually on mapping of soils and other resource information over the United States. This breaks down to $130 million for conservation technical assistance mapping, which

includes wetlands mapping (at 1:7,920 scale); $10 million for watershed and subwatershed resource mapping, and $6 million for river basin mapping at 1:100,000 and 1:5,000,000 scales. The remaining $66 million is spent on soil mapping at scales of 1:12,000 to 1:31,680. Soil surveys are a major SCS product, published as printed reports and paper maps but now also becoming available as digital data bases.

The SCS in conjunction with the ASCS and the USGS is proposing a 1:12,000 digital orthophotography program. If implemented, this would considerably increase the store of digital data nationally and at a scale of considerable use by local investigators.

The National Resources Inventory (NRI) is another major data product of the SCS. The NRI contains soil characteristics and interpretations (slope, depth, permeability, salinity, and acidity), land cover, land use, erosion, land treatment, conservation treatment needs, and vegetative conditions. The NRI is used to monitor trends in the status and condition of natural resources and to help formulate state and national policy. Data are collected for the NRI by primary sampling units (PSUs). There are approximately 300,000 PSUs over the United States and data are collected by field units in over 1,000,000 points (three per PSU). The point data are then summarized for statistical purposes and trend analysis.

The USGS has the responsibility for the National Mapping Program for the United States. This program includes the production of 1:24,000 scale (7 1/2-minutes) quadrangle maps for the conterminous United States, Hawaii, and Puerto Rico and 1:62,500 (15-minute) topographic quadrangle map coverage of Alaska. In addition to this primary map series, the USGS also produces intermediate and small scale map products at 1:50,000, 1:100,000 and 1:250,000 scale in both quadrangle and county formats.

Many of the spatial activities of the USGS have previously been reviewed by the MSC. The report Spatial Data Needs: The Future of the National Mapping Program (1990) reviewed user needs for the USGS spatial data products, and the report Research and Development in the National Mapping Division, USGS: Trends and Prospects (1991) evaluated the USGS's R&D program. We reaffirm the conclusions and recommendations in these two reports and are encouraged by USGS efforts to be responsive in implementing these recommendations.

The USGS has assigned the National Mapping Division (NMD) with responsibility for conducting the National Mapping Program. In 1990, they completed the first version of the 1:24,000-scale map coverage for the conterminous 48 states and Hawaii, and expect completion of the initial mapping of Alaska by late 1992. Now that the initial hardcopy coverage of the United States is basically complete, the NMD is focusing on ways to maintain the currency, accuracy, and usefulness of their maps.

Many of the federal and state agencies rely on the 1:24,000 quadrangle map series to form the base data upon which other information is referenced. The use of the 1:24,000 quads by other federal and state agencies is one method for reducing duplication of data collection activities. The NMD produces digital spatial data in various formats in response to requests from a wide variety of customer.

In 1987, the NMD undertook a new system development effort called Mark II. This program is to implement advanced technologies and production procedures to satisfy their requirements through the year 2000. The major aspect of this program will be the population of the National Digital Cartographic Data Base (NDCDB) with spatial data representing the content of the 1:24,000 primary map series and other smaller scale series.

In 1991, the NMD changed the focus of the Mark II system development effort to be more of an incremental "commercial off-the-shelf" (COTS) procurement. This re-named Automated Cartographic System (ACS) marries advanced technology developed for other federal cartographic systems with the latest COTS hardware and software. The system is to encompass an open-ended architecture so additional COTS developments may be added in the future. The ACS will be used to support products for the proposed the SCS/ASCS/USGS digital orthophotography program.

The USGS is attempting to accelerate the data collection effort for the creation of the NDCDB. This is being done through data exchange with federal and state agencies, and by contracting out some of the digitizing to private companies. The USGS is working with the USFS for acquiring large amounts of 1:24,000 DLG data. Hypsographic data at 1:100,000 scale in digital form is also being produced by the DMA in a cost-share agreement. Finally, both the USFS and the BLM are developing DEM data to USGS standards under a work-share agreement.

The USGS is also using GIS technology and spatial data analysis for water-resource studies. There is currently a nationwide computer network

of over 200 locations staffed with GIS users working on a variety of water-related studies. The system has been used to work with groundwater models, define drainage basins and river systems, assess runoff in urban areas, and examine trends in water quality.

The National Water Quality Assessment (NAWQA) program also has a major influence on USGS spatial data activities. Sixty NAWQA study units will complement the normal 423 long-term water sampling stations to collect data for water-quality studies. Spatial data will be used to classify lands and detect change for tracing pesticides and trace metal concentrations in the environment, and for hydrologic simulation models.

Spatial data activities typically occur in several agencies in each state. The common agencies include the functions of natural resources, environmental protection, agriculture, transportation, taxation/assessment, emergency management, and planning and community development.

Some state-level spatial data are collected and used at the scale of 1:24,000, and some agencies or programs require larger or smaller scale data. Spatial data are used for environmental applications, though each state's transportation agency has specific requirements related to the road and other transport systems. Large scale maps are also used in the assessment of land values and taxation in a function that is frequently but not exclusively delegated to local governments. There are also various other applications that make use of demographic and economic data.

GIS technology is being implemented in many agencies in virtually every state today (see Warnecke, 1992). The acquisition of spatial data is a very significant percentage of the effort (and resources) for each GIS project. Most state agencies have similar requirements for spatial data and therefore there is significant potential for sharing of acquisition efforts and data. The USGS 1:100,000 and 1:24,000 maps are the most common source for base data; the SCS soil maps (1:15,840) are also frequently used. Conversion of these maps to digital form is therefore a focus for most of the state spatial data development efforts. (Since the USGS also

has a program underway to digitize their 1:24,000 maps, there is a strong potential for data sharing with the states in this activity, see Chapter 8.)

Several states have passed or are developing legislation that affects spatial data (see Table 4.5). In some cases the legislation creates or designates a coordinating body. In others it provides access to funding for local governments to modernize and implement land records (see North Carolina example in Chapter 7). Also it creates requirements for spatial data to support a program or regulation. In still others it affects the availability of spatial data developed by state agencies.

Growth management, mining reclamation, redistricting, tax mapping, transportation planning, traffic safety, hazardous materials, natural resources, economic development, and land records modernization are key areas of spatial data activity at the state and local levels.

In the past, most state spatial data activities have been for special programs, often with federal funding and/or direction. However, this situation is beginning to change. States are emerging as an important key player in the collection, dissemination, and coordination of geographic and land information.

One of the major stumbling blocks to the collection of spatial data by state agencies has been funding. To make matters worse, the federal government has delegated more responsibilities to the states (and likewise, states to local government) without matching funds in most cases. In spite of these problems, more and more activity is occurring at state levels to collect and process spatial data using GIS tools.

While much of the state level spatial data acquisition in the past was performed on an individual agency basis, most states now have some level of recognition of the need for sharing and cooperative efforts. Thirty-five or more states have established some type of coordinating mechanism or lead agency for GIS or spatial data. In some states that recognition is formalized by assignment of a lead agency with statewide responsibility or by the formation of a GIS coordinating committee. Some states have implemented legislation and some have prepared or are preparing statewide plans for spatial data development, maintenance, and use. In other states informal arrangements have been established between and among pairs or groups of agencies. In some states, however, cooperation suffers from interagency rivalries, incompatibility between specifications or between systems, or lack of resources to establish coordination mechanisms.

There is a growing awareness of the mutual benefit to be gained by communication among states on this topic. Several meetings and surveys

have been conducted in recent years in an attempt to improve the communication among states. At the invitation of Georgia Governor Zell Miller, representatives of 39 states met during the 1991 GIS/LIS meeting in Atlanta. The meeting resulted in the formation of the National States Geographic Information Council to address coordination and common issues within the states. One of the most repeated themes at this meeting is that the federal agencies need to work together more effectively to form stronger spatial data partnerships with the states.

Some state programs involve direct cooperation or coordination with related federal agencies and programs. Some others are operated in response to federal requirements for transportation, environmental protection, or other functions.

Where state programs are funded by or are performed in conjunction with a federal program, some level of state-federal cooperation exists. The level of cooperation varies from establishment of standards and sharing arrangements to cooperative efforts for digitizing the USGS quad sheets. Federal reporting requirements affect some state spatial data activities. Both the DOT and the EPA have had an important effect in this area. Many forms of environmental control and regulation emanating from the federal government require the development and use of spatial data, sometimes to federal specifications and others merely to meet federal reporting requirements. In some states specific programs such as the selection of nuclear waste disposal sites (e.g., Pennsylvania) and determination of oil revenues (e.g., Alaska) have been guided by federal programs or regulations. Furthermore, the Bureau of Census' TIGER data forms the legal basis for the most recent redistricting process.

The National Geodetic Survey (NGS) operates a state advisory program in 26 states. The NGS state advisor provides technical assistance and stimulus for improved spatial data activities.

Western states have developed special relationships with federal land management agencies that do have limited activities in the east. In some of these western states there is sharing of data between the state agencies and the BLM and the USFS, though in other states these entities operate independent of each other. In any case, all these relationships appear to be ad hoe and often narrowly defined.

There has been a great deal of interest and activity in the development of data exchange standards. The current practical reality, however, is that only de facto standards based on available data or formats of software-specific systems are recognized in state and local governments. As a default choice, the USGS DLG (1:100,000) and the Bureau of the Census TIGER formats are often used because data are available at low cost and contain complete coverage of each state.

Probably the most significant spatial data problem facing states today is a lack of resources and a shortage of funding for data acquisition and management. States often look to the federal government for assistance, particularly with regard to spatial data required for federal mandates. In most cases, however, the federal government is not able to provide the desired assistance.

Coordination of spatial data handling among state agencies has been a chronic problem. Differing requirements, limited resources, and organizational rivalries have impeded cooperation in spatial data activities. This problem is being addressed in many states today and the situation is improving significantly. Even though many states now have a coordinating body and data sharing among agencies is improving, they are still depend on a continuous flow of accurate and timely spatial data.

Local governments are major creators, maintainers, and users of spatial data. Analyses have indicated that as many as 90 percent of local government (approximately 80,000 municipalities, towns, townships, regional planning entities, and school and special districts) operations involve some use of spatial data. Many have stated that they spend more on GIS activities than the federal and state governments combined. Local governments use spatial data for a wide range of activities that include: real estate assessments, land-use planning, public works, water and sewer utilities, resource management, environmental control, solid waste management, emergency management, and health care management.

Local governments generally require medium-or large-scale maps and data. While some local government activities, particularly in rural areas, can be supported by the 1:24,000 scale maps and TIGER data, most local government activities in urban areas require data at scales of 1:4,800 to 1:1,200 (1'' = 100'). Local governments have extensive requirements for cadastral maps since the ownership parcel is a key entity in many local government functions. Planimetric features such as roads, hydrography, and buildings are also important features; soils and forest cover are important resource features. Many local governments operate public works and utilities departments that require data on the location of public facilities and utilities. All of these features must be mapped at medium or large scales depending on the level of urbanization and the parcel sizes.

There has been a major movement to develop and implement GIS among local governments in the past several years. The major effort of each GIS project has been the development of large-scale and medium-scale spatial data bases containing cultural, infrastructure, and resource features. These projects usually involve the creation of new base maps from high-resolution aerial photography and geodetic control. As a result, these projects can take five or more years and cost several million dollars. Because of the resources required there are examples (e.g., Indianapolis) of data sharing among the local government and utility organizations within a geographic area. However, there has been limited cooperative activities between local governments and state or federal agencies.

Some states have established programs of assistance and/or regulations specifying mapping standards for local governments. Most of these focus on equitable tax mapping throughout the state. Notable among these are South Carolina, Missouri, Oklahoma, and New York. In these cases, mapping standards employed by local governments tend to be uniform and of high quality throughout the individual state.

Recently some states, such as Florida, Georgia, and Vermont, have established growth management policies and/or regulations that have placed requirements on local governments for the acquisition and use of spatial data. These requirements are sometimes accompanied by resources for compliance. Wisconsin has implemented the Wisconsin Land Information Program that provides access to state funds when local governments implement a land records modernization plan.

Spatial data development in a local government is now typically incorporated in a GIS program. The GIS program in turn is typically a consortium of several departments across the city or county organization or of multiple city, county, and utility organizations. Because this can cross a multitude of jurisdictional boundaries, these local projects require considerable organizational cooperation. The lead agency is usually a GIS user department such as planning or public works. GIS data activities have traditionally taken place outside of the conventional information services department.

There is little sharing of data between local governments and federal agencies. In recent years there has been very little federal funding provided to local governments and so little sharing has taken place.

The FEMA flood insurance program has generated maps of flood zones that are used extensively by local governments. These maps are required for definition of locations for flood insurance purposes. Even though the scales and formats of these maps are often incompatible with other local government spatial data, they are used out of regulatory necessity.

Federal regulations significantly affect local governments. EPA regulations in particular are generating requirements for digital data among local governments. For example, the recent EPA regulations (e.g., the non-point discharge elimination standards) are stimulating many local governments to collect spatially referenced data for purposes of storm water pollution permitting. This requires the local governments to integrate geographic, water, and topographic features with other local spatial data.

Generally the same standard issues facing states (above), face local governments. In fact the issues may be exacerbated by the often dramatically different scales used at the local level.

Local governments suffer from the same lack of funding and resource problems as states. The magnitude of the effort required to acquire spatial

data for a local government is often a serious problem. While new technological tools and procedures are being developed, the development of a spatial data base continues to be a major concern.

There is about the same level of communication or coordination of spatial data activities among local governments as there is with state and federal agencies. Within a specific area, such as a county, there may be a cooperative GIS project that involves multiple organizations, but there are very few effective mechanisms for communication with other local governments to obtain guidance or resource sharing. A few states (e.g., Wisconsin) have a program for guidance or technical assistance from the state to local governments.

The private sector makes broad use of spatial data technology and has done so for over two decades; they are both producers and consumers of digital spatial data. Many uses involve digitizing proprietary single-purpose maps—for example, an electric utility company computerizing its "outside plant" facilities: utility poles, rights-of-way, and transformer substations. Timber companies were early adopters of GIS technology to inventory and analyze timber stands. While these applications involve spatial data, in many cases they result in the acquisition of proprietary and/or special purpose spatial data that are not of potentially broad usefulness within the national spatial data infrastructure. One exception is the sale of some of the spatial data by telephone companies to local governments as part of the enhanced 911 emergency service.

On the other hand, many private sector applications call on the same generic spatial data sets time and time again, specifically: boundaries or centroids of statistical areas (counties, census tracts, or zipcodes) and street centerline spatial data sets.

Typically, a person (or small group) at a private company becomes aware of a GIS or desktop mapping package, sees that spatial data technology could he useful to the company, and buys a copy of the product. They rapidly discover that the new system cannot deliver any value without a copy of an appropriate spatial data set. Practically without exception such a person assumes that the spatial data set is available from the government at low cost. What they discover is that either:

• the appropriate spatial data set (census tract boundaries, for example) is available only from the private sector;

• a government data set like the Bureau of the Census' TIGER file is sufficient for many of the applications and it can be acquired from a private company cheaper and/or in a more convenient ready-to-use format;

• a proprietary enhanced version of a government spatial data set is necessary to make the application work, because of some shortcoming of the government product; or

• the required spatial data set does not exist and the company must have it built from scratch at a substantial cost.

Incidentally, this experience is not restricted to private companies; in many cases government agencies at all levels can only find the spatial data sets they need from private suppliers.

The private sector has many advantages over government in use of spatial data technology for several of the following reasons:

• Private applications are generally simpler or at least more focused: the impact of spatial data technology on their bottom line.

• Private sector procurement practices are simple and streamlined, so companies have better access to the latest technologies at lowest cost.

• Private sector decision making is streamlined; action can be taken quickly.

In addition, private sector spatial data vendors in the United States have the tremendous advantage—practically unique in the world—of unrestricted royalty-free usage of federal spatial data sets like TIGER and DLGs, as well as copyright-free access to federal paper maps as a basis for proprietary spatial data sets.

As a result of these advantages, a robust private sector community of users and vendors of spatial data technology has evolved, which consists of a large group of users, software vendors and data vendors. Users represent nearly every sector of business: fast food chains selecting new sites, catalog marketers targeting customers, taxi companies dispatching cabs, insurance companies evaluating hurricane risks, food wholesalers routing delivery trucks, sales managers delineating sales territories, phone companies planning communication infrastructure for future developments, and banks complying with home mortgage disclosure regulations. Every

one of these application sectors is a consumer of the spatial data infrastructure and is expected to grow for years to come.

There is a growing number of software vendors offering a range of products from GIS, desktop mapping, desktop marketing, routing, and dispatching. In every case these software systems are useless without access to a spatial data set. Increasingly, the customers of software vendors demand "plug and play" spatial data sets; they want to concentrate on getting their jobs done without the costly diversion of digitizing or massaging a spatial data set. Many vendors only break even on sales of their software packages but make profits reselling spatial data sets to their customers.

Spatial data set vendors are either companies that simply transform data sets like TIGER or DLG to the internal representation required by systems like AutoCAD, or companies with more ambitious goals of creating proprietary spatial data sets from scratch or by significant investment to improve a public data set like TIGER. The latter companies have specialized to serve geodemographic markets or emerging IVHS (intelligent vehicle-highway system) applications like in-vehicle navigation and route guidance.

It is significant to observe that private spatial data set vendors dominate the supply of spatial statistical area boundary data bases as well as street centerline data sets. In aggregate, the combined data base creation and maintenance budgets of these companies could exceed federal spending on these data sets. In a previous report of the MSC (MSC, 1990), it was stated that de facto control of significant spatial data sets could default to the private sector. This has probably come true to some degree considering commercial activities of land title companies and the extent that private companies have leveraged off the running start provided by the Bureau of the Census release of TIGER in 1989.

Traditional private sector advantages of efficiency and competitiveness have served the government well when, for example, building of portions of TIGER was put out to bid in the mid-1980s. The private sector as a whole responded aggressively and positively to spatial data set technology with dynamic growth in all three sectors defined above. But private information businesses tend inexorably to monopoly or at best oligopoly. At present, a competitive market exists for some spatial data sets, and the rapid growth of "Business GIS" indicates that value is being delivered at current prices.

As we say several times in this report, what is missing from the federal perspective is a vision of spatial data use beyond fulfilling missions of individual agencies. The value of spatial data technology in promoting productivity or stimulating commerce is not a major concern of the FGDC or of most of its constituent agencies. Expansion of the vision of spatial data use beyond federal agencies' needs will be considered as a future subject for a Mapping Science Committee study.

Academia serves primarily in a support role for the improvement of the National Spatial Data Infrastructure. Although the nation's colleges and universities have extensive personpower, GIS experience, and specialized spatial data sets, there are few programs that coordinate their efforts for input to the NSDI. In almost all cases, academia must respond to the requirements of those agencies supporting their research. It is clearly not the role of academia to structure the NSDI, although the productivity of many of its scientists would he enhanced if the NSDI were more robust.

Many specialized GIS research and teaching laboratories have recently been established in U.S. universities. In many cases these laboratories work closely with state-level institutions to assist their states to rationalize and improve their spatial data and to educate students in the methods and techniques of processing and using spatial data. Examples of leading university activities include the National Center for Geographic Information and Analysis (NCGIA—a consortium of the University of California at Santa Barbara, the State University of New York at Buffalo, and the University of Maine at Orono), the Center for Mapping at the Ohio State University, the University of South Carolina, and the University of Wisconsin—Madison. Research programs at these and other universities are advancing basic knowledge in the field of geographic information and analysis. Examples include analysis of error in spatial data bases; the use and value of geographic information; development of new data gathering techniques such as Ohio State's GPS van (Bossler et al., 1991); remote sensing research and others. The NCGIA receives its principal support from the NSF along with additional support from the USGS, the EPA, and others; the Ohio State University's Center for mapping receives funding from the National Aeronautics and Space Administration (NASA), the USGS, and other state and federal agencies; and the University of Wiscon

sin—Madison is supported by the U.S. DOT, the NSF, the USDA, and various state, local, and private organizations.

Additional research is critical. The recommendations in the MSC's (1991) previous report on Research and Development in the National Mapping Division, USGS: Trends and Prospects outlines work that is applicable not only to the USGS but also every research organization that is part of the NSDI. We reiterate our emphasis for continued research on standards.

The explosive growth of spatial data handling technology and applications has whetted the appetite of users for new and improved capabilities to analyze, model, and apply the data to meet their needs. The private sector seems to have satisfied some of these needs through the development of new hardware platforms and software to process the data. Still other needs or desires require innovative R&D to extend data handling and modeling capabilities. Although some results are several years away from the marketplace, the private sector has demonstrated its rapidity to incorporate new spatial data handling advances into commercial products. To realize fully the potential benefits of new technology, researchers in government, the private sector, and the academic community should work synergistically.

BEST/WSTB (1992). Review of EPA's Environmental Monitoring and Assessment Program (EMAP): Interim Report, Board on Environmental Studies and Toxicology (BEST) and Water Science and Technology Board (WSTB), National Research Council, Washington, D.C., 25 pp.

Bossler, J. D., C. C. Goad, P. C. Johnson, and K. Novak (1991). GPS and GIS Map the Nation's Highways, Geo Info Systems 1(3), 26–37.

DMA (1991). Digital Products Study: Uses, Standards and Specifications , Defense Mapping Agency, Fairfax, Virginia.

DOI (1992). Audit Report: National Wetlands Inventory Mapping Activities, U.S. Fish and Wildlife Service, Office of the Inspector General, Department of the Interior, Report No. 92-1-790, 54 pp.

FGDC (1991). A National Geographic Information Resource: The Spatial Foundation of the Information-Based Society, Federal Geographic Data Committee, First Annual Report to the Director of OMB, 10 pp. plus 41 pp. of appendixes.

FGDC (1993). Manual of Federal Geographic Data Products, Federal Geographic Data Committee, Washington, D.C.,

MSC (1990). Spatial Data Needs: The Future of the National Mapping Program, Mapping Science Committee, National Research Council, National Academy Press, Washington, D.C., 78 pp.

MSC (1991). Research and Development in the National Mapping Division, USGS: Trends and Prospects, Mapping Science Committee, National Research Council, National Academy Press, Washington, D.C., 63 pp.

OMB (1988). OMB Bulletin 88–11, 1988

Warnecke, L. et al. (1992). State Geographic Information Activities Compendium, Council of State Governments.