SPATIAL DATA AND THE URBAN FABRIC
The basic premise of most federal mapping programs has not evolved as the nation has changed from an economy based on natural resources to one dominated by personal services, manufacturing, and industry The nation grew from its knowledge and exploitation of natural resources. Crossing the continent required (and resulted in) a knowledge of the physical and biological geography. The mapping of its topography gave government and private industry a context for great plans and small initiatives; the mapping of its geology, a sense of wealth and the location of building materials; the mapping of its soils, an assessment of the potential to feed itself and others; and the charting of its waters, the ability to navigate, generate power, and harvest fish.
As the nation has matured, these mapping needs have lost their priority though not their utility. Delivery of human services, assessment of environmental conditions, responses to natural hazards such as earthquakes and hurricanes, deployment of law-enforcement personnel, education of children, economic development, transportation of goods, delivery of services, and the electoral process itself demand a broader and more detailed information set: one that defines, characterizes, and embraces the urban fabric of a once-rural nation.
The federal focus on mapping natural resources, almost to the exclusion of all else, fails to recognize the growing needs of the federal agencies for information more commonly referenced to street addresses, voting precincts, and land parcels. We refer to these needs within the context of an urban fabric, a fabric that includes both urban and rural areas
and consists of information concerning the built and human environment versus one dominated by natural resources. The expenditures for acquiring and encoding information defining and pertaining to places are made in an ad hoc manner and redundantly by federal agencies, local governments, and the private sector. Thus, what could be a national investment frequently goes no further than an expense.
We divided this fabric into the following four pieces:
the land base and use of land,
the ownership of the land,
the transportation (street) network that serves it, and
the addressing schemes that provide common geographic reference.
We discuss these pieces in the above context.
The point of departure for this discussion is the findings and recommendations in the report—Spatial Data Needs: The Future of the National Mapping Program (MSC, 1990)—that digital, spatial data bases be developed to serve both public and private needs of the country. We studied the case of the urban fabric to consider the needs of urban areas for digital spatial data, to determine how well currently available systems for supplying and maintaining spatial data meet these needs, and to determine how the availability and accessibility of urban spatial data in the future might be improved.
An urban spatial data base is a digital system that defines location and/or spatial references (addresses) of objects, people, and events. The committee identified four general kinds of urban spatial data. This chapter addresses the potential contribution of each toward meeting the needs for an adequate spatial data infrastructure for urban areas of the United States and the current status of each one. In Chapter 9 we present recommendations for technical, organizational, and institutional changes and for cooperative activities to meet the needs for urban spatial data in the next decade. The four components of an urban spatial data base are (1) the urban land base system (planimetric and topographic); (2) the cadastral system; (3) the street centerline system; and (4) the political and administrative boundary system (not further discussed).
In some urban areas, government and private organizations are cooperatively developing GIS. However, individual organizations usually develop their own GIS to respond to their needs. No commonly accepted standards for geographical accuracy, content, format, or transferability of data exist. Consequently, sharing information between different organizations as well as relating information from different organizations for common geographical areas is difficult and sometimes impossible.
The federal government also needs spatial information within urban areas for many of its tasks. Its access to this information is difficult, if not impossible, when so many different systems for collecting, updating, and maintaining geographical information exist and when the terms of accessing that information differ from one urban area to another and from one organization to another.
LAND BASE SYSTEMS COMPONENT
Land base refers to activities and data sets (both analog and digital) associated with defining the location and extent of physical (natural and manmade) features on the Earth's surface. The land base typically incorporates:
a common coordinate reference system,
the portrayal of physical features in a planimetric and orthogonal view, and
the representation of topographic relief and features.
Land base systems include:
techniques for collecting information, through aerial photography and remote sensing, about the location, extent, and nature of physical features;
techniques for representing this information graphically and textually;
linkages to other descriptive information about the features;
techniques for replacing and disseminating information, e.g., maps; and
a process for maintaining the temporal currency of the features.
The scale, accuracy, and feature representation, as well as the level of textual and statistical detail characterizing each feature, are varied with little consistency and no standards between the activities and products of similar organizations. However, land base maps are typically the foundation of most spatial information systems, serving to register other layers or themes of information to a particular geography.
No organization or level of government is responsible for building, maintaining, or setting the standards for land base systems. Because of their central import to analog and digital spatial information systems, many organizations have traditionally developed base maps to support their traditional requirements; other organizations have informally adopted (and sometimes modified) the base maps of another organization to satisfy, if only partially, their requirements.
The map products of the USGS and the SCS have served, most frequently, as the base map for federal mapping activities. The traditional natural resource orientation and the scale of the products limit their utility and fail to serve agencies with an urban applications orientation. The USGS and SCS map products are also used extensively by state and local governments and the private sector, with comparable limits on their utility.
The land base mapping of state governments is generally more diffused than that of the federal levels. One consistent exception is the mapping activities of the states' departments of transportation. State agencies rely heavily on federal sources for large-area mapping and to a significant yet lesser extent on the mapping activities of local governments.
Land base maps with wide ranges in type, accuracy, currency, and content are a common product of local government. The land base is constructed by using many different techniques and standards, frequently relying on the private sector as a supplier of some or all of the sources referenced above.
Land-base mapping by the private sector occurs both in support of federal, state, and local government activities and as a component of private business, most frequently in land acquisition, site development, resource management and development, and providing and servicing utilities and transportation networks.
Tens of thousands of organizations are building and maintaining land bases in the United States. The cooperation and coordination in building and maintaining the land base systems rarely go beyond ad hoc relationships within organizations and between organizations interested in the same geography. The ramifications are obvious: lack of consistency between similar or identical data sets, redundancy, lack (gaps) in coverage, and lack of temporal currency among others. The cost to the national economy in terms of productivity and misdirected revenues is extraordinary.
There are no systematic or programmatic opportunities other than isolated and local arrangements to facilitate the sharing of responsibilities in developing land bases between the various levels of participants: federal, state, and local governments and the private sector. If indeed there is a national benefit to be derived from recognizing and building a better NSDI, surely the institutional processes and programs for sharing responsibility as well as data must be established.
We are not without models—limited in coverage though they may be—that might serve as points of initiation. Among them are:
cooperative geologic mapping programs of the USGS and state geological organizations (see P.L. 102–285);
the high altitude aerial photography program of the USDA, the USGS, and state governments;
the cadastral mapping program and the previous funding incentives of North Carolina and county tax assessors;
the funding incentives for land record modernization in Wisconsin; and
GIS consortia such as in Louisville, Kentucky; Knoxville and Nashville, Tennessee; and Indianapolis, Indiana.
Rationale for Federal Involvement
Federal needs justify an initiative to reach a more cooperative and less
redundant spatial data collection process. The benefits derived by other levels add to the public good and national productivity.
The needs of the federal government span all levels of the geography and detail of land base systems. At the same time, the federal agencies are not always best positioned to collect, validate, maintain, or aggregate information required by federal agencies. The examples are many, including locations of buildings, construction of secondary roads, movement/migration of people, and siting and characterization of public and privately owned utility facilities.
THE NATION'S CADASTRE
As William Chatterton (1991, past chairman American Bar Association Committee on Land Records Improvement, personal communication) noted, markets are now linked globally through sophisticated communications technology and have blended into one global trading system. Globalized production and marketing require capital to flow easily across national boundaries. A recent Nobel Prize for economics went to the pioneers of the efficient-market theory in which prices quickly reflect information about property. The work of these Nobel Prize winners shifted the focus to getting information to market as quickly as possible. The growing volatility of the world capital market is a sign that the old system is increasingly going out of control.
The savings and loans crisis and the well-publicized problems in the banking and insurance industries similarly tell us that old safety mechanisms, designed to maintain financial stability in a world of self-contained national economies, are as obsolete as the world they were designed to protect. Even minor glitches in telecommunications and computer systems now have the potential to wreak major havoc in the financial marketplace worldwide.
Transactions in land are the touchstone of a market economy. Public records should, therefore, enable a prospective landowner, whether he is located in New York, Los Angeles, Tokyo, or London, to determine, quickly and unambiguously, the rights, responsibilities, and risk associated with owning a specific tract of land. We already see that land information systems are being driven by the requirements of the market. Most land
transactions in the United States now require title insurance, so that the multibillion dollar market in mortgages can at least be assured, to the extent of the mortgage title policy, that it has a first lien on the premises.
The United States lacks what is commonplace in most of the developed nations and many others: an integrated and comprehensive national land records system. Historically, the responsibility for maintaining land ownership information (i.e., a cadastral system) has been delegated to local governments, with state intervention and coordination occurring only limitedly and inconsistently.
Cadastral system refers to those activities and data sets associated with parcel-based land information (McLaughlin and Nichols, 1987). It incorporates:
a common definition of the parcel and a unique primary parcel identifier,
a cadastral mapping program with geodetic control, and
linkages to a series of records.
Information contained in the cadastral system include:
data about how the parcel itself was created and any subsequent changes,
core cadastral data from primary sources (e.g., ownership data from the registry office, value data from the assessment office, etc.), and
other data that may be referenced to the parcel (such as building and improvements data).
Cadastral information may be related to other spatial data in a variety of ways, including
assigning coordinates to the cadastral parcel,
registering the cadastral map to a base mapping system, and
relating the parcel identifier to other indexes (e.g., street addresses).
The states have the primary responsibility for building and maintaining cadastral systems. They have the constitutional mandate for administering the system of real property law and for matters pertaining to the planning, development use, and taxation of land. They are also responsible for establishing the administrative arrangements (such as registry offices) required to support this mandate.
Much of the daily responsibility has been traditionally delegated by the states to local government. Currently, there are approximately 100 million parcels of taxable real property. The records for these parcels are maintained by approximately 80,000 state and local government entities (including counties, municipalities, towns, townships, etc.).
A few states, such as North Carolina and Wisconsin, have undertaken to develop management programs for county land records resulting in statewide data bases. To date these programs have been primarily concerned with developing uniform standards and with assisting (through funding and technical assistance) in local land records improvement programs. North Carolina, for example, instituted a land records management program in 1977, and developed cooperative programs with more than 80 of its 100 counties (see Chapter 7).
Rationale for Federal Involvement
The federal government is responsible for a diverse group of mandates and functions that require parcel-based information. These include aboriginal land tenure; the federal government's significance as a land owner; its role in real estate and asset/facilities management; its role in acquiring property for specific projects; various taxation roles; its regulatory role with respect to real estate financing, interstate commerce, agricultural support programs, environmental assessment, hazardous waste management, etc.; and civil defense and emergency preparedness roles.
The problems associated with the existing cadastral arrangements have been extensively documented in studies over the past 30 years. The basic
record-keeping systems are often archaic, based on legal principles predating the U.S. Constitution (the rudimentary deed registry concept). Information may or may not be available for a specific parcel (e.g., registration of deed is generally not compulsory) and may be distributed among a number of public agencies. There is generally no systematic procedure for efficiently retrieving the data, and current cadastral index maps are generally not available. The quality of the data stored and their legal validity are often unknown. The ability to integrate parcel data from different sources is generally very limited. The costs associated with these problems show up in the transaction and regulatory costs of the real estate marketplace, the assessment and collection costs of the taxation system, the cost of the record-keeping itself, and others.
Responding to the Problems
Efforts to respond to these problems date back more than 30 years and include proposals aimed at improved survey standards and coordination of parcels, administrative and legal improvements to the deed registry system, automation of land records, development of cadastral mapping programs, integration of parcel data bases, etc. In the early 1980s, many of these proposals were bundled together under the label of the multipurpose cadastre concept, as described in two reports prepared by the National Research Council (Committee on Geodesy, 1980, 1983).
The impact to date has been modest. Although many local governments have invested in specific land record improvements (especially the introduction of computers), a much smaller number have undertaken more comprehensive reforms. The reasons for this limited impact are many and complex but are much more a function of policy and institutional concerns than any underlying technical or financial issues.
The most recent effort to deal with the cadastral problem from a national perspective has been the Department of the Interior study (DOI, 1990) on land information mandated by the Federal Land Exchange and Facilitation Act of 1988 (P.L. 100–409). Essentially, it once again reviewed the problems associated with land information management in the United States (with an emphasis on cadastral records); examined a number of recent initiatives at the federal, state, and local levels; and called for a nationwide land information management system. The report's recommendations have gone unheeded. The timing of undertaking land records
reform will never be better because the technology is ripe and the task becomes more complex with time.
STREET CENTERLINE SPATIAL DATA BASE
The transportation network is the third and increasingly important component of the urban information fabric. Its importance transcends the utility of a common street map because it is a base for defining, organizing, and accessing places (and their associated information) within both complex urban environments and rural areas.
In a digital format the importance is magnified and its uses are expanded manyfold. What might have served as a guide or descriptor of pathways through a particular geography can now serve as an index to large volumes of tabular and statistical data, as a framework for depicting those statistics spatially, and as an analytical tool.
In its basic form, a Street Centerline Spatial Data base (SCSD) is a computerized street map, where streets are represented as centerlines and the characteristics of the streets (the attributes) are appended. The practice that has developed from almost three decades of experience has led to records that uniquely identify each segment and intersection of the street network; differentiate between the left and right side of each segment and anomalies such as cul de sacs; encode street names, zip codes, census geographies, and other area descriptors; provide address ranges for both sides of the street segment; and incorporate coordinate references and scale at various levels of accuracy, among other characteristics.
Significance and Applications
A national SCSD with these characteristics, if kept current and accurate, would support a wide variety of administrative functions by all levels of government and would also provide a basic spatial framework for extensions to serve a wide a variety of other applications of commercial importance. Applications include facility site selection e.g., schools, health facilities, and transit nodes), socioeconomic planning studies using census data, legislative redistricting, and analysis of demand for and supply of human services based on administrative records address-matched to areas
with census data. Operational Intelligent Vehicle Highway System (IVHS) applications of SCSD in logistical operations (e.g., fleet management) promise to reduce driving time and fuel usage. Direct benefits extend to increased productivity of professional drivers and commuters alike, reduced emissions from burning fossil fuels, and reduced dependence on foreign energy sources.
Current Status of Street Centerline Spatial Data Bases
Although no single national data base fills all requirements for an SCSD, products from three agencies provide part of the content of a national street-centerline resource.
USGS DLG Program
During the 1980s the USGS embarked on a program to produce DLGs corresponding to most nationwide paper map quadrangle publications. The 1:100,000 scale DLG series was completed in the 1980s and served as the coordinate and linework basis for nonmetropolitan portions of the Bureau of the Census TIGER data base (see below).
The transportation layer of the 1:24,000 DLG series in April 1992 existed for 5,360 of the 55,000 quadrangles. This series is important because its coordinate accuracy is sufficient for many SCSD applications, and it is derived from the largest-scale (most detailed and accurate) uniform map series available in the United States. Despite coordinate accuracy, the data in the series are limited in applicability because they lack street names and address ranges and are infrequently updated, the coverage is limited, and the accuracy levels are inadequate for many uses.
U.S. Postal Service's ZIP + 4 Data Base
Although technically not an SCSD, the U.S. Postal Service's (USPS) ZIP + 4 file is an important national resource. ZIP + 4, the most detailed postal data base, relates all U.S. addresses to nine-digit numeric codes, each of which usually ties a small number of addresses to one side of a city block. The first five digits are the familiar zip code, which corresponds to a postal delivery unit. Each zip is divided into about 40 sectors (the next two digits), each of which can have 99 segments (the last two digits). In contrast to DLGs and TIGER, ZIP + 4 is updated monthly to permit large
volume mailers to precode mailing addresses with postal codes. Mailers benefit from postal rate discounts; the USPS gains by streamlined bulk-mail processing.
Unlike DLG and TIGER, ZIP + 4 is not organized as an SCSD. Each logical record in ZIP + 4 can define a range of addresses, part of a rural carrier route, a box number, or even an individual person or company. The ZIP + 4 file contains no spatial coordinates or census codes. Yet, the ZIP + 4 is an important resource to the national SCSD because (1) it is a complete definition of mailing addresses throughout the United States; (2) addresses usually are the most prominent geographic identifiers in people, property, or event-oriented data files; (3) it is updated monthly, making it by far the most current geographic reference data base; and (4) it is widely used as the standard for street naming.
TIGER (Marx, 1990), a by-product of the 1990 census, is the closest of the three examples to a broadly useful SCSD. It covers the entire country and is the authoritative repository of the 1990 census geographic definitions. Although TIGER's coordinate content is immense, accuracy and graphical appearance vary because of past cartographic and data processing practices. TIGER coordinates are useable for many applications but fall short of the consistent accuracies of the 1:24,000 DLGs and the needs of many users. Another deficiency of the TIGER files is that only streets in urban areas carry address range and ZIP code attributes. Consequently, only 55 to 60 percent of U.S. addresses are represented in TIGER, which severely limits TIGER's utility. Moreover, as TIGER is a work product of the 1990 census, a schedule of periodic updates is not published.
Problems Caused by Deficiencies of Available SCSDs
A principal problem with the present situation is that there are three spatial data bases, developed to serve the needs of their respective agencies, each containing data that are not easily coordinated with data from the others. Levels of spatial accuracy differ, one (ZIP + 4) contains no spatial coordinates, and the temporal updating of information is different and uncoordinated for the three sources. Because many user needs are not met by any of the above sources, a second problem is created
when users of SCSDs frequently add attributes of interest by themselves to one of the existing SCSDs without consulting with or cooperating with other public or private users. The result is a costly duplication of effort and lost opportunities for acquiring a superior data base.
Although designed to facilitate specific agency internal operations, both TIGER and ZIP + 4 are widely used in a broad range of state, local, and commercial applications. Again, inconsistencies persist. The Bureau of the Census actively markets TIGER and encourages TIGER use with workshops. TIGER is used for marketing studies, dispatching taxis, making inventories of municipal street signs and streets, and routing fleets of vehicles, for example. On the other hand, the USPS uses copyright restrictions to discourage use of ZIP + 4 for purposes other than preparation of bulk mailings. Nevertheless, information from the ZIP + 4 data base is being tested by at least one state as a mechanism for allocating state income tax to school districts and by marketers for demographic analysis of customers and prospects.
Because versions of an SCSD are being developed independently, little thought is given to issues of standards and formats. The flow of information is now dependent on mandates and missions of specific agencies, legal and market considerations, budgets and profits, poor communication linkages (organizational, personal, and technical), and generally the lack of coordination and cooperation. Adding to the complexity and inconsistency, several private companies have created proprietary data bases combining the best data from TIGER and ZIP + 4 and proprietary data sources.
Need for Greater Coordination or Consolidation
There are inter-and intragovernmental opportunities for shared development, collection, and use of spatial data. Additionally, there are valid roles and potential for private sector participation in cooperative efforts as well as in providing value-added products. As an example, information can gain value through the accumulation of incremental improvements, features, detail, and validation or definition of accuracy as it is passed from one governmental level to another or from the governmental bodies to the private sector and on to the end user. The added value is as much a function of the entity having the resources and the mandate (whether legal or profit motivated) to contribute.
With foresight, the development life cycle of information can be used to add value and accessibility to information. Moreover, the design of the information resource can become dependent on the interactions of the various participating levels and with coordination the actions of any particular level can be maximized, realizing the potential efficiencies of a robust NSDI.
There is a need to find the resources to forge the permanent cooperative arrangements that will be required to produce and keep current the SCSD. Many application areas have special needs that are not covered by the current SCSD. For example, TIGER is still missing the kind of attribute information required for transportation network modeling and logistics. The most important are direction of flow, turn restrictions at intersections, travel times on segments by time of day, and turning penalties at intersections. In addition, there should be a minimum set of transportation-related attributes. State departments of transportation could add their route identifier and beginning and ending milepoints to each segment in the SCSD so that they can more easily reference other attribute information, such as pavement condition, accidents, and signs.
Conclusions: The Need to Ensure Access, Use, and Maintenance
An urban SCSD is an essential component of the urban information fabric. There is a clear federal need for a city block/street-level spatial data infrastructure, evidenced in support of the missions of the Bureau of the Census and the USPS. Yet each agency continues to maintain its data base independently.
Clearly, no single federal agency has responsibility for the development and maintenance of a fully functional SCSD. Several agencies have responsibility for pieces of information contained in the SCSD but, like a jigsaw puzzle that is never put together, none has the responsibility nor the duty of ensuring that these pieces fit together. The only two nationwide data bases (ZIP + 4 and TIGER) are byproducts of the operational missions of their organizations. Furthermore, the ZIP + 4 data base is copyrighted by the USPS to control distribution of outdated copies, the use of which could snarl mail processing. The only general-purpose data base—the 1:24,000 DLG—is incomplete and lacks key data fields required for broad usage. Furthermore, none of these agencies has plans to augment data bases with information about traffic-flow restrictions, which would support
future IVHS applications and other logistical operations such as dispatch routing, which has substantial commercial value and utility to governmental units of all levels.
There is a natural but undefined opportunity for sharing spatial data through the entire domain of federal, state, and local governments and the private sector regardless of the specific geography involved. The commercial utility of an SCSD is great enough to encourage several private companies to fund independent multimillion dollar projects to improve public SCSD resources, turning them into proprietary data holdings. Moreover, each of these developers, collectors, and users of spatial information is making considerable investments that mask significant opportunity costs because of redundancy of effort, undeveloped opportunities for efficiencies, and insufficient financial resources available to any one sector.
Committee on Geodesy (1980). Need for a Multipurpose Cadastre, National Research Council, National Academy Press, Washington, D.C., 112 pp.
Committee on Geodesy (1983). Procedures and Standards for a Multipurpose Cadastre, National Research Council, National Academy Press, Washington, D.C., 173 pp.
DOI (1990). A Study of Land Information, Department of the Interior, Washington, D.C., 61 pp. plus appendixes.
Marx, R. W., ed. (1990). The Census Bureau's TIGER System, Special Issue, Cartography and Geographic Information Systems 17(1), 133 pp.
McLaughlin, J., and S. Nichols (1987). Parcel-based land information systems, Surveying and Mapping 47(1), 11–29.
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.