Geographic Representation

It is estimated that the United States covers 3,586,498 square miles, which includes 144.2 million privately owned and 8 million publicly owned land parcels. Each parcel has at least one public or private owner. In this model, each of these parcels would be represented as a closed polygon that approximates its legal boundaries or at least a point in the initial phases. Eventually, as in many developed countries such as Australia there would be a set of mutually exclusive and comprehensive polygon features with geographic coordinates that define every part of the United States. It is estimated that about 70 percent of the parcels are already represented in digital form and that the effort to complete the coverage is technically and economically feasible. In fact, most state governments require local governments to produce tax maps that depict the boundaries of each parcel. Even though these parcel polygons are only a sketch of a legally defined parcel they can fulfill the basic needs for a national system. Once the initial set of parcels has been established, over time further integration can be done to address the more complex accuracy issues, including the absolute accuracy of the boundary data. There are several acceptable procedures for converting existing tax maps into digital parcel data. Furthermore, many communities have utilized legal descriptions and ground survey measurements to produce highly accurate parcel boundaries. For those parts of the country that are not covered by tax maps or definitive descriptions of parcel boundaries it will be acceptable to initially represent a parcel with a single point. These points can be gathered directly on the ground through global positioning system technology, geocoding information from the legal description, or capturing them from geographically registered imagery. Following this approach it would be possible to quickly represent each land parcel in the United States as a polygon or point-level feature that can be associated with an owner or owners. Of course, the long-range goal would be to have complete coverage of nonoverlapping parcels with shared boundaries that are as accurate as possible.


The benefits of any parcel data system are associated with the information relating to the parcel. While a tax map can provide a graphic depiction of parcel boundaries it does not convey information about the ownership, surface rights, easements, land use, or value of the land. Local governments require this information to build an equitable property tax system and many also use it to support applications relating to land use and emergency response. There are common needs in each community and there are also needs for regional coordination. At the national level the need for sufficient parcel attribute information must be balanced with legitimate concerns about personal privacy and confidentiality. Therefore this national model should be limited to only a basic set of attributes that would support discovery and navigation of relevant parcels. From a functional viewpoint the base set of attributes would enable one to

  1. Uniquely identify each parcel in the national database;

  2. Link a parcel uniquely back to its source provider;

  3. Provide basic information concerning the parcel geometry;

  4. Locate its street address; and

  5. Identify the owner type.

(Note that in cases where parcels do not normally have a street address, such as agricultural or timber lands, address information would not be required, and the address field would be empty or null values.)

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