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

Before about 1960 the dominant medium for recording and transmitting information about geographic location was the map, a highly conventionalized analog image, usually ink on paper. The map has certain unique values as well as profound limitations.

In a sense, the total store of information about the geography of a nation in the past was equivalent to all existing maps. Because it is difficult or impossible to aggregate or cumulate maps or mapped information in any practical way, this store of information was widely dispersed and frequently encapsulated in particular applications. Except for a few major programs, most importantly the national topographic map series produced by the National Mapping Division (NMD) of the United States Geological Survey (USGS), mapping has traditionally been application specific. Maps are almost never made strictly for the sake of mapping; they are tools, having value only in allowing their users to do other tasks: for example, manage land records, run paper companies, build highways, carry out a census, or search for minerals.

Many public and private sector organizations that traditionally made maps or caused them to be made were not primarily mappers and viewed mapping strictly as a cost of doing business. Once used or applied, much map information was stored or discarded. It never became accessible to other potential users, and thus could not be said to contribute significantly to a national store of knowledge.

One could in the pre-1960 era conceive of a social-economic entity called "the cartographic enterprise," that is, all individuals and institutions involved in the production, use, and dissemination of maps. Except for the widespread use of USGS topographic maps as a source of common base

data, there was little integration among the components of this entity. The cost and time associated with the compilation, drafting, printing, and distribution of paper maps prohibited such integration.

The introduction of the digital computer in the 1960s and 1970s led to two major developments in mapping. First, the printed map can be produced by using the new, more flexible digital technology. Second, the printed map is increasingly being supplemented and replaced by computer-based geographic information systems (GIS), which treat maps as a series of spatially integrated layers. While paper maps originated to supplement human memory and vision, GIS do this and more: they supplement human cognitive or information-processing capabilities as well.

In the GIS era it is no longer adequate to speak of a "cartographic enterprise" with its connotation of simply making paper maps. Instead we must develop new conceptions more useful in describing and analyzing how geographic data (any data referenced to location) are acquired, processed, disseminated, and used.

Unlike maps, strings of geographic or spatially referenced digital data can be aggregated, transformed, and shared. Spatial data can now be more easily isolated and abstracted from the particular application in which it was developed and channelled into other settings and other GIS where it can be reused, enhanced, and routed to other potential user communities. The old "top down" model (especially appropriate for base data from NMD and other federal agencies) is inadequate to represent the multidirectional alternative information flows that are now technically feasible.

Spatially referenced digital data can perhaps be thought of as molecules of water that in aggregate form a circulating fluid, flowing freely from application to application (or from system to system). Conceived of in this fashion, information can be seen to take on value and become a marketable commodity, quite apart from the context, need, or application for which it was originally developed.

An important qualification to this concept must, however, be clearly stated. It is often said that our nation's economy is becoming information-based, and statistics are produced to show its considerable economic value. It is critical to recall (as many journalists do not) that while information seems to have a reality of its own, it takes on value only with reference to authentic value-producing activity, that is, only when it is about something. For example, even though a timber products firm might like to buy rather than produce basic mapping of its resources and facilities

and might assume that there would be a market for such information per se, the need and therefore the market for such data may exist only as long as the primary business (timber production and sale) itself exists. This is obvious but often seems to be overlooked in discussions of the information economy.

Considerable quantities of spatial data are generated for special or even unique purposes, and it is unlikely that much of it needs to circulate widely. In addition, proprietary business data, conceivably of very general interest, also would not be publically shared.

Because of rapidly advancing capabilities and uses of technology, it is often difficult to understand and visualize the most profound implications of the computer for handling geographic information. GIS require vast amounts of digital spatial data to function and because these data are not commonly derived by digital sensing of geographic reality (much of which is not physical anyway), the users have been forced to rely on data derived from analog maps. Consequently, much of the total investment in GIS for decades has been in a continuing, tedious, and expensive process of converting analog (paper) maps into digital data bases. Furthermore, given the application-oriented nature of mapping, including GIS, the committee has seen examples of waste and redundancy as the same maps are repeatedly digitized by different organizations.

The maps or categories of information on maps that are repeatedly being digitized at high cost (primarily to the tax-paying public) are what are commonly called base maps or base data. It can be safely assumed that the most commonly digitized base maps are the 55,000 sheets of the NMD's 1:24,000-scale topographic map series and that the most commonly digitized data categories include graticule, shorelines, drainage, political boundaries, and transportation routes.

In retrospect, it might have been wise early on to digitize these categories nationwide, as a one-time-only crash effort. This was not done, and the nation continues with a disproportionate share of its spatial data resources devoted to repeating base data digitizing rather than creating new, improved data.

The MSC (1990), in its report Spatial Data Needs: The Future of the National Mapping Program, focused on the role played by the USGS in providing spatial information to other federal agencies and to the nation as a whole. During most of the USGS's history, this information took the form of topographic maps, most importantly and most recently nationwide coverage at the scale of 1:24,000.

However, in the past few decades the demand for spatially referenced information shifted dramatically from analog maps to digital data. As a result, the USGS, the only organization offering a comprehensive national reference map series, came under intense pressure to respond to this change with an altered or expanded program. Spatial Data Needs was produced at the USGS's request, and it contained a series of recommendations designed to assist the USGS to adapt to changing user requirements.

In Spatial Data Needs the phrase "national spatial data infrastructure" appeared several times, its meaning clarified only by context. At that time no effort was made by the MSC to define the concept precisely, or to examine the roles played by government agencies other than the USGS. However, in 1990 the responsibilities of the MSC changed to include a broader perspective on federal mapping organizations (with the word mapping used here as a very general term encompassing all spatial data collection and use) and, less directly but crucially, the mapping activities of the nation as a whole. With this change in perspective, the need to better articulate our objectives became apparent. Therefore, the MSC considers that its mission includes responsibility for the NSDI as follows:

The MSC has for 2 years been gathering information about the programs of a wide variety of mapping organizations in the public sector (federal, state, and local) as well as from a more limited selection of firms in the private sector. The MSC is now in a position to provide both a more comprehensive definition of a national spatial data infrastructure and a series of recommendations to improve its utility. The purpose of this report is to:

• describe the concept and evolution of the NSDI;

• document the success or failure of the current NSDI, with evidence based on an examination of specific spatial data domains;

• analyze impediments to success and, where possible, propose ways these impediments can be removed; and finally,

• propose the creation of incentives, which probably will nee to involve many institutional changes, that will lead to the creation of a more useful and cost-effective NSDI for the nation as a whole.

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.