SEPTEMBER 1993

Prepared by
Federal Geographic Data Committee Ground Transportation Subcommittee

Table 1: Areal Coverage Requirements

In the contiguous (continuous) networking sense for ground transportation, it is apparent that the conterminous 48 States are a common denominator that all agencies can agree on. However only two of the respondents indicated a requirement for only this level of areal coverage. Nine of the respondents indicated at least a 50 State interest, with four of these indicating a need also for territorial coverage. Four respondents also indicated the need for international coverage, particularly into bordering Canada and Mexico, to support eventual international networking connections.

While the majority of the member Federal agencies that responded have a need for maximum national areal coverage extending to all 50 States, some agencies indicated a requirement for ground transportation data only within specific areas of the 50 States for which they have oversight or mission responsibility (and to a limited extent to surrounding areas connected by proximity or by environmental or ecosystem relationships). These agencies include the USFS, BLM, NPS, and MTMC.

2.1.2 Locational Accuracy

The locational accuracy required by a particular agency is related to the scale at which its maps are produced or from which its digital data are derived. Some agencies require that only a spatial or topological relationship be maintained rather than adherence to a particular scale or locational accuracy standard. Usually the agencies that require a more rigorous locational accuracy standard produce or use cartographic data at larger scales. The responses from the member agencies are shown in Table 2.

This question reveals as much about map scales in use in various agencies as it does about actual locational accuracy requirements. For example, USGS produces and maintains cartographic products in all the map scales listed in the above table. The Census adapts map scale to the size of the tract to be portrayed or product to be derived. An unintentional bias in the phrasing of the question may have limited many of the respondents to answer the question only in reference to the scales suggested in the question. It is clear, however, that a lot of mapping takes place at the 1:100,000 and 1:24,000 scales, since all of the respondents indicated a requirement for ground transportation data at one or both of these scales.

Overall, the responses indicate that a minimum horizontal locational accuracy desired for ground transportation networks is that associated with 1:100,000-scale map products for large networks, with the ability to thin data to smaller scales up to 1:2,000,000. Smaller scales (1:2,000,000) are adequate for simulating traffic flows and keeping file sizes at levels sufficient for large area analysis, but the larger scale (1:100,000) is required for national highway planning and the more precise location of rail/highway crossings, accident locations, etc. For urbanized areas, ports, parks, military reservations, government installations, and other small geographic areas or areas with complex transportation networks, 1:24,000-scale locational accuracy is preferred.

2.1.3 Map Projection

On the issue of establishing a standard map projection, member agency responses to this question included the following: 1) that the projection be documented; 2) that the geographic coordinate system be useable on the projection; 3) that an equal area projection was a preference for small scale maps; 4) that a UTM grid was a preference on large scale maps; and 5) that there is no one projection that can solve all user needs.

Due to technological advances in software conversion packages for projections, the issue of map projection was perceived as relatively unimportant. None of the respondents had a request for a specific projection, but asked for the above mentioned characteristics if in fact a near consensus was reached on the adoption of a standard map projection. Most respondents indicated that it is more important to know the projection of the map or data, and to have the ability to convert between projections as needed.

Generally, while it was felt that establishing a standard map projection may be a desirable objective to achieve consistency in any given map or data product series, it was recognized that the selection of a specific map projection is driven by the specific user application for the data or product. Therefore, most respondents felt that the choice of map projection should be based on the user's requirements and capabilities; map and data producers and users should be free to use the best suited projection for a given application. It was generally agreed, however, that data base content referenced to latitude and longitude will make the combining of data sets, such as population data, boundary files, and other modal networks, derived from various graphic and digital data base sources more straightforward. For this reason, it is recommended that absolute latitude and longitude be adopted as the most useful standard for linking data bases to support national level ground transportation networking.

2.2 Modal Networks

The survey asked participating agencies to identify the modes for which geographically referenced data are required, and how the networks are defined within each agency. Although the questionnaire was oriented toward network data, the respondents also included several point or facility features which were important to their GIS concepts. The following paragraphs define the networks in more detail, as seen by the responding agencies.

2.2.1 Highway

2.2.1.1 Existing Products.

Automated highway data are currently included in products of the FHWA, USGS, and Census. FHWA is the institutional sponsor of the National Highway Planning Network (NHPN), along with the Department of Defense (DoD). The NHPN is maintained on mainframe computers at Oak Ridge National Laboratory (ORNL). It includes some 400,000 miles of rural and principal urban arterials and identifies the Strategic Highway Network (STRAHNET) of highways important to national defense. The NHPN was developed at a scale of 1:2,000,000, or an approximate locational accuracy of 1200 meters. An ongoing project of FHWA and ORNL is the upgrading of the locational accuracy of this system to correspond to a scale of 1:100,000, or about 100 meters. The USGS Digital Line Graph (DLG) databases represent all roads and most trails in their area of coverage, although the attribute data are insufficient for most transportation analysis. The Census Bureau TIGER (or Topologically Integrated Geographic Encoding and Referencing) files represent all roads and streets which lead to at least two residences. Thus, existing products cover a wide range of definitions for the highway and road networks.

2.2.1.2 Stated Requirements.

FHWA's ongoing efforts will create a highway network at scale of 1:100,000 for the 50 States, Puerto Rico, and the District of Columbia, suitable for national planning purposes. This network will include State designated rural and urban principal arterials, plus any other State-proposed National Highway System link not functionally classified as a rural or urban principal arterial. The network has been made a part of FHWA's Highway Performance Monitoring System (HPMS) and will be updated annually. The level of detail and locational accuracy of the FHWA highway network seems to meet the requirements of most other agencies with national level planning responsibilities, such as DoE. Agencies with a more localized focus require higher resolution, but a smaller coverage area. The USFS desires to record forest development roads open to public travel, together with all public roads in the "area of interest." The NPS desires all roads listed in the NPS Park Roads manual, together with trails and parking lots. The BLM wishes coverage of low volume roads and trails for foot, horse, bicycles and snowmobiles, although networking capability is desired only for low volume roads meeting USGS Category 3 criteria or higher, and for linking major trail systems with minor feeders. MTMC has a traffic engineering interest in roadways and parking lots on DoD installations. This broad disparity in the definition of the network of interest suggests that a roadway database for general Federal use should consist of a hierarchical sequence of geographically consistent files.

2.2.2 Railway

The FRA was the only respondent to describe the extent of the rail network. FRA's view is that the system consists of all railroad track in the United States, and the agency expects to represent most of this system as GIS data. Early work is capturing all rail routes, but not the intricate local rail of metropolitan areas and classification yards, nor the trackage associated with urban transit systems. FRA's initial rail database has been publicly released by the Bureau of Transportation Statistics (BTS). The USGS collects DLG data for all railroads in the United States--indeed, USGS data formed the initial geographical basis for FRA's rail database. DoE also owns a proprietary rail database containing all U.S. railroads, including spurs into DoE installations. The Subcommittee is aware of, but has not examined, other rail databases of a proprietary or sensitive nature which are not available for general Federal or public use.

2.2.3 Inland Waterway

Respondents who desired waterway data referred to it in the context of networking with other modes. Today, both USGS and the CoE maintain data on the waterway system. USGS collects information on certain surface water features, including streams and rivers. DoE also has an existing barge database which includes all waterways with depths of nine feet or greater. A consortium of Federal agencies and academic contributors, including CoE, DoE, Vanderbilt University, MTMC and others, is engaged in creating a non-proprietary waterways base for general public use. This database is being developed at a 1:100,000-scale and will include all waterways supporting commercial traffic (generally considered navigable if depth is at least 9 feet), intracoastal waterways, and connections to sea lanes. The network covers the level of major movement of commercial goods from major terminals, such as grain elevators. Its extent is fairly well defined in Congressional legislation. Docks, locks, terminals and other waterway facilities will be identified in separate point databases.

2.2.4 Pipeline Network

Six of 11 respondents indicated interest in a national pipeline database. The USGS collects DLG data on pipelines at present, which provides at least a geographically referenced network against which quantitative attribute data can be registered. DoE also has data and experience working with oil and natural gas pipeline systems. MTMC has an interest in pipelines on behalf of DoD, since petroleum product sourcing and transportation is a critical portion of the deliberate war planning process. FHWA, MarAd and the CoE Navigation Data Center expressed only a conditional interest. The subcommittee's view is that development of a national pipeline network should be a future goal, following the development of highway, rail and waterway networks.

2.2.5 Power Transmission Lines

Only Census specified a need for power transmission network data, with the remark that such visible, above-the-ground features could serve as boundaries for statistical areas. USGS has collected DLG data on at least some transmission lines, while DoE has done considerable work with power transmission networks in the past. For the user community represented by committee members, there appears to be little interest in power lines as transportation entities. The subcommittee sees development of a national power transmission line database as another future goal, following the development of higher priority modes.

2.2.6 Other Modal Databases

Several respondents cited needs for locational and attribute data on facilities or networks which were not within the sphere of interest of others. These included the following:

2.2.6.1 Seaports

Only FHWA specified a need for port data in response to the original questionnaire. MTMC later expressed a need for such information. It is likely that this small number was due to a network orientation in the questionnaire, which did not elicit facilities data requirements. Both MarAd and MTMC have considerable corporate information on seaports, but these data are scattered and not now integrated into a GIS structure. MTMC is currently developing an automated seaport database for support of its enhanced Port Operational Performance Simulator (POPS), a software complex for determining port throughput capacity under various scenarios. It is anticipated this database will be in the public domain. POPS is not currently geographically based, but MTMC is interested in modeling the transportation subnetworks on the port and also in evaluating the impact on port throughput of safety restrictions (such as quantity-distance restrictions on ammunition transshipment). This requires detailed information on areas outside the port gate.

2.2.6.2 Airports

Both FHWA and NPS noted a need for airport data. FHWA desired the information for intermodal considerations, while NPS regarded airport data as a useful non-network addendum. USGS, DoE and the Department of Transportation (DoT) have developed airport databases. USGS collects DLG data for aircraft facilities and runways, while DoE has worked with the Federal Aviation Administration (FAA) on airspace networks, airports and air traffic activities. DoT's Volpe National Transportation Systems Center (VNTSC) has combined and edited FAA files on airport data, and issued the resulting structure as a GIS database in support of MTMC's Strategic Transportation Analysis (STA) program. This information has also been released to the public on CD-ROM by the BTS.

2.2.6.3 Fixed Transit Facilities

Of the questionnaire respondents, only Census expressed interest in urban trolley and subway system networks, noting their possible use in routing enumerators. The subcommittee agreed that development of a set of urban transit facility databases would be useful but of lower priority than the national networks described above.

2.2.6.4 Intermodal Facilities

Only MTMC noted a specific need for location and throughput capacity data for facilities at which shipments move between rail and highway modes. However, FHWA and FRA interest in intermodal analysis suggest they could also use such information. The subcommittee's collective view is that development of a national database on this matter should be pursued subsequent to development of the modal network databases.

2.2.6.5 Parking Lots and Marinas

The NPS identified vehicle parking lots and small-craft anchorages as needed in their developing GIS. These data are not required at a national level by other agencies.

2.2.6.6 Bridges

MTMC and the BLM cited a need for bridge capacity and condition data. FHWA also has a need for bridge data and maintains some of these data in the National Bridge Inventory (NBI). Engineering assessment of the passability of bridges and overpasses can be accomplished using the NBI and other programs such as the American Association of State Highway and Transportation Officials' (AASHTO) Bridge Analysis Record System (BARS). MTMC, DoT, and BTS are pursuing methods of assigning a military load classification code to the NBI bridge set. This information may also satisfy the BLM data requirement. USGS maintains DLG data on bridges, providing location, name or identification, whether or not covered and whether or not multiple decked. This base can support at least a partial check on State furnished bridge locations, given common keying or any other applicable cross-identification algorithm.

2.3 Intermodal Connection Needs

Most of the respondents who have a national level network responsibility (in contrast with agencies whose interest is more focused on specific local areas like Federal lands) expressed interest in developing intermodal connections between national highway, railway and waterway networks, and to a lesser extent with a national pipeline network. The primary purpose for such intermodal connections would be to facilitate the analysis and display of freight flows by combinations of alternative modes. For example, FRA is interested in identifying truck and rail interchange locations in order to simulate the movement of intermodal shipments; MTMC is interested in identifying alternative routes for military cargo movements and comparisons of transhipment times among these routes; and MarAd is interested in examining intermodal landside access to seaports.

There was no elaboration by the respondents on how intermodal connections should be depicted among the current single mode networks. However, in order for intermodal connections to be established, the single mode networks should be at or near the same level of locational accuracy and level of detail (e.g., 1:100,000 or 1:24,000). Furthermore, on order to be able to implement routing or flow algorithms which assign flows to the network, intermodal connections must be represented either by a node or by a pseudo link whose attributes reflect the time, cost, and inconvenience associated with transferring from one mode to another.

3. MINIMUM NETWORK ATTRIBUTE NEEDS

3.1 Highway

The data elements listed below (in no particular order) are those regarded as minimum essential by at least one respondent to the questionnaire. Also shown for each element is the set of respondents submitting the item. A leading asterisk indicates that the data element is one included in the NHPN database maintained at ORNL. The rural arterial and urban principal arterial links in the NHPN will include a linear referencing system and attributes derived from the HPMS. This does not mean that the data will be available for the full range of roads and trails each respondent sees as the scope of the base. The presence of USGS as a nominating user implies that the element exists in USGS DLG or DLG-E standards.

3.2 Railway

The data elements listed below (in no particular order) are those regarded as minimum essential by at least one respondent to the questionnaire. Also shown for each element is the set of respondents submitting the item. A leading asterisk indicates that the data element is one currently carried in the FRA National Railroad Database. As before, the presence of USGS as a nominating user implies that the element exists in USGS DLG or DLG-E standards.

3.3 Waterway

The data elements listed below (in no particular order) are those regarded as minimum essential by at least one respondent to the questionnaire. Also shown for each element is the set of respondents submitting the item. A leading asterisk indicates that the data element is one currently carried in the National Waterway Network Database. As before, the presence of USGS as a nominating user implies that the element exists in USGS DLG or DLG-E standards.

3.4 Other Networks

At this time, no information has been developed for the Subcommittee on member data requirements for pipeline, power transmission, fixed transit and other networks considered candidates for future development.

4 NON-TRANSPORTATION RELATED ATTRIBUTES

The questionnaire asked what non-transportation geographic features would be needed to make network features useful in GIS applications. The list below is compiled from the member agency responses. It presents the non-transportation related features at a gross level and indicates the depth of interest in such data among the agencies that responded. Although some of the responses (features) are arguably directly related to ground transportation, the agency felt compelled to list them to be sure that the features were collected.