PART III: The Application of GIS to Multimodal Investment Analysis


Transportation modeling activities have traditionally recognized that movement through transportation networks is complex, especially when considering configurations of several modes of travel (1). Current and historic patterns of mode choice are used to dictate mode split, or the proportion of trips that will be undertaken by each available mode. Agricultural products are transferred by truck, rail, and waterway. Passenger travel is accomplished by private automobile, public transportation (rail or bus), airplane, waterway, and bicycle. Multimodalism considers the contribution of each of these modes to the overall movement patterns of people and goods, while intermodalism considers the interconnectedness of these modes. Along with the consideration of a multimodal system, models have been extended to account for social, economic, environmental, and aesthetic opportunities are referred to as "unified", "integrated", or "comprehensive" (2).

In reality, movement of people and goods does not occur solely on the basis of a single mode. In other words, grain may be moved from the field by truck to a storage location. From the storage location it may be moved by train to a processing plant and ultimately delivered by truck to stores or consumers. Trip chains for passenger travel can also involve multiple modes. A work trip commute might involve driving an auto to a rail or bus stop, with the possibility of additional transfers within or among modes to complete the trip. A significant new improvement to modeling efforts can calculate modal transfer links based upon spatial and temporal proximity (3). In this way, travel activity modeling can be more realistically simulated.

Visualizing transportation networks, especially involving multiple modes is very cumbersome in a non-graphic environment. A zonal trip time/distance matrix by mode can be constructed, however, without an explicit spatial element. On the other hand, a (geo)graphical map that defines network topology is not only more easily interpreted in geographic space, but it also defines the relationships between the transportation network and surrounding activities such as population densities, employment locations, and land use patterns. These activities are factors in dictating not only the level of transportation demand, but also the spatial distribution of movement, as well as the modal requirements of transportation demand. The spatial arrangement of transportation demand and supply of transportation facilities are perfectly suited for a geographic information system (GIS) environment for visualization purposes and also for the data management requirements associated with both demand and supply characteristics.

The purpose of this paper is to highlight the key the key issues involved with the application of GIS to multimodal investment analysis. As will be discussed, there are considerable data requirements involved. GIS offers a variety of analytic functions that enhance data base management activities as well as spatial analysis. Using GIS for multimodal investment analysis will merge transportation network analysis with regional economic impact modeling. This represents an innovative use of GIS, especially in the area of feedback between travel patterns and economic activities. GIS technology is still evolving, with advancements in modeling capabilities currently in the development stages. Advances in GIS, along with tremendous improvements in computing speed, graphics quality, and Internet accessibility may change the face of modeling activities in the near future.

Role of GIS in transportation modeling efforts

Transportation demand analysis has been greatly enhanced by the use of GIS. Using travel demand characteristics such as population, employment, and land use, "what-if" scenarios can be tested (4). The graphical, map-based interface provided by GIS enhances data input and management capabilities. GIS data aggregation functions can be used to easily assign demand characteristics to nodes on a transportation network. Once transportation demand indices have been associated with nodes, the data can be ported to an urban transportation planning system (UTPS) package such as TranPlan. After the modeling procedure is completed, results are transferred back into the GIS for graphic display of projected traffic activity. One of the most common uses for GIS in modeling continues to be the display of transportation system attributes (5).

Other uses of GIS for transportation modeling include traffic analysis zone and transportation network generation. Polygon analysis (overlay and buffer) can help to determine optimal zone sizes and geography. Two objectives of traffic analysis zone construction are homogeneity and contiguity, which can be easily tested with a GIS (6;7). When zones define areas that exhibit homogenous household and land use characteristics, transportation demand can be more effectively predicted. In addition, the network topology capabilities of GIS assist in transportation network preparation. Operations such as ArcInfo’s "clean" and "build" operations can determine whether network links are complete and continuous.

GIS and regional economic models

Regional economic modeling has been traditionally carried out with limited spatial specificity. Economic characteristics for each geographic unit (e.g., region, state, county, MSA, city) along with the likelihood of each region to interact with other regions, are the foundation for analysis. Conceptually this can be structured in a matrix format, where geographic space does not need to be represented realistically (i.e., map form). Instead, cells of the matrix signify discreet geographic units and the attribute data provides economic, social, and spatial definition. Although such a model includes a "distance decay" factor for spatial interaction, the distance measure is commonly a straight-line distance or average travel time or distance by a single mode along a fixed route or shortest path. The optimization of travel routes or transportation facility usage levels do not have a high level of importance in the modeling process. Transportation modelers predict movement patterns and economic modelers predict levels of economic activity. The convergence of these two efforts could produce a valuable, integrated analytic tool. There are few published examples of GIS applications for regional economic analysis that consider transportation infrastructure. The following is a brief summary of three examples.

Brooks, London, Henry, and Singletary (8) employ GIS to analyze the impacts of infrastructure investments on employment and income distribution. In the case of transportation investments, the GIS is used to calculate highway density measures for each of the Census County Divisions (CCDs) in the state of South Carolina. Their results suggest that highway accessibility has a significant impact on employment levels. An input-output (IO) table was then used to estimate employment impacts related to output and income effects. The resulting model can then be used to simulate the impacts of proposed highway improvements on employment and industrial output.

In 1994, Hartgen and Li (9) reported about the use of GIS for transportation corridor analysis in a 10-county rural area of North Carolina. Their research estimated the growth impacts of interstate exits following improvements to the roadway. They also analyzed the impacts that resulted from decreasing travel times from manufacturers to shipping ports and also for changes in commuter sheds because of increased accessibility. Using the GIS they were able to generate forecasts of travel volumes which then impacted assignments. Their analysis, however, did not consider multiple mode choice opportunities.

In a third article, Nyerges (10) provides a thorough description of the transportation modeling process that accounts for region-wide population, employment, and household forecasts within a multimodal framework. He describes GIS support for travel demand forecasting in the Puget Sound Region of Washington. As is common to the demand forecasting process, the assignments and mode choice are iterative, however, the economic impacts to the region have no explicit feedback function to the regional demand element. Exogenous forecasts for employment and residential growth are supplied at the traffic analysis zone level.

Overall, the use of GIS in regional economic modeling efforts is primarily for data generation and integration purposes. The complexity of economic impact assessment related to transportation investments has been recognized by researchers (11). Most of the actual forecasting is performed external to the GIS with traditional modeling operations such as IO analysis. The missing element here is the true integration of transportation modeling with regional impact assessment, which then feeds back into transportation demand. The pieces of this process currently exist, however, the methods for integration have not yet been fully developed.

Sub-models and feedback

It is likely that a model analyzing regional demographic, economic activity, and a multimodal transportation network will be comprised of several sub-models. Estimates of zonal population change, industrial output and capital investment will be used in order to predict impacts of transportation system alternatives. The issue of simultaneous interactions has been only minimally addressed in the transportation and regional economic modeling literature. Economic development generally occurs in anticipation of infrastructure investments or as a function of the benefits of existing facilities (12). One relies upon the other - where a transportation facility in itself cannot create economic activity without the existence of other inputs. Not only are there interactions between transportation investments and population and economic activities, but also among demand characteristics. Population changes are effected by industrial locations. Industrial locations depend upon labor availability. Industrial location activities rely on the availability of other industries in order to achieve agglomeration thresholds, and so on. Each of these relationships becomes a sub-model within the overall regional investment analysis structure.

In addition to sub-models, feedback becomes an important issue in the modeling process. Similar to the way that traditional UTPS operate in an iterative fashion, incorporating changes in congestion levels to affect traffic assignment to network links (13), a regional transportation investment model needs to incorporate feedback from changes in economic activities that produce transportation demand (14); see Figures 1 and 2. Impacts of transportation investments can typically occur over many years and produce changes in business investments and residential development patterns. These changes then impact transportation network utilization by increasing or decreasing spatial demand levels. Other research has focused on analysis zone definition as also having an impact on feedback opportunities (15).

Figure 1. Components of macroeconomic transport systems simulation (image unavailable)

Source: Kresge and Roberts (1971).

Figure 2. Functional components of the macroeconomic model (image unavailable)

Source: Kresge and Roberts (1971)

Sub-models (or the full model) could be made available at the point of data entry so that users (data inputs) could test their data or generate estimations or projections based upon their current data. Each user controls his own layer of data but can also view and manipulate other layers while not modifying them. New data providers (users) would be added as a new layer of information. In order to establish a new user (layer), criteria will need to be structured regarding (a) geography, (b) time series/interval, (c) meta data, and (d) type, such as economic, environmental, network, and social.

Advantages of GIS

Researchers have identified many advantages of using GIS for transportation modeling. The primary advantages include speed, analytical capabilities, visual power, efficiency of data storage, integration of spatial databases, and capabilities for "finer-grained" spatial analysis (16; 17; 18). By its nature, geographic information is rarely beneficial to only a single user or location. Typically geographic attributes are common to region-wide locations. Initial start-up investments in GIS usually involve large investments in base map layers of geographical data. For example, cities will often want countywide data because planning activities usually account for extra-jurisdictional areas to accommodate growth. Environmental data is typically collected and maintained by a state or regional organization, transportation facility data is handled by state, county, and/or local agencies, business data may be available locally. It is not unusual for these different types of data to be collected and reassembled by individual users. This may be a function of different data needs related to accuracy, software compatibility, and geographic resolution among organizations. A GIS can serve to integrate all of these data types from different data sources (19).

It is not unusual for users to be unaware of available data that meet their operational requirements. Better communications, coordinated data collection efforts, and information exchange can in the long run lead to cost savings and better decision-making (20). Dueker and Vrana (21) generally refer to these as efficiency, effectiveness, and enterprise benefits. Agency efficiency and effectiveness benefits are most commonly discussed in the literature. The third type of benefits, enterprise benefits, take the form of overall information management activities within an organization. An example of interagency cooperation that can produce enterprise benefits is the case of the Pennsylvania Department of Transportation (PennDOT). The process that PennDOT used in constructing their GIS system included the input and from the state departments of agriculture, commerce, community offices, environmental resources, state data center, state library, and governor’s office (22). Such a comprehensive approach in the initial phases of database construction anticipates future data integration and sharing opportunities, as well as providing the collective experience to establish a durable GIS system. By having access to an increased amount of information, individual organizations can enhance their own data resources. Spatial data when combined or overlaid can result in a synergistic effect - the combination of layers is more valuable than the sum of the individual layers (23). This type of data enrichment is another benefit that can be realized by organizations that share data.

GIS tools - accessibility, gravity models, and spatial interaction

The movement of people, goods, services, resources, and information all happen within identifiable network systems (24). The measurement of accessibility can take on a variety of operational forms - based upon assumptions of the attraction between origins and destinations (gravity) and ease of movement through the network. One of the most fundamental forms is referred to as "relative accessibility", where the distance or cost that separates two locations is an indicator for the potential of interaction (25). The distance to CBD measure is an example of this. If a set of points or locations are all potential origins or destinations, an "integral accessibility" index measures the degree of interconnection of a location i to all other locations, j :

Ai = , where Ai = integral accessibility and aij = relative accessibility (26). The point with the lowest accessibility index (shortest overall distance to all other points) is most accessible and also most central (27).

Relative and integral accessibility does not explicitly account for variable supply and demand characteristics within a network of travel origins and destinations. In general, there are few situations where trip origins are unlimited from a location and few destinations have unlimited capacities as trip ends. Gravity models are examples of accessibility measures that are able to account for attraction, opportunity, or capacity among points as well as distance and/or cost of travel (28). For instance; Iij = k Xi Yj f(cij) ,where interaction Iij between locations i and j is assumed to depend on conditions at i and j as well as on interaction costs cij, Xi is a measure of the propensity of i to generate interaction and Yj is a measure of the propensity of j to attract interaction. Xi and Yj can represent production-attraction constraints, which are most frequently used in transportation planning models. Singly or doubly constrained models can be dynamic, reflecting changing supply or demand conditions of locations over time (29). Constrained models balance trip productions and attractions so that total zonal outflows and inflows are equal.

Each of these network analysis methods is available within a GIS in the form of shortest path, tours, routes, allocation, and flow operations (30). Attributes of the network as well as the distance or travel time between each location influence the level of potential interaction between two or more locations. These measurements are easily obtained in GIS packages such as ArcInfo, MGE, and TransCad, which have network analysis capabilities.

The potential of GIS in multimodal investment analysis

Distributed data input and maintenance

With the increasing spread of Internet access and utilization, data traditionally collected and disseminated from centralized locations (such as transportation network, census data, REIS, county business patterns, and TAZ data) can instead be entered by separate agencies as inputs to the database. The model would then use the most recent data available (if desired) from these sources. The database would be designed to handle the level of data aggregation for each entity including time series data (at appropriate intervals). A graphical (web-based) approach would allow for spatially indexed data entry as a user interface (map-based rather than non-graphical database). Another type of data that could be captured automatically would be network volume information, such as congestion levels, queuing times, accidents, transfer delays, and road (link) conditions. This data could also be collected and maintained at appropriate time increments based upon modeling requirements.

General database requirements

To adequately model the various interactions between transportation activities - both on the supply and demand sides - a

Table 1. Examples of database information needs

Submodel Database Fields Spatial




Feature Type
Passenger Demand Trip production Household income Tract, city, county Current Area
Car ownership --- --- ---
Household size --- --- ---
Occupation --- --- ---
Population density --- --- ---
Trip attraction Business type: Parcel, tract Current Area
Industrial --- --- ---
Commercial --- --- ---
Services --- --- ---
Employment density --- --- ---
Ag. Transport Demand Production estimates Crop production: County Annual Area
Soybean --- --- ---
Oat --- --- ---
State livestock Livestock production: State Quarterly Area
Cattle --- --- ---
Poultry --- --- ---
Pork --- --- ---
County livestock Livestock production: County 5-year Area
Cattle --- --- ---
Poultry --- --- ---
Pork --- --- ---
Transfer points Grain elevators Capacity Site Annual Point
Services available --- --- ---
Turnover rate --- --- ---
Costs --- --- ---
Processors Process type Capacity Site Annual Point
Services available --- --- ---
Turnover rate --- --- ---
Costs --- --- ---
Production inputs Supplier type Capacity Site Annual Point
Services available --- --- ---
Turnover rate --- --- ---
Costs --- --- ---
Manufacturing Demand County Business Employment County Annual Area
Patterns Payroll --- --- ---
Production --- --- ---
State-to-state Destinations State 5-Year Area
commodity flows Origins --- --- ---
Value of shipments --- --- ---
Commodity class --- --- ---


The concept of distributed processing has inherent implications for inter- as well as intra-organizational data management. Several issues arise when considering a framework for data exchange among organizations. These issues tend to revolve around security, propriety, cost recovery, and maintenance (31). The reluctance to distributed processing is also a function of the lack of experience that organizations have operating in such an environment. This is not surprising considering that network technologies, especially for desktop computing, have only been widely implemented in the past ten years. On the other hand, Internet connectivity has been experiencing exponential growth over just the past five years. Distributed processing in the form of relational database management and open database connectivity (ODBC), coupled with graphical internet access (WWW) and web-based application programming (JAVA) is opening up vast possibilities as well as challenges to organizational data management.

Internet-based GIS

The Internet has become an efficient means to share geographic data. The World Wide Web (WWW) with graphic capabilities, along with file transfer (i.e., FTP), and web application programming (such as JAVA and ActiveX controls) are bringing GIS functionality to the Internet (32). The current limitations of internet GIS are that: a) HTML documents use graphic images (such as gif and jpeg) and are not object-based with topologic data structures, b) maps can only be associated with other text and images through hypertext links, and c) users cannot edit map contents but only manipulate the view of the map (zooming, panning, etc.) (32). Basic operations like geographic queries, buffering, and overlay analysis techniques are not widely available on internet GIS, however, given the current rate of innovation in web-based GIS, it is likely that true interactivity will be available in the very near future.

Examples of Internet mapping are currently being produced with ESRI software. Objects (locations) can be queried by selected parameters. Maps can be panned and zoomed to display desired locations. An example mapping web site can be seen at: (City of Ontario, CA) and (U.S. Bureau of the Census).


Geographic aggregation

The concept of multiple contributors to a common textual database is not technologically challenging. On the other hand, a shared geographic database with data being entered and maintained with different spatial units from different organizations (such as zip codes for business activity, counties for agricultural productivity, census tracts for population and housing data, and municipal boundaries for property taxation rates) calls for either standardization of spatial units or a spatial structure that relates incongruous data reporting boundaries. In addition, point data will also need to be appropriately aggregated for location specific activities.

Preserving organizationally based boundaries for data reporting and analysis means that algorithms for relating layers of polygon data with non-common boundaries will rely on buffering, data aggregation and disaggregation routines. An example is relating population changes for a census tract and changes in business activity for a zip code to a node in the transportation network. It is likely that the zip code contains many census tracts, thus a share of business activity will need to be allocated to a census tract because it only represents a portion of the total zip code area (see Figure 3).

Figure 3. Layers of a GIS database for transportation planning (image unavailable)

Source: Adapted from Prastacos (1991)


In this case there are the following alternatives:

1. disaggregate to tract level on a proportional area basis

2. assign data directly from zip code to tract

3. aggregate tract data to zip code and use the larger area for data analysis

4. overlay a uniform grid over spatial units and disaggregate to these units for analysis purposes

The research on constructing traffic analysis zone boundaries addresses many of these issues. Typically, different types of geographic data from different data sources will have non-corresponding levels of aggregation. The aggregation or disagregation techniques mentioned above can be used to achieve common areal units from different geographic data types.

Time series/intervals

Similar to spatial aggregation issues, it is also important to consider that a shared GIS database will involve different levels of "temporal aggregation." Currently, model inputs such as population, business activity, and agricultural production are not reported only at different spatial units, but also at different time intervals. Decennial releases of population and housing characteristics do not correspond with the census of agriculture data. Highway trip count data is not always available on an annual basis to correspond with other annually released data. With the inclusion of real time congestion and transport activity data in a distributed database structure, common time increments for data storage or modeling purposes will need to be defined. Just as with spatially incongruous data, aggregation and disaggregation methods can be used to produce comparable temporal units such as months, quarters, and years.

Insufficient network topology

In transportation modeling, networks tend to be simplified in order to reduce the amount of data that needs to be maintained (33). Generalizing a transportation network usually results in the representation of major arterials and highways, which carry the greatest proportion of daily trips. For inter-city transportation modeling of a single mode this is not problematic. A highway network has a relatively simple topologic structure. However, when multiple modes are being analyzed, especially when transfers or connections are not possible at certain intersection points, more data must be captured to account for these characteristics. Depending on the level of detail required by a modeling effort, a realistic transportation network will also need to represent turn (transfer) inhibitions and penalties, bridges, underpasses, overpasses, one-way restrictions, and weight (vehicle type) restrictions (34). Figure 4 gives an example of some multimodal network topology issues in an urban context.

Figure 4. Transit network connectivity (image unavailable)

Source: Sutton (1996).


Relationship of data layers

One capability that is currently lacking in the GIS/transportation modeling environment is the dynamic link between layers of spatial data. There is no interaction between layers of transport facility (links or nodes) or demand (people or freight) data on an automated or dynamic basis. Demand characteristics have to be mechanically transferred to nodes before analyses can be performed. Likewise, projected trip activities from a node are not converted into areal impacts for population and employment changes. This problem is referred to as "cross-layer referencing" - not only between nodes (points) and polygons, but also links (lines) and polygons, or polygons and polygons (35). For instance, a change in population levels for a region may be a function of employment in a region - one data layer should interact with the other to show this relationship. This is typically done by using centroids, however, there is still the missing dynamic link between centroids on separate layers. This dynamic link could be established if attribute data (fields) could actually be a spatial analysis function such as a buffer with aggregation. A dynamic link between objects should also be established based upon specified distance criteria. In any event, real time, a continually updated spatial relationship would aid the modeling process.

Legal, economic, access issues for data sharing

Along with the exchange and sharing of geographic data a variety of legal, economic, and access issues arise. These issues act as hurdles to public and private organizations in the process of coordinating data collection and management activities. Without universally accepted accuracy standards, businesses or public agencies may be liable for distributing inaccurate data, which leads to safety hazards or economic loss. Unreliable geographic data used for technical purposes (e.g. navigation and civil engineering) may lead to hazardous situations.

Economic considerations almost certainly arise in relation to data sharing activities (36). Geographic data collection and management are expensive to support. How should costs be shared? Should the original provider of the data absorb all of the costs, assuming that he will be the beneficiary of other data collection efforts for which they did not pay for? In a data-sharing environment, how can cost sharing be structured equitably? Because there are profit-making opportunities connected to the use of geographic data, the potential for ‘free-riders’ is high. This also relates to the issue of data access. How will data access be controlled among cooperating organizations? Will there be a hierarchy of access levels based upon pricing or confidentiality? Controlling data access becomes a very complicated issue considering the potential for improprieties of data handling and dissemination.

More specifically there are potential problems with allowing public access to government databases. When electronic data is made available to the public there is the potential for accidental for deliberate destruction, loss, alteration, or degradation of data (37). In addition, the release of "secret" information regarding defense and toxic chemical locations is an example of sensitive locational data. On the other hand, the release of previously unavailable government geographic data may serve to increase the publics awareness of government sponsored activities that the public may be critical of. At this point in time the issues are primarily institutional. Technologies for data access and dissemination already exist. Large amounts of data collected by the federal government are currently being provided in CD-ROM and Internet downloadable format. Transportation, demographic, economic, and environmental data is readily available to public users.


The application of GIS to multimodal investment analysis stands to make significant improvements in the areas of impact assessment and modeling. In addition to the graphic nature of GIS, the database capabilities along with spatial analysis tools provide a useful platform for regional economic modeling. The potential for distributed database management activities can result in a more functional, flexible, and accessible data structure - but at the same time it presents a number of challenges to modelers to insure data integrity. The benefits of a GIS-based investment analysis system extend beyond the capabilities and analytical functions of GIS. Organizational benefits related to more efficient and effective data acquisition and maintenance can represent substantial savings and improvements to data quality.



  1. Loureiro, Carlos Felipe G., and Bruce Ralston. "Investment Selection Model for Multicommodity Multimodal Transportation Networks." Transportation Research Record 1522, 1996, pp. 38-46.
  2. Lewis, Simon; Peter Cook; and Marcelo Minc. "Comprehensive Transportation Models: Past, Present and Future." Transportation Quarterly, 44(2), 1990; pp. 249-266.
  3. Miller, Harvey J., and John D. Storm. "Geographic Information System Design for Network Equilibrium-Based Travel Demand Models." Transportation Research-Part C, 4(6) 1996, pp. 373-389.
  4. Ralston, Bruce A.; George Tharakan; and Cheng Liu. 1994. "A Spatial Decision Support System for Transportation Policy Analysis in Bangladesh." Journal of Transport Geography, 2(2), pp. 101-110.
  5. Stokes, Robert W.; and Giancarlo Marucci. "GIS for Transportation: Current Practices, Problems and Prospects." ITE Journal, 1995, pp. 28-37.
  6. Ding, Chengri. "Impact Analysis of Spatial Data Aggregation on Transportation Forecasted Demand: A GIS Approach." Urban and Regional Information Association, URISA, 1994, pp. 362-375.
  7. Bennion, Wayne M.; and Wende A. O’Neill. "Building Transportation Analysis Zones using Geographic Information System." Transportation Research Record 1429, 1994, pp. 49-56.
  8. Brooks, K.R.; J. B. London; M. S. Henry; and L. A. Singletary. "Analysis and simulation of employment and income impacts of infrastructure investments in a state-wide GIS framework." Computers, Environment, and Urban Systems, 1993, 17, pp. 129-151.
  9. Hartgen, David T.; and Yuanjun Li. "Geographic Information Systems Applications to Transportation Corridor Planning." Transportation Research Record 1429, 1994, pp. 57-66.
  10. Nyerges, Timothy L. "Geographic Information System Support for Urban/Regional Transportation Analysis," in Susan Hanson, ed. The Geography of Urban Transportation, Guilford Press, New York, 1995.
  11. Huddleston, Jack R.; and Prem P. Pangotra. "Regional and Local Economic Impacts of Transportation Investments." Transportation Quarterly, 44(4), 1990, pp. 579-594.
  12. Rephann, Terance J. "Highway Investment and Regional Economic Development: Decision Methods and Empirical Foundations." Urban Studies, 30(2), 1993, pp. 437-450.
  13. Levinson, D., and A. Kumar. "Integrating Feedback into Transportation Planning Model: Structure and Application." Transportation Research Record 1413, 1993, pp. 70-77.
  14. Kresge, David T., and Paul O. Roberts. "The Systems Approach to Transport Planning." in John R. Meyer, ed. Techniques of Transport Planning, The Brookings Institution, Washington, D.C., 1971.
  15. Choi, Keechoo, and Tschangho John Kim. "Integrating Transportation Planning Models with GIS: Issues and Prospects." Journal of Planning Education and Research 13, 1994, pp. 199-207.
  16. Hartgen, David T.; Yuanjun Li; and George Alexiou. "Super-Regional, Very Long Range Transportation Modeling with a Geographic Information System." Transportation Research Record 1413, 1994, pp. 78-87.
  17. Anderson, Larry D. "Applying Geographic Information Systems to Transportation Planning." Transportation Research Record 1305, 1991, pp. 113-117.
  18. Niemeier, Debbie A. and M. Kate Beard. "GIS and Transportation Planning: A Case Study." Computers, Environment, and Urban Systems, 17, 1993, pp. 31-43.
  19. Simkowitz, Howard J. "Integrating Geographic Information System Technology and Transportation Models." Transportation Research Record 1271, 1990, pp. 44-71.
  20. Onsrud, Harlan J., and Gerard Rushton, eds. Sharing Geographic Information, Center for Urban Policy Research, New Brunswick, New Jersey, 1995.
  21. Dueker, Kenneth J., and Ric Vrana. "Systems Integration: A Reason and a Means for Data Sharing," in Harlan J. Onsrud and Gerard Rushton, eds. Sharing Geographic Information, Center for Urban Policy Research, New Brunswick, New Jersey, 1995.
  22. Basile, Ralph E.; Thomas TenEyck; and Anthony J. Pietropola. "Geographic Information System for the Pennsylvania Department of Transportation." Transportation Research Record 1305, 1991, pp. 118-122.
  23. Evans, John, and Joseph Ferreira, Jr. "Sharing Spatial Information in an Imperfect World: Interactions Between Technical and Organizational Issues" in Harlan J. Onsrud and Gerard Rushton, eds. Sharing Geographic Information, Center for Urban Policy Research, New Brunswick, New Jersey, 1995.
  24. ESRI 1992
  25. Pirie, G.H. "Measuring Accessibility: A Review and Proposal." Environment and Planning A, 11, 1979, pp. 299-312.
  26. Ingram, D.R. "The Concept of Accessibility: A Search for an Operational Form." Regional Studies, 5, 1971, pp. 101-107.
  27. Garrison, William L. "Connectivity of the Interstate Highway System," in Brian J.L. Berry and Duane F. Marble, eds. Spatial Analysis, Prentice-Hall, Inc., New Jersey, 1968, pp. 239-249.
  28. Rietveld, Piet. "Employment Effects of Changes in Transportation Infrastructure: Methodological Aspects of the Gravity Model." Papers of the Regional Science Association, 66, 1989, pp. 19-30.
  29. Weibull, J.W. "An Axiomatic Approach to the Measurement of Accessibility." Regional Science and Urban Economics, 6, 1976, pp. 357-379.
  30. Slavin, Howard. "An Integrated, Dynamic Approach to Travel Demand Forecasting." Transportation, 23, 1996, pp. 313-350.
  31. Meredith, Paul H. "Distributed GIS: If Its Time is Now, Why is It Resisted?" in Harlan J. Onsrud and Gerard Rushton, eds. Sharing Geographic Information, Center for Urban Policy Research, New Brunswick, New Jersey, 1995.
  32. Peng, Zhong-Ren. "An Assessment of the Development of Internet GIS." University of Wisconsin, Milwaukee, Dept. of Urban Planning, 1997.
  33. Choi and Kim 1994
  34. Sutton, John C. "Role of Geographic Information Systems in Regional Transportation Planning." Transportation Research Record 1518, 1996, pp. 25-31.
  35. Shaw, Shih-Lung. "GIS for urban travel demand analysis: Requirements and alternatives." Computers, Environment, and Urban Systems, 17, 1993, pp. 15-29.
  36. King, John Leslie. "Problems in Public Access Policy for GIS Databases: An Economic Perspective," in Harlan J. Onsrud and Gerard Rushton, eds. Sharing Geographic Information, Center for Urban Policy Research, New Brunswick, New Jersey, 1995.
  37. Dangermond, Jack. "Public Data Access," in Harlan J. Onsrud and Gerard Rushton, eds. Sharing Geographic Information, Center for Urban Policy Research, New Brunswick, New Jersey, 1995.


  1. Fotheringham, Stewart and Peter Rogerson, eds. 1994. "Spatial Analysis and GIS Taylor & Francis: London.
  2. Hsiao, Chien-Hua, Ching-Teng Lin, and Michael Cassidy. 1994. "Applications of Fuzzy Logic and Neural Networks to Automatically Detect Freeway Traffic Incidents." Journal of Transportation Engineering, 120(5):753-772.
  3. Martinelli, David R. and Hualiang Teng. 1996. "Optimization of Railway Operations Using Neural Networks Transportation Research, 4(1):33-49.
  4. Millhouse, Scott and Matthew Gifford. 1997. "Net Results: Adding Neural Nets to Standard GIS Has Helped the Corps" Civil Engineering, 58-60.
  5. Prastacos, Poulicos. 1991. "Integrating GIS Technology in Urban Transportation Planning and Modeling." Transportation Research Record 1305, 123-130.
  6. Ricci, Robert C. and Walter E. Gazda. 1994. "Transportation Modeling: an essential tool for evaluation of IVHS." IVHS Review, Winter/Spring.
  7. Shmueli, Deborah, Ilan Saloman and Daniel Shefer. 1996. "Neural Network Analysis of Travel Behavior: Evaluating Tools for Prediction." Transportation Research C, 4(3):151-166.
  8. Sutton, John and Terry Bills. 1996. "Using GIS for Regional Transportation Planning in Southern California." Paper presented at the GIS-T Conference in Kansas City.

Multimodal Investment Analysis: Phase 1 Contents

CTRE is an Iowa State University center.

Address: 2711 S. Loop Drive, Suite 4700, Ames, IA 50010-8664

Phone: 515-294-8103
FAX: 515-294-0467

CTRE Communications: Marcia Brink
CTRE Webmaster: Michele Regenold


Iowa State University--Becoming the Best