The page you requested has moved and you've automatically been taken to its new location.
Please update your link or bookmark after closing this notice.
Office of Planning, Environment, & Realty (HEP)
The Orange County Transportation Authority (OCTA) provides transit service in a generally automobile-oriented part of southern California. Consequently, OCTA must plan its services and operations for maximum effectiveness and efficiency. To help with this challenge, OCTA has instituted a geographic information system (GIS) that provides detailed information on the demographic and land use characteristics of all locations throughout its service area. Examples are presented of how OCTA has used its GIS to enhance regional transit service planning. These GIS techniques can also be used for regional-level of analysis of transportation and land use policies.
Techniques highlighted here include:
In addition to describing these techniques and their applications, this case study describes the system, staff, and data resources that OCTA has utilized to develop its GIS.
The Orange County Transportation Authority is the consolidated transportation agency responsible for planning, funding, and operating county-wide transportation services in Orange County, California. Orange County has more than 2.5 million residents in 33 municipalities and surrounding unincorporated areas. The County has more than 1.24 million jobs, many of which attract workers from other counties. Figure 1 is a map of the county.
Figure 1. Orange County, California
Development patterns in Orange County present a challenge to the provision of effective transit services. Much of the county and its transportation system developed after World War II when travel has been dominated by the automobile. During that time, the urban areas within Orange County have grown and spread toward each other. The result is a relatively dispersed land use pattern that is served by an extensive network of freeways and arterial streets.
Within this setting, OCTA manages a regional transit system that is among the largest and fastest-growing in the nation. More than 60,000 individuals per day ride OCTA's 400 buses on 73 local, express and rail feeder routes. Ridership has increased nearly thirteen-fold since 1972, when county-wide bus service was launched with five leased vehicles serving approximately 1,300 daily riders. Figure 2 provides a map of the current fixed route transit system.
Figure 2. OCTA Fixed Route Transit System
Even with the auto-oriented development patterns in the county, mass transit plays a vital role in the transportation system. Demographic data indicate a growing transit-reliant population in the county. The transit-reliant population includes:
In addition to the transit-reliant population, a growing number of people in the county are choosing transit for their commutes to and from work. OCTA's bus system serves all employment concentrations within the county and also provides connections to nearby commuter rail and light rail services. By attracting automobile drivers to transit, OCTA's bus system helps to reduce traffic congestion and improve air quality.
Because of the critical services provided by the transit system and the challenges presented by development patterns in the county, OCTA must manage its resources carefully in the provision of bus services. To accomplish this objective, OCTA must deal with complex questions on transportation behavior, demographics, and land use. This task requires data, spatial analysis techniques, and skilled personnel. OCTA uses a geographic information system - a GIS - as a primary tool in answering these questions.
OCTA initiated its GIS program in 1991 with the acquisition of its first ArcInfo license, first running on a PC and then on a single UNIX Sun Sparc Workstation. Since then, OCTA has expanded the GIS role in transit applications from ad hoc mapping support to analytical tasks that produce information for multiple users in different departments. Information requests emanate not only from the transit planning area within the Planning and Development Division, but also from the Operations Division, External Affairs, and other departments.
OCTA continues to use ESRI products - ArcInfo and ArcView - as its primary GIS software. The GIS unit uses ArcInfo primarily for data management and editing, along with spatial analysis and hardcopy map production. The GIS unit requires powerful hardware to support the processing, storage, and plotting requirements of GIS software. Staff performs computing tasks with Windows NT workstations connected to a Unix server. The GIS unit has a Hewlett Packard DesignJet color plotter for making large maps and several printers for smaller plots and transparencies. The staff uses digitizers to convert hardcopy maps into digital layers.
Various other OCTA staff members use ArcView for querying, viewing, mapping, and small-scale spatial analyses. These staff members run ArcView modules - customized user-friendly versions of Arc View that automate common tasks - that have been distributed throughout the Authority, taking advantage of ArcView implementation in the desktop computer environment. These modules give the users easily used capabilities for querying, viewing, and map plotting. The modules include tools for analyses of bus stops, design of transit routes, and market research analyses. OCTA uses other extensions and modules of ArcInfo and ArcView for specialized analyses, such as network modeling tasks with Network Analyst.
The GIS unit is staffed with a supervisor and two analysts who are trained in GIS programming, spatial analysis, mapping, spatial database design, and database management. The GIS unit also employs interns and consultants as needed. OCTA also has a digital graphics department that often works with the GIS staff on the preparation of reports, web pages, brochures, and posters that communicate information developed by the GIS unit.
The most important, and usually most expensive, element of a GIS is data. Over the years, OCTA has developed and maintained spatial data of three principal types: land-use data, socioeconomic data, and transportation data. OCTA maintains this information in a master database that is expanded regularly with a data dictionary for various GIS project applications. Table 1 lists the key map layers maintained by OCTA. To support the maintenance of more complex data structures, some data, such as the bus stop inventory, are stored in a relational database management system (RDBMS). While most of the map layers have been prepared by OCTA, some of them are acquired from other sources, including public agencies and private vendors that specialize in spatial data products.
Table 1. GIS Data Layers Maintained by OCTA
|Map Layer||Description of Data|
|Existing and Proposed Transit Routes||Contains all bus routes of the current system. A separate database is maintained for a proposed system.|
|Bus Stops Inventory||Contains all bus stops for the current bus system. These bus stops assist in determining transit accessibility.|
|Existing and Proposed Master Plan of Arterial Highways||Includes the arterial street network for Orange County. Attributes include traffic data such as average daily traffic, capacity and, as a result, level of service.|
|Street Centerlines||Contains streets of all types in Orange County including freeways, arterials, and local roads. Address ranges are provided in the attribute table of this layer so that the user can geocode locations by street address.|
|Transportation Projects||Includes transportation projects that are either programmed or planned.|
|Transportation Analysis Zones (TAZ)||The TAZ layer contains geographic zones, which share similar socioeconomic characteristics. The TAZ data is used as input for a transportation forecasting model.|
|Aerial Photos||These photos are digital images that can be shown along other layers. They can be used for verifying land use and defining transportation analysis zones (TAZs).|
|Travel Surveys||Contains various market research surveys with origin-destination and trip generation data. For example, on-board transit surveys, senior travel survey and college student travel survey.|
|1990 Census Information||Contains census data at the tract, block group and block level.|
|Land Use||Contains an inventory of land use throughout the county (conducted by the Southern California Association of Governments, the regional MPO). The resolution of the land use layer is two- to three-acre parcels. Each parcel is coded with one of many land use categories. This layer was developed with the aid of high-resolution aerial photographs.|
|Political Boundaries||Contains district boundaries for local governments.|
|Bikeways||Contains all facilities used for bicycle transportation. This layer is useful since buses are now equipped with bike racks.|
|Park and Ride Facilities||Includes park and ride lots. These lots are heavily used by transit patrons.|
|Geocoded Employer Database||Contains all employment locations. Employers are categorized by industry codes.|
|Major Transit Destinations||Includes all primary locations used by transit patrons. This layer is useful for locating transit routes for the fast growing segments of the transit-reliant population.|
Since its inception at OCTA, the GIS unit has evolved into a valuable technical resource that is utilized for many interdepartmental tasks. Some uses of GIS at OCTA for regional transit service planning include:
GIS also assists with many other aspects of transit service planning, including maintaining an inventory of bus stops; visualizing passenger count data; administering and analyzing on-board surveys by linking survey data with transit route data; displaying survey data; linking passenger count data to transit route data to analyze the productivity of specific lines; and analyzing the travel characteristics of paratransit patrons to determine their potential for fixed-route bus service.
Two examples of OCTA's use of GIS for transit planning are presented below. The first describes how the spatial distribution of socioeconomic data is revealed in more detail by employing spatial overlay techniques. The second example illustrates the analysis of transit accessibility using GIS. Potential extensions to regional transit and transportation planning are also discussed.
The examples presented below are only a few of the many applications of GIS-based analysis for transit planning. An additional application, in which OCTA used GIS to map transit survey data, is described in Transportation Case Studies in GIS (FHWA, 2000).
Socioeconomic data are vital for determining locations for transit routes and facilities. Often, socioeconomic spatial data reside in a zonal layer - census tracts or transportation analysis zones (TAZs), for example - where each zone contains such attributes as population, housing units, and employment. A single zone in such a layer may cover a very large area. While socioeconomic data aggregated to these large zones are useful for regional analyses, they can support only a relatively crude analysis of accessibility to individual bus routes and bus stops. OCTA recognized this problem in analyses using current socioeconomic data at the census tract level. To solve the problem, OCTA used GIS tools with land-use information to disaggregate the census data.
To accomplish the disaggregation, OCTA overlaid the census tract layer shown in Figure 3 onto the layer shown in Figure 4 representing small-area land use. The land use layer has a higher resolution of detail since each feature in the layer is a two- to three-acre area that is encoded with a specific land use class.
Figure 3. Population Density by Census Tract
Figure 4. Orange County Land Use Map
The approach to disaggregation is easily illustrated with a larger-than-normal tract with a moderate-sized population that is concentrated in a few contiguous blocks. Computed at the tract level, the average population density understates the population density - and potential transit market - that exists within a portion of the tract. GIS tools can produce a better representation of the population distribution by tapping information from the land-use layer on actual locations of housing within the tract. The resulting estimate of population locations reveals that most residents of the tract are located in high-density housing in a compact subarea of the tract. Tracts and land use can be combined to determine a more accurate spatial distribution of population within the tract. This is done by dividing the total population of the tract by the land area within the tract devoted to residential use, to obtain a revised estimate of population density. This density is then assigned to the residential area, while non-residential areas are assigned a population density of zero. Figure 5 illustrates the resulting layer that more accurately represents population distribution and density within tracts.
Figure 5. Refined Population Density Map
The same approach can be applied to zone-level employment data as well by using GIS to link it spatially with data on the location and extent of office, commercial, and industrial land uses in each zone. The more precise distributions of population and employment are much more useful to transit planners in efforts to locate bus routes and bus stops to serve areas of high population and employment. Thus, employing the automated spatial analysis utilities of a GIS, OCTA uses multiple map layers to produce a less aggregated and more spatially accurate layer.
GIS can also be used to perform the reverse of this process by aggregating data associated with smaller areas into a larger zone. In general, one of the most useful capabilities of GIS is the manipulation of data that is rich in attributes across different levels of aggregation for use in various analyses.
Pedestrian accessibility is an important factor in the determination of the alignment of a transit route and the locations of bus stops. On-board surveys in Orange County have consistently shown that 80 to 90 percent of bus riders walk to and from their bus stops.
The calculation of accessibility requires the identification of the "catchment area" around each bus stop - that is, the area from which potential riders would be willing to walk to and from the stop. GIS tools can be used to identify catchment areas in several ways. The selection of a particular approach is determined largely by the availability and accuracy of the necessary spatial data. This section presents three ways of determining catchment areas around bus stops.
Route Buffer Catchment Areas - The simplest approach to building transit catchment areas is to create a buffer around an entire route, as depicted in Figure 6. The buffer is centered on the route of interest and is defined by the maximum distance that riders find convenient to walk to and from the system - typically one-quarter mile or so. This method is commonly used to describe fixed-route service areas for federal reporting purposes. The approach implicitly assumes that all locations within the buffer are accessible to the route. However, this assumption may overestimate accessibility because the route is only accessible at bus stops. Nevertheless, this approach can be useful when bus stop locations are not available in the GIS. The approach is commonly used, for example, to examine the general accessibility of alternative future transit routes for which bus stop locations are not yet known.
Catchment Area Created From a One-Quarter-Mile Buffer Surrounding a Transit Route
Stop Buffer Catchment Areas - The second approach to building catchment areas is to create a buffer around each individual stop served by a route. This approach improves the accuracy of resulting measures of accessibility to the route but requires the availability of a bus stop layer that has coordinate locations for each stop. Figure 7 illustrates the results of this approach for the same route used in Figure 6 to illustrate the use of a route-level buffer.
While the stop-level approach clearly improves the accuracy of the analysis, it implicitly assumes that all locations within each buffer are within the defined walking distance of the bus stop. The geometry and connectivity of the street system, however, may result in actual walking distances that are much further that the distance implied by the buffer. Dead-end streets, cul-de-sacs, freeways, and interchanges - all characteristics of auto-dominated development patterns - can cause circuitous walk paths that are not recognized by the buffer approach.
Catchment Area Created From One-Quarter-Mile Buffers Surrounding Transit Stops
Network-Based Catchment Areas - The third approach to building catchment areas is to identify all links in the street network that can be reached from a bus stop by walking along the network for less than the specified maximum walking distance. This approach requires network analysis tools and a spatially accurate street network that identifies facilities that can be used by pedestrians - streets, roads, and pedestrian-ways in contrast to freeways and expressways.
Several methods can be used to exclude facilities not available to pedestrians. These facilities can be removed entirely from the layer that describes the streets-and-highways system. Alternatively, the not-walkable facilities can be excluded from the selected set before any network operation is performed. Finally, these facilities may be assigned artificially long distances that would prevent their inclusion in any shortest path to and from a bus stop.
The addition of pedestrian-only facilities may be important to an accurate portrayal of possible walk paths to and from bus stops. Pathways in residential areas and pedestrian connections in commercial areas can play important roles in providing access to bus stops. Depending on the location and importance of these facilities, their addition to the street network may be important to the accuracy of the results.
Most GIS software has network analysis capabilities that can be used to create network-based catchment areas. The user specifies the origin points (stop locations) and maximum walk distance. The network tools then determine the walk network around each origin point (bus stop) that lies within the user-specified distance. Figure 8a illustrates the results of this analysis for the same route shown in Figures 6 and 7. Figure 8b focuses on three of the stops on that route, contrasting the results from the buffer-based approach and the network-based approach. Because the network-based approach is able to recognize disconnects in the street network, it is able to identify the walk-accessible service area for each bus stop with much better accuracy.
One-Quarter-Mile Walk Networks Created Around Transit Stops
Comparison of One-Quarter-Mile Buffer and One-Quarter-Mile Walk Network at Transit Stops
The methods described above assist in the measurement of accessibility by identifying the street network within walking distance of a transit stop and the total population and employment accessible to a stop. Accessibility data for individual stops can be further compiled to determine accessibility measures for individual routes, for the entire transit system, or at a traffic analysis zone (TAZ) level. Potential route, system, and TAZ-level measures might include:
In developing route-level or system-level accessibility measures, care needs to be taken so that populations in adjacent, overlapping catchment areas are not counted twice. Two overlapping areas may be combined into a single area, or may be split into two separate areas halfway between the stops.
The following example illustrates how regional transit accessibility can be evaluated by plotting the population accessible to OCTA's fixed-route service within a one-quarter-mile walk. This analysis relies on the disaggregation of socioeconomic data as well as the walk network analysis described above. Figure 9 illustrates the population accessible to existing transit service throughout the county. Figure 10 demonstrates the change in accessibility as a result of a proposed change in bus service. The changes as identified from this analysis are shown in Table 2.
Figure 9. Population Accessible to Existing OCTA Bus System
Figure 10. Change in Population Accessible to OCTA Bus System
Table 2. Pedestrian Access for Existing and Proposed Systems
|Existing System||Proposed System||Change|
|Number of Residents with Walk Access||1,149,422||1,197,800||48,378|
|Percent of All Residents||47.7||49.7||2.0|
|Number of Jobs with Walk Access||1.056,791||1,078,609||21,818|
|Percent of All Jobs||74.7||76.3||1.6|
Using other methods, relationships between residential and employment areas can be evaluated to determine different measures of accessibility or mobility. For example, the number of jobs within a 60-minute bus ride of a particular census tract or bus stop can be estimated. This would require estimates of employment accessible from each bus stop, as well as GIS representations of the transit network (bus routes), transfer locations, headways, and travel speeds. Some of these data may be obtained from travel demand models with a transit network. Travel demand models are used to calculate similar measures, but being more generalized, they may not provide results as detailed as GIS-based methods.
OCTA has realized three kinds of benefits from the use of GIS in transit planning: better communication of technical information, more efficient development and maintenance of data sources, and more comprehensive analyses of transit planning questions. To help improve communications between planners and decision-makers on technical matters, OCTA has used GIS to transform information from tabular formats into visual displays. This application is particularly effective because so much of the information - about transit routes, bus stops, and demographic distributions - is naturally tied to specific locations.
To enhance the efficiency of data-related tasks, OCTA has used GIS as a way to organize and connect geographically the data sources maintained by different departments in the agency. Because this connection means that each data source is immediately available to all departments - not just the department that develops and maintains it - each department no longer has to duplicate efforts the efforts of others to maintain and update copies of potentially useful data sources.
Finally, to improve analytical capabilities, OCTA has used GIS and its data-integration tools to expand the breadth and depth of work to answer real-world planning questions. These analyses have previously been either not possible or very cumbersome without the ready capability of GIS to tie together different information based on geographic relationships. As a result, OCTA is now able to uncover and communicate insights from the data that previously have been difficult to detect without GIS.
OCTA's GIS techniques greatly expand the agency's ability to measure accessibility to transit. A few practical limitations should be noted, however, that are relevant to any agency applying similar data and techniques. In particular:
Some potential future applications of the GIS techniques illustrated here include:
OCTA has successfully applied GIS to its transit planning efforts for nearly 10 years. The result is a reduction in the time, effort and funds needed to perform analyses used in the decision-making process. During OCTA's implementation of GIS, the GIS staff has learned these key lessons:
OCTA will continue to use GIS for its transit planning efforts in the future. The agency intends to make spatial data available to in-house, non-GIS users via an intranet. The data will include demographic data, route alignments, and analysis results. Therefore, instead of viewing hardcopy maps, planners will be able to interactively create customized maps, using GIS data, within a web browser. This is one of the future applications that OCTA's GIS unit will work toward as it continues to provide a high-quality service to transit planning and other business functions throughout the agency.
Hsiao, Shirley; Jian Lu, James Sterling, and Matthew Weatherford. Use of Geographic Information System for Analysis of Transit Pedestrian Access. Transportation Research Record 1604 (1997).
|Orange County Transportation Authority||Shirley Hsiao|