Case Study:
Tren Urbano

Methodology
Regional Accessibility
Description
Scope
Computation
Area Walkability

Application

Conclusions

References

Case Study:

Tren Urbano

Methodology

Regional Accessibility

Description

To compare the various project alternatives, the authors calculated a measure of accessibility of population to employment opportunities. The accessibility measure selected is a gravity-model-based measure known as the Hansen Model. This is formulated as:

Where A_i is accessibility from zone i to the employment opportunities in the San Juan Metropolitan Region; O_j is opportunities (total employment) in zone j; and C_ij is the travel time for a trip from zone i to j. f(C_ij) is an impedance function, which is adopted from the San Juan Regional Transportation Plan (SJRTP):

where a, b and gare model coefficients, given by SJRTP. C_ij is the travel time from zone i to zone j, and e is the base of the natural logarithms. The constant a is set to one for convenience; since accessibility is a relative measure in this case, doing so will not affect the results.

To determine an overall weighted accessibility score for a set of zones, A, the measure for each zone is multiplied by the population of that zone. This product is then summed across all zones for which accessibility is to be calculated. The result is divided by total population in this set of zones:

Equation

While this measure is unit-free and has no intrinsic meaning, it provides an indication of the level of accessibility of jobs for each zone, with jobs weighted more heavily the closer they are (in terms of travel time) to the zone.

Scope

The research team calculated this accessibility measure for the following groups:

By mode, based on the network travel times for each mode. Three distinct public transit modes exist in San Juan: bus, Publico, and rail (under the Tren Urbano build alternative). These were also combined into an overall public transit accessibility measure by selecting the shortest zone-to-zone travel times of the three modes.
By income group, for five income groups. This was done based on the median income level for each zone. Zones were divided into five groups, depending on which income range they fell into. Then, the weighted accessibility index as shown above was determined for each group of zones.
In addition to absolute accessibility, relative accessibility was also compared among modes and income groups. This was done by dividing the group's accessibility score by the regional average accessibility score.

Computation

The data sets used in the study include:

Origin-destination tables for inter-zonal travel time and travel costs by modes and by types of trips (1990 Base Year and 2010 estimates);
Demographic and socioeconomic information, including total population and median income, from the 1990 Census; and
Employment by TAZ distribution in four sectors - basic, retail, service, and government; and
TAZ geographic boundary files.

C programming was used to compute accessibility scores based on travel model output. These scores were then imported into ArcInfo/ArcView GIS to produce tables and maps.

The entire process of designing the Tren Urbano analysis, preparing data, writing programs, analyzing data, and reporting on results was conducted by a research assistant over a two-year period. The project team estimates that once the data and programs have been prepared, doing the calculations and producing tables and maps could take less than a day for the experienced analyst. In each run of the analysis, however, a significant amount of time is taken up by importing and exporting data to and from the GIS package. The project team believes that these procedures could be automated by someone knowledgeable in programming and GIS. This might take a couple of months of staff time but would save significant effort in the long run, if the analyses are to be repeated.

[TOP]