MMA Process Fundamentals

The process of developing multimodal accessibility scores is simple in concept but challenging in practice. There are many decisions to make along the way, and processing data for numerous origin-destination pairs can be computationally cumbersome. For this reason, a set of geoprocessing tools for ArcGIS have been developed to guide analysts through the process. The geoprocessing toolbox is documented here. This section provides insight into the major phases and components of an MMA analysis.

The basic procedures for MMA processing are presented in the diagram below:


The process of developing accessibility scores starts by determining the modes to be analyzed and defining scenarios by selecting appropriate land use and network data. Examples of potential data sources are listed in the figure above. Scores are developed for each scenario by using network data to evaluate travel times and summarize the number of activities reachable by each population group. Comparisons across different scenarios can provide useful insight for analyzing a variety of planning issues.

Determine which modes will be analyzed

Depending on the focus of the analysis, you may only need to calculate accessibility for a single travel mode, such as walking or transit. In other cases, complete multimodal analysis may be required. The most commonly evaluated modes are walking, biking, transit (walk access), and auto. The modes selected will determine the data used in the analysis.

Define scenarios

What conditions will you analyze and what comparisons are desired? Scenarios include combinations of land use and network data. Thus, the selection of data sources is critical in scenario definition. Important considerations include the temporal and spatial scope of the analysis, the modes to be analyzed, budget for obtaining vendor data, and availability of open data sources such as GTFS feeds. Each scenario can blend alternative land use and network data. For example, suppose you want to assess future accessibility based on proposed transportation improvements and in light of potential changes in land use. You may choose to define four scenarios as shown below:

Land Use Existing Future
Existing “Base” “Transportation-only”
Future “Land use-only” “Combined”

Comparing the “Transportation-only” and “Land use-only” scenarios to the “Base” scenario provides insight into how much each component (transportation improvements or land development) can be expected to change accessibility over existing conditions. Comparing the “Combined” scenario against the others shows how synergies between transportation and land use interact to enhance accessibility above what can be accomplished through focusing only on transportation or land use.

  • Coming soon: Determinants of data sources

Score scenarios

For a given scenario and for each mode, calculate accessibility scores. The calculation of scores itself is relatively simple, consisting of simple table operations, such as calculating a decay factor in a new column; joining activity data based on destination zone IDs; and summarizing accessibility activity, grouping by origin ID. See the figure below for a diagram illustrating these steps for three zones. Optionally, accessibility scores can be summarized for groups of origins, with averages weighted by population groups (to keep things simple, this is not shown in the diagram). Using the MMA geoprocessing toolbox, these steps are automated.


Once a skim has been created, the calculation of access scores is accomplished through a series of simple table joins and field calculations, as shown above. Results are summarized for each origin, making analyses and comparisons by population group possible.

Accessibility scores, once developed, can be mapped to show heatmaps highlighting the most and least accessible zones in the study area. Comparisons across modes using ratios can also be mapped to show the contours of modal competitiveness within the study area.

Compare scenarios

Once each scenario has been scored, comparisons across scenarios can be made. These comparisons may reveal how combined land use and transportation projects enhance accessibility, as described in the four-scenario example above. They may provide insight into how alternative project configurations or site locations impact accessibility and travel behavior. Or they may produce scores for ranking projects on a case-by-case basis to prioritize investments.

In all cases, comparisons among scenarios can be made for multi-zone areas to understand the average changes in accessibility that would be experienced by different population groups. Ideally, projects will benefit all populations and help connect disadvantaged groups to greater opportunity.