The Metropolitan Area Planning Council (MAPC) has developed an accessibility-based sketch planning model for quick-hit analysis of travel behaviors, including mode choice, non-motorized travel, average trip length, trip distribution, and ride-hailing service utilization throughout the Boston region. The sketch model responds to the need for reliable planning-level insight into urban travel behaviors that are sensitive to land use, urban design, and travel costs with minimal reliance on complex regional travel models. It utilizes multimodal accessibility as the overarching analysis framework, linking travel behavior to the effectiveness with which different travel modes (walking, driving, public transit, e.g.) connect households, jobs, schools, and shopping.

The sketch accessibility model was developed and piloted as part of MAPC’s analysis of the West Station Area Transit Study. This site provides an overview of the core components of multimodal accessibility analysis, a walkthrough of the modeling steps piloted in the West Station area, examples of model outputs for comparison across hypothetical land use and transportation scenarios, detailed model documents, and links to toolkit support.

The map below shows the West Station study area and model extents. The “focus area” is the local area where land use and network changes are envisioned. The “window area” is a buffer around the focus area approximating the limits for walking and biking trips coming from or going to the focus area. Non-motorized trips are modeled using fine-grained land use (blocks) and network (all streets) data in the focus and window areas. The “remainder area,” all modes are modeled using coarser geographic resolution (regional TAZs and travel newtorks).

What is multimodal accessibility?

Analysis components

Multimodal accessibility is an established transportation analysis metric that is recently generating new insights into travel behavior. It is a relatively simple metric, comprised of a few key components:

  • Zones are geographic aggregations, such as census blocks or travel analysis zones (TAZ’s), that generalize locations where trips come from or go to.

  • Activities are the reasons we travel: jobs, schools, stores, etc.

  • Population groups are defined by demographic characteristics that impact how, why, when, and where people travel.

  • Costs or Impedances refer to the time and money required to travel from each zone to each other zone. Cost information is stored in skims, which are origin-destination matrices recording different cost components for quick summarization to total costs and/or application in cost-sensitive travel analyses.

  • Decay refers to the concept that travel is less likely to occur between two zones if the cost to travel between them is high. Specific rates of decay vary by mode and travel purpose.

These components feature prominently in the summary analysis of existing conditions in the West Station Area.

Basic processing steps

The basic processing steps in developing multimodal accessibility scores include:

  • Estimate travel costs among zones. Costs may include travel time, fares, tolls, fuel costs, parking charges, and other expenses.

  • Calculate generalized cost among zones as a function of each mode’s relevant costs.

  • Apply decay formulas to estimate the “weight” of each OD pair, a rough approximation of travel probability based on generalized travel costs.

  • Broadcast destination-end activities into the decay matrix and multiply by decay weights to get each OD pair’s “weighted activity” estimate.

  • Summarize weighted activity by origin zone to determine the access score.

The last two steps can be run on alternative axes so that access from origin-end activities to each destination zone is summarized rather than access to destination-end activities from each origin zone.

These processing steps are codified in the enhanced multimodal accessibility (emma) python library for generalized application and used in the West Station analysis python library developed for MAPC’s pilot study.

Relationship to travel behavior

Multimodal accessibility offers an intuitive lens to understand the relative impacts of planning alternatives on travel behavior. Access scores are sensitive to:

  • local land use changes;
  • transportation costs and system performance (reflected in travel times);
  • policy variables (parking charges, e.g.).

Access scores, in turn, can be used to assess mode choice and trip distribution by mode. In general, higher accessibility scores by a given mode (especially relative to other modes) indicate higher likelihoods that the mode will be used in daily travel.

Details of the sketch travel model developed for the Boston region using accessibility are provided in the Model Documentation. The model estimates trip generation, mode choice, and trip distribution to assess travel impacts in a localized study area under alternative land use and transportation scenarios.