Skip to contentUnited States Department of Transportation - Federal Highway Administration FHWA HomeFeedback
< Policies to Reduce Greenhouse Gas Emissions Associated with Freight Movements | Table of Contents | Achieving Intermodal Interoperability for Freight Movements >

Linking Transportation and Land Use

Mike McKeever, Sacramento Area Council of Governments
Bruce Griesenbeck, Sacramento Area Council of Governments

Introduction

Linking transportation and land use refers to the process of guiding development and expansion of communities with the goal of better coordination of land use and transportation that accommodates pedestrian and bike safety, mobility, enhances public transportation service, improves road network connectivity, and includes a multi-modal approach to transportation. Typically, this is accomplished through supporting land-use development patterns to create a variety of transportation options.

Under increasing pressure from population expansion, development of large tracks of open lands into residential subdivisions or strip-style commercial shops is occurring in communities throughout the United States. At the same time that roads are being widened and new roads are being constructed, the facets of transportation such as bike trails, sidewalks, and other facilities that link activities and users are lagging. The objective of linking transportation and land use is to define and manage this growth of communities in a fashion that balances land use and transportation needs. To achieve this objective, there are a number of important resources that need to be available to planners. The following discusses four priority areas where federal research is needed to support the linkage between transportation and land use.

Improved Data and Modeling Capacity to Support Integrated Scenario Planning

Metro, the Portland (Oregon) regional government, adopted a long-range growth vision in the mid 1990s1. That action spawned a number of regional-scale scenario planning exercises around the country, starting in Utah with Envision Utah, and most recently becoming a statewide Blueprint Planning program sponsored by the California Department of Transportation for all of the regions throughout that large state. The process became popular enough in California that this Fall the state legislature passed and the Governor signed SB375, a landmark law that requires regional planning agencies to integrate climate change, transportation, land use, and housing planning2. There are several initiatives to advocate for inclusion of some of the concepts in SB375 in the new Federal Transportation bill, which is just starting the re-authorization process. FHWA has actively encouraged Metropolitan Planning Organizations (MPOs) to engage in scenario planning activities that integrate transportation, land use, and air quality decisions.

Most of the regional scenario planning initiatives share the following characteristics:

The Metropolitan Planning Region (MPO) is the most cost-effective scale to build that data, and create modeling and analysis tools necessary to adequately support serious planning. The local level is too small and costly, the state level often is too big and unwieldy in the Western states and too small in the Eastern states. While the MPO is the right scale to build a parcel-level geographic information system, forecasting tools, scenario building models (including three-dimensional visualization capability), and travel and air emissions forecasting models, it still takes money and management-level commitment to make it happen. Many of the technical tools and methods should not be different from one region to another. Some standardization would help cut costs, increase the reliability of results, and support good inter-regional planning to address the cross-border impacts.

Longitudinal Household Travel and Activity Survey

Household travel surveys have advanced in recent years to include important information on the activities that people engage in during the course of the day. Additional data and collection procedures are needed to provide a robust dataset of the transportation and land-use characteristics that influence travel choices. The resulting data set will serve as a cost-effective basis to develop a transferable protocol for activity-based travel models throughout the country.

Traditionally, the surveys were concerned only with the number and location of vehicle and transit trips, then walk and bicycle trips were added. These surveys focused on the primary purpose of the trip as discrete decisions made by travelers, then linking trips into tours was examined to begin to understand how people and households decide on a set of activities and their relative importance.

The locations of trips traditionally was concerned only with the Travel Analysis Zone (TAZ) of the trip, then exact addresses and locations were collected as research pointed out that travel decisions for transit and non-motorized trips depended on very small units of distance and time (i.e., feet not miles, and small not large numbers of minutes).

The surveys traditionally did not collect any information about the location of trips other than TAZ, then research concluded that geographic information about each destination is important to understanding why that location was chosen. The land-use and travel-choice characteristics that influence location choices include street pattern, density and mix of surrounding uses, transit and pedestrian system characteristics, and safety and security among other features.

Originally, the surveys were conducted for only 1 or 2 days for each respondent, but household activities are often scheduled across a week or more for some important mandatory and discretionary trips. Only one survey (in the Puget Sound region) has conducted a multi-year longitudinal survey. By tracking changes in travel and activity over years the long-term behavioral patterns can be analyzed and incorporated into models.

Traditionally, no information was collected on the health of the respondents. More recently, the relationship between land use, transportation system, and health has been examined. Health data include the amount of physical activity, especially walking, exposure to air toxics from vehicles, and personal safety.

Conducting surveys has become more difficult for several reasons. People are more wary of surveys because telemarketers and criminals have used the primary recruitment for household surveys, which are telephone calls to the home. Also, cell phones are now a common, if not the predominate, communication device that makes random household selection within the survey area more difficult. With more attention to detailed spatial data have come concerns about privacy. The surveys have had a requirement that the person and household's private information was not released. Now that it is possible to know the exact address of each trip of each person, attaching the person's name and other vital information is more likely.

Revising the protocols for household travel surveys would address each of these issues in meeting its objectives of a comprehensive survey of urban travel behavior; such a study currently is being planned in California. Key features of the California study include:

Transportation Cost and Pricing Research

An improved travel survey could be used in conjunction with activity-based travel models to better understand the long- and short-term impacts of costs in travel choices.

Part 1 – Exogenous Costs: Vehicle acquisition, disposition, and use study. Using the first wave or two of the longitudinal study described above, estimate model of vehicle acquisition, disposition, and use. This would make it possible to capture the full range of operating costs of different vehicle types, plus the intra-household dynamics of who uses what vehicle for what trips. This would make it possible to actually model the true costs of vehicle transportation (rather than single-point averages), and it would net significant data on vehicle activity by type of vehicle for use in emissions modeling.

Part 2 – Pricing Policy: Modifications to travel models to allow for evaluation of pricing (HOT lanes, toll roads, parking pricing, road pricing). Parts 1 and 2 together would provide the first comprehensive treatment of true transportation costs using an activity-based modeling system for household-based travel.

Commercial Vehicle Activity Model Transfer

This would be a unique research effort focused on the transferability of a commercial vehicle/freight activity model from one region to another. The “donor” region would be Calgary, Alberta, Canada. Research would focus on model structural modification and calibration to fit in the Sacramento test region and any other test regions in the country. By implementing this model, SACOG would have the first true activity-based demand simulation model for both household- and commercial-based travel.

Workshop Presentation

Innovations for Tomorrow's Transportation FHWA, January 9, 2009. Mike McKeever, Sacramento Area Council of Governments, Executive Director, mmckeever@sacog.org.

Sacramento Area Council of Governments. Map of area included in the subject line. List items include 6 counties, 22 cities, 2.2 million people.

Evolution of Planning. Bullet list with three items. (1) Metropolitan Transportation Plan (MTP) adopted in 2002 - disappointing performance. (2) Blueprint growth strategy adopted 2004. (3) New MTP adopted 2008 - better performance, indicated by Fewer vehicle miles traveled, Higher non-auto mode shares, Reduced carbon emissions per capita.

How to Best Manage Growth?. Horizontal bar chart showing amount of growth through 2050. The number of people reaches 1.7 million; the number of jobs reaches 1 million; the number of dwellings reaches 840,000.

Smart Growth Principles. Bullet list with seven items. (1) Housing Choices. (2) Transportation Choices. (3) Compact Development. (4) Use Existing Assets. (5) Mix Land Uses. (6) Protect Natural Resources. (7) High Quality Design.

Information-driven planning. Three versions of the area map are shown, with various sectors highlighted. (1) Base Case Scenario: (MEPLAN - Land Economics, PICAS). (2) Alternative Scenarios: (I-PLACE begin superscript 3 end superscript S). (3) Regional Transportation: (SACMET/4Ds, SACSIM).

Citizen Input - Over 5,000 participants at workshops and forums (PLACE3S modeling). Four photographs shows people participating in meetings at a large table covered with documents and maps.

Regional Forum 2004 - 1400 people. One large photo shows many people at tables in a conference area with projection screens and a display area at the far end.

Elected Officials Summit - Oct. 2004. Large photo shows a meeting of people in a room with tiered seating in a u-configuration, and two projection screens and a dais in the open end.

Base Case Urban Footprint - 2050. A large map shows areas of existing (year 2001) and planned (year 2050) development and boundaries.

Preferred Scenario Urban Footprint - 2050. A large map shows areas of existing (year 2001) and planned (year 2050) development and boundaries.

Less Urban Land. A vertical bar chart shows additional urbanized land through 2050 in square miles. The value for base case scenario reaches 661 square miles. The value for preferred blueprint scenario reaches 304 square miles.

A vertical bar chart shows housing choice in percent, indicating new housing stock for two lot sizes, small and large. The 2050 base case plots 20 percent small, 80 percent large. The 2050 preferred scenario plots 69 percent small, 31 percent large. The 2035 MTP projections plot 62 percent small, 38 percent large. The 2004-06 built units plot 34 percent small, 61 percent large. The 2007 units for sale or under construction plot 67 percent small, 33 percent large.

Cost-Effective Solutions with Good Performance Benefit. Line chart showing daily congested vehicle miles traveled, in person trips per day. The plot for actual vehicle miles traveled starts at about 2000 in 1995 and swings upward to a value above 3000 for 2004. The plot for old MTP swings from this value steadily upward to a value approaching 14,000 by the year 2035. The plot for the new MTP swings less steadily to end at a value approaching 8,000 by the year 2035.

Quality Transit For All. Line chart showing daily transit person trips in person trips per day. The plot for actual trips starts at 75,000 in 1995 and swings upward to a value of about 100,000 for 2006. The plot for the old MTP extends to a value above 150,000 by the year 2035 in an almost linear fashion. The plot for the new MTP swings sharply upward to a value above 300,000 by the year 2035.

12% per household carbon dioxide reduction by 2035. A table matrix shows weekday travel, energy, and carbon dioxide impacts in 2035 for MTP and No Project scenarios. For weekday VMT in thousands, the MTP value is 84,879 and the No Project value is 90,664, with a savings of 5,785. For daily gas and diesel in gallons, the MTP value is 5,053,000 and the No Project value is 5,564,000, with a savings of 511,000. For daily carbon dioxide in tons, the MTP values is 50,200 and the No Project value is 55,280, with a savings of 5,080. Source: SACOG, October 2007.

Senate Bill 375 (Steinberg). The goal of SB 375 is to reduce GHG emissions from cars and light trucks through incentives for better development patterns so people can choose to drive less.

SB 375 Does 4 Things. Bullet list with four items. (1) Adds new Sustainable Communities Strategy to the Regional Transportation Plan - leveraging existing transportation funding incentives to support growth in good locations. (2) Adds new CEQA provisions to assist land use decisions that implement the Sustainable Communities Strategy. (3) Adds new modeling provisions to accurately account for the transportation impacts of land use decisions. (4) Adds a new provision for determining the regional need for housing so that it will be consistent with the Sustainable Communities Strategy.

Percentage Growth Greenhouse Gases, 1990-2004. Vertical bar chart shows percentage growth of greenhouse gases between 1990 and 2004 in five categories. The plot for cars and light trucks extends to 50.2 percent. The plot for agriculture, forestry, etc. extends to 11.6 percent. The plot for industrial processes and products extends to 23.4 percent. The plot for electricity extends to 4.9 percent. The plot for net other extends to 8.9 percent. Source CARB GHG Inventory 2007.

Growth in VMT will overwhelm savings from LCFS and AB 1493. Line chart showing percent values over time, indexed to the year 2005 = 100 percent, for five data sets. The plot for 1990 carbon dioxide emissions is flat from 2005 to 2030 at about 87 percent. The plot for vehicle miles traveled increases steeply to reach a value of 170 percent by 2030. The plot for carbon dioxide emissions swings slowly upward to a value of 110 percent by the year 2030. The plot for fuel with low carbon standard swings slowly downward to a value of 90 percent by 2030. The plot for vehicle standards (AB 1493) swings downward to a value below 75 percent by the year 2030. Plots are based on CEC, CARB, and CALTRANS data.

Travel Demand Models. SACMET is traditional 4-step model indicated as follows: Based on 1533 traffic analysis zones; Households cross-classified (persons, workers, income); TP+ software platform; Developed 1994; Used, improved since then; 3 MTP's adopted (1996, 1999, 2002); 2 FTA New Starts; Basis for many city/county models

Travel Demand Models (continued). SACSIM is Activity-based tour model, indicated as follows: Based on parcel/point land use data; Synthetic population (includes age, gender, FT/PT worker, student status); Custom software (DAYSIM) w/in a TP+ shell; Developed 2006; First application: analysis of 2008 MTP.

Difference in Unit of Analysis. Bullet list with two items. (1) SACMET = TAZ and Trip, indicated by Geographic area with number of people, jobs, etc.; Trips are 'disembodied' and treated as a 'gravitational' event. (2) SACSIM = Persons and Tours, indicated by Population represents variety of people in a 'real' way; Travel is a an outgrowth of activities-a way of stringing activities together.

SACMET Representation of Travel. Bullet list with four items. (1) Trips to/from TAZ, not Households. (2) Trip rates = averages. (3) Mode of trips unrelated. (4) Non-home-based trips disconnected. A graphic for home-based trip and a graphic for non-home based trips illustrates routes from TAZ to work, school and from work to other TAZ.

SACSIM Representation of Travel. Bullet list with five items. (1) People represented (1 FT worker, 1 PT worker, 2 school-age kids). (2) 'Tours' for each person generated, indicated as follows: Tour = chain of trips beginning and ending at home; Tours defined by activities (work, school, shop, meal, etc). (3) Mode prior mode choice affects later mode choice on tour. (4) Non-home-based travel 'attached' to person. (5) Characteristics of people and place of residence retained through all activities and travel.

Typical Weekday Travel for 4-Person Household. Symbolic graphic indicates routes between buildings and areas for two adults and two juveniles. The graphic depicts 16 person trips, 11 vehicle trips, and 45 vehicle miles traveled.

Density and Non-Auto Travel. Vertical bar chart plots percent non-auto travel for four categories of density at place of residence in terms of jobs plus dwellings per acre. For density less than or equal to 4, the plot extends to about 4 percent. For density more than four to 10, the plot extends to about 11 percent. For density more than 10 to 20, the plot extends to about 13 percent. For density greater than 20 the plot extends to about 24 percent. Plots are based on survey data.

Density and Non-Auto Travel. Vertical bar chart plots percent non-auto travel for four categories of density at place of residence in terms of jobs plus dwellings per acre This chart repeats plots from WP4-27 but with SACSIM data added. For density less than or equal to 4, the plot extends to about 5 percent, which is higher than the 4 percent of the survey. For density more than four to 10, the plot extends to about 10 percent which is lower than the 11 percent of the survey. For density more than 10 to 20, the plot extends to about 14 percent, which is higher than the 13 percent of the survey. For density greater than 20 the plot extends to about 27 percent which is higher than the 24 percent of the survey. Plots are based on survey and SACSIM data.

Density and VMT. Vertical bar chart plots vehicle miles traveled per hour for four categories of density at place of residence in terms of jobs plus dwellings per acre. For density less than or equal to 4, the plot extends to about 59. For density more than four to 10, the plot extends to about 38. For density more than 10 to 20, the plot extends to about 26. For density greater than 20 the plot extends to about 20. Plots are based on survey data.

Density and VMT. Vertical bar chart plots vehicle miles traveled per hour for four categories of density at place of residence in terms of jobs plus dwellings per acre. This chart repeats plots from WP4-29 but with SACSIM data added. For density less than or equal to 4, the plot extends to about 60, which is higher than the 59 of the survey. For density more than four to 10, the plot extends to about 39, which is higher than the 38 of the survey. For density more than 10 to 20, the plot extends to about 30, which is higher than the 26 of the survey. For density greater than 20 the plot extends to just under 20, which is nearly the same as the survey. Plots are based on survey and SACSIM data.

Research Priorities. Bullet list with four items. (1) Improved data and modeling capacity to support integrated scenario planning. (2) Longitudinal household travel and activity survey. (3) Transportation cost and pricing research. (4) Commercial vehicle activity model transfer.


1 2040 Growth Concept, Portland Regional Government Metro Council, 1995.

2 California Senate Bill 375, Signed into California Law September 30, 2008.

< Policies to Reduce Greenhouse Gas Emissions Associated with Freight Movements | Table of Contents | Achieving Intermodal Interoperability for Freight Movements >

FHWA Home | Feedback
FHWA