Final Report
DISCLAIMER
The contents of this report reflect the views of the authors, who are responsible for the facts and the accuracy of the data presented herein. The contents do not necessarily reflect the official views or policies of the State of California or the Federal Highway Administration. This report does not constitute a standard, specification, or regulation.
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies
Final Report
Prepared for the
State of California Business, Transportation and Housing Agency,
California Department of Transportation
By:
DKS Associates
and the
University of California, Irvine
In association with
The University of California, Santa Barbara
And Utah State University
July 27, 2007
Sources of funding for this study:
Federal Highway Administration ( FHWA) State Research and Planning Program,
and the State of California, Department of Transportation,
Division of Research and Innovation
Final Report
Abstract
There is a growing interest in California in “ smart- growth” land- use and transportation strategies designed to provide mobility options and reduce demand on automobile- oriented facilities. This study focuses on models and tools available for use by cities and counties in California for assessing the potential effects of smart- growth strategies.
The majority of regional agencies and local jurisdictions in California currently use a version of the Urban Transportation Modeling System ( UTMS), commonly referred to as the “ four- step travel demand model.” This study provides a review of the steps in the UTMS process to identify where sensitivity to smart- growth strategies may be limited during the modeling process, and suggests ways that improvements could be made.
The greatest degree of modeling smart- growth sensitivity was found among UTMS models used by larger Metropolitan Planning Organizations ( MPOs) or Congestion Management Agencies ( CMAs). Several larger MPOs in California are also implementing new types of models, such as activity- based travel models or integrated land use/ economic/ transportation models. Some local jurisdictions also already use advanced models or travel demand models with high levels of smart- growth sensitivity. The report suggests that if local jurisdictions are already using models with “ moderate” to “ high” levels of smart- growth sensitivity, they should continue to enhance their models.
However, many local jurisdictions’ models have very little sensitivity to smart- growth land use or transportation strategies. In such cases, the study suggests the appropriate use of a planning tool and/ or post- processing application that incorporates “ 4D elasticities” ( e. g., Density, Diversity, Design and Destinations). The report finds that 4D elasticities tools can be used as part of local planning, public participation, and decision- making processes, such as: reviewing major land- use development proposals, preparing updates to city and county general plans and specific area community plans, and during regional “ visioning” and other public participation processes. Therefore, local jurisdictions with low- sensitivity models should consider using a 4Ds methodology to gain increased sensitivity to smart- growth strategies, either applied in “ sketch- planning” software ( such as I- PLACE3S, INDEX), or as a spreadsheet post- processor to a travel demand model.
However, before a decision is made to implement a 4D elasticities tool, the available travel demand model should first be tested to determine its sensitivity to smart- growth strategies. In addition, the report suggests that methods used to capture smart- growth sensitivity ( either via improvements to a travel model and/ or supplemental tools) should first be calibrated with local data and tested for reasonableness before being applied.
The report cautions against using 4D elasticities tools for conducting detailed corridor planning of streets or highways, for transportation impact studies of proposed land- use projects or traffic impact fee programs, or for CEQA or NEPA documentation - unless they are applied in specific ways ( which are described). Other significant findings, conclusions, and recommendations are provided in Chapter 7.
Final Report
ACKNOWLEDGMENTS
The work is this project was performed by a team of researchers from DKS Associates, the University of California at Irvine, the University of California at Santa Barbara, and Utah State University. The members of the research team were as follows:
Research Team
Member
Organization
John Gibb
DKS Associates
Kostas Goulias
University of California, Santa Barbara
Ming Lee
Utah State University; currently University of Alaska at Fairbanks
Miriam Leung
DKS Associates
William Loudon
DKS Associates, Project Manager for the Research Team
Michael Mauch
DKS Associates
Michael McNally
University of California, Irvine, Institute of Transportation Studies
Terry Parker
Caltrans HQ Division of Transportation Planning
Joe Story
DKS Associates
The authors of this report wish to acknowledge the significant contribution of the Terry Parker, the Caltrans Project Manager, who provided vision, direction, and oversight throughout the study.
The authors and Caltrans would also like to acknowledge the participation and contributions of the individuals who served on the study’s Technical Advisory Committee, whose input and review ensure relevance for the readers of the report. The Technical Advisory Committee members were as follows:
Technical Advisory Committee
Member
Organization
Marc Birnbaum
Caltrans Local Development/ Intergovernmental Relations ( LD/ IGR) Program
Jimmy Chen
City of Irvine
Anup Kulkarni
Orange County Transportation Authority
Bill McFarlane
San Diego Association of Governments
Ron Milam
Fehr & Peers
Bruce Griesenbeck
Sacramento Area Council of Governments
George Naylor
Santa Clara Valley Transportation Authority
Jerry Walters
Fehr & Peers
Zhongren Wang
Caltrans HQ Traffic Operations
William Yim
Santa Barbara County Association of Governments
Final Report
The authors also wish to acknowledge the valuable contributions of the individuals who participated in the local jurisdiction case study analyses, and/ or who commented on the draft versions of this report:
Local Agency Staff who provided Case Study Information:
Commenter
Organization Represented
Keith Berthold and Darrell Unruh
City of Fresno
Linda Marabian
City of San Diego
Paul Ma
City of San Jose
Tim Bochum, Kim Murry, and Brian Leveille
City of San Luis Obispo
Caroline Quinn
City of West Sacramento
Other Reviewers and Commenters:
Chuck Purvis
Metropolitan Transportation Commission
Eliot Allen
Criterion Planners, Inc.
Authors of the final report by Chapter:
CHAPTER
AUTHOR( S):
Executive Summary
William Loudon
Chapter 1 – Introduction
William Loudon
Chapter 2 – Overview of Travel Models and their Uses in Local Planning
William Loudon, Joe Story
Chapter 3 – Review of the Conventional Transportation Planning Model: Characteristics, Sensitivity to Smart- Growth Strategies, and Areas for Possible Improvement
Michael McNally, William Loudon, Joe Story, Michael Mauch, John Gibb
Chapter 4 - Overview of New Methods for Reflecting Smart- growth
Ming Lee, William Loudon
Chapter 5 - Travel Modeling Practice in California
William Loudon, Ming Lee, Miriam Leung, Kostas Goulias, Joe Story, Michael Mauch
Chapter 6 - Sensitivity Test of 4D Elasticities
Ming Lee
Chapter 7- Conclusions and Recommendations
William Loudon
Appendices
Michael McNally,
William Loudon
Final Report
TABLE OF CONTENTS
Executive Summary
Overview.............................................................................................................. E- 1
Challenges with Current Travel Modeling Practice............................................. E- 3
Options for Improving Travel Modeling Practice to Gain Smart- growth Sensitivity............................................................................................................ E- 4
New Methods for Gaining Smart- Growth Sensitivity......................................... E- 6
Conclusions and Recommendations.................................................................. E- 10
Chapter 1 – Introduction
1.1 Project Purpose and Objectives............................................................... 1- 1
1.2 Smart- Growth Strategies.......................................................................... 1- 3
1.3 Research Approach.................................................................................. 1- 4
Chapter 2 – Overview of Travel Models and Their Use in Local Planning
2.1 Uses of Models in Local Land- use and Transportation Planning............ 2- 1
2.1.1 Policy Development ( Sketch Planning)....................................... 2- 2
2.1.2 General Plan................................................................................. 2- 3
2.1.3 Specific Plan................................................................................ 2- 3
2.1.4 Transportation Investment Study/ Corridor Study........................ 2- 4
2.1.5 Traffic Impact or Development Fee Program.............................. 2- 4
2.1.6 Traffic Impact Analysis/ CEQA Analysis for New Development2- 5
2.1.7 Transportation Project EIS/ EIR under NEPA/ CEQA.................. 2- 6
2.1.8 Transit New Starts Project Analysis............................................ 2- 6
2.2 Types of Transportation Planning Models............................................... 2- 7
2.2.1 Sketch Planning Tools................................................................. 2- 7
2.2.2 Conventional Models ( 4- Step Models)........................................ 2- 8
2.2.3 Activity- Based Models................................................................ 2- 8
2.2.4 Micro- level Models...................................................................... 2- 9
2.3 The Conventional ( UTMS) Transportation Planning Model................... 2- 9
2.3.1 Limitations of Travel Demand Models...................................... 2- 10
2.4 New Methods for Reflecting Smart- growth.......................................... 2- 10
Chapter 3 – Review of the Conventional Transportation Planning Model: Characteristics, Sensitivity to Smart- Growth Strategies, and Areas for Possible Improvement
3.1 General Characteristics............................................................................ 3- 1
3.2 Representation of the Traveler/ Decision Maker and the Unit of Travel. 3- 4
3.2.1 General Approach........................................................................ 3- 4
3.2.2 Common Limitations and Improvement Options........................ 3- 4
3.3 Representation of Land- uses.................................................................... 3- 6
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page TOC- 1
Final Report
3.3.1 General Approach........................................................................ 3- 6
3.3.2 Common Limitations and Improvement Options........................ 3- 7
3.4 Representation of the Transportation System.......................................... 3- 9
3.4.1 General Approach........................................................................ 3- 9
3.4.2 Common Limitations and Improvement Options...................... 3- 10
3.5 Trip Generation...................................................................................... 3- 12
3.5.1 General Approach...................................................................... 3- 12
3.5.2 Common Limitations and Improvement Options...................... 3- 13
3.6 Trip Distribution.................................................................................... 3- 15
3.6.1 General Approach...................................................................... 3- 15
3.6.2 Common Limitations and Improvement Options...................... 3- 16
3.7 Mode Choice.......................................................................................... 3- 17
3.7.1 General Approach...................................................................... 3- 17
3.7.2 Common Limitations and Improvement Options...................... 3- 18
3.8 Route Choice and Assignment............................................................... 3- 20
3.8.1 General Approach...................................................................... 3- 20
3.8.2 Common Limitations and Improvement Options...................... 3- 21
3.9 Time of Travel....................................................................................... 3- 22
3.9.1 General Approach...................................................................... 3- 22
3.9.2 Common Limitations and Improvement Options...................... 3- 23
3.10 Conclusions............................................................................................ 3- 24
Chapter 4 – Overview of “ 4 D Elasticities” Methods for Analyzing Smart - Growth Strategies
4.1 Introduction.............................................................................................. 4- 1
4.2 The “ 4D Elasticities”............................................................................... 4- 2
4.3 4D Elasticities Post- Processor ................................................................ 4- 5
4.4 I- PLACE3S.............................................................................................. 4- 9
4.5 INDEX................................................................................................... 4- 12
4.6 Another Tool: URBEMIS....................................................................... 4- 16
Chapter 5 – Travel Modeling Practice in California
5.1 Transportation Planning and Modeling Requirements in California....... 5- 1
5.2 Common Practice by Local Jurisdictions................................................ 5- 4
5.3 Application of Smart- Growth Sensitive Methods in California.............. 5- 6
5.3.1 Sophisticated Conventional Planning Models............................. 5- 6
5.3.2 Activity- Based Planning Models................................................. 5- 7
5.3.3 4D Elasticities.............................................................................. 5- 7
5.3.4 I- PLACE3S.................................................................................. 5- 8
5.3.5 INDEX......................................................................................... 5- 8
5.4 Case Studies of Local Travel Modeling Practice .................................... 5- 8
5.4.1 Irvine.......................................................................................... 5- 11
5.4.2 Fresno......................................................................................... 5- 15
5.4.3 San Diego................................................................................... 5- 20
5.4.4 San Jose...................................................................................... 5- 25
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page TOC- 2
Final Report
5.4.5 San Luis Obispo......................................................................... 5- 30
5.4.6 West Sacramento....................................................................... 5- 32
Chapter 6 – Sensitivity Test of 4D Elasticities
6.1 Overview of the Sensitivity Tests............................................................ 6- 1
6.2 Development of the INDEX Sensitivity Tests......................................... 6- 1
6.2.1 Case Study Area........................................................................... 6- 1
6.2.2 Coding of Land- uses.................................................................... 6- 3
6.2.3 Coding of the Transportation Network and Services................... 6- 6
6.2.4 Benchmarking Baseline Conditions............................................. 6- 8
6.2.5 Creation of Development Scenarios............................................ 6- 8
6.2.6 Comparison of Scenarios........................................................... 6- 13
6.2.7 Modification of Development Scenarios................................... 6- 17
6.3 Lessons Learned from the Sensitivity Test............................................ 6- 21
Chapter 7 – Conclusions and Recommendations
7.1 Overview of Study Findings.................................................................... 7- 1
7.2 Study Conclusions................................................................................... 7- 4
7.2.1 Local Model Sensitivity to Smart- Growth Strategies.................. 7- 4
7.2.2 Supplemental Methods................................................................. 7- 4
7.3 Study Recommendations......................................................................... 7- 6
7.3.1 Local Jurisdiction Practice Regarding Local Travel Modeling... 7- 6
7.3.2 Local Jurisdiction Practice Regarding 4D Elasticities Tools....... 7- 7
7.3.3 Research, Development and Training.......................................... 7- 7
Appendices:
Appendix 1: List of Study Participants........................................................................ A1- 1
Appendix 2: Definitions of Acronyms......................................................................... A2- 1
Appendix 3: Glossary of Terms................................................................................... A3- 1
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page TOC- 3
Final Report
LIST OF FIGURES
Executive Summary
Figure E- 1 Logical Progression of Steps to Improve UTMS Sensitivity to Smart- Growth Strategies..................................................................................... E- 5
Chapter 3 – Review of the Conventional Transportation Planning Model: Characteristics, Sensitivity to Smart- Growth Strategies, and Areas for Possible Improvement
Figure 3.1 Logical Progression of Steps to Improve UTMS Sensitivity to Smart- Growth Strategies................................................................................... 3- 25
Chapter 4 – Overview of New Methods for Analyzing Smart - Growth Strategies
Figure 4.1 4D Formulation.................................................................................. 4- 4
Figure 4.3 Support of Community Planning with INDEX................................ 4- 14
Chapter 5 – Travel Modeling Practice in California
Figure 5.1 SJVGRS Model Process................................................................... 5- 18
Figure 5.2 Final 2030 SANDAG Forecast Models............................................ 5- 23
Figure 5.3 West Sacramento Travel Demand Model Structure......................... 5- 33
Chapter 6 – Sensitivity Test of 4D Elasticities
Figure 6.1 Case Study Area Illustration............................................................... 6- 2
Figure 6.2 The Case Study Area within the City of West Sacramento................ 6- 3
Figure 6.3 Land- use Parcels within the Case Study Area.................................... 6- 5
Figure 6.4 West Sacramento Transit, Pedestrian and Bikeway Map................... 6- 7
Figure 6.5 GIS Layers of West Sacramento INDEX Study................................. 6- 7
Figure 6.6 Land- use Parcels and Streets of the Proposed Development........... 6- 10
Figure 6.7 Proposed Points of Interest............................................................... 6- 10
Figure 6.8 Reduced Residential Parcels in Scenario 2...................................... 6- 11
Figure 6.9 Bus Transit Line in Scenario 4......................................................... 6- 12
Figure 6.10 Modified Study Area and the Proposed Development................... 6- 18
Chapter 7 – Conclusions and Recommendations
Figure 7.1 Logical Progression of Steps to Improve UTMS Sensitivity to Smart- Growth Strategies..................................................................................... 7- 3
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page TOC- 4
Final Report
LIST OF TABLES
Executive Summary
Table E- 1 Summary of 4D and UTMS Sensitivity to Smart- Growth Strategies. E- 9
Chapter 1 – Introduction
Table 1.1 Intended Effects from Smart- Growth Strategies on Travel Behavior. 1- 5
Chapter 3 – Review of the Conventional Transportation Planning Model: Characteristics, Sensitivity to Smart- Growth Strategies, and Areas for Possible Improvement
Table 3.1 UTMS Limitations and Areas for Improvement................................ 3- 26
Chapter 4 – Overview of New Methods for Analyzing Smart - Growth Strategies
Table 4.1 4D Elasticities...................................................................................... 4- 4
Table 4.2 Fehr & Peers “ Do’s and Don’ts” for Use of 4D Elasticities................ 4- 7
Table 4.3 Fehr & Peers’ Guidelines for Application of 4D Elasticities.............. 4- 8
Table 4.4 I- PLACE3S Modules and Examples of the Indicators, User- defined Inputs, and Formulas of each Module.................................................... 4- 11
Table 4.5 INDEX Travel Indicators .................................................................. 4- 15
Chapter 5 – Travel Modeling Practice in California
Table 5.1 MPOs in California.............................................................................. 5- 3
Table 5.2 Summary of Six Case Study Cities.................................................... 5- 10
Table 5.3 Comparison between West Sacramento and SACMET Models....... 5- 35
Chapter 6 – Sensitivity Test of 4D Elasticities
Table 6.1 INDEX Land- Use Type and West Sacramento Land- Use Match Up. 6- 4
Table 6.2 Assumption of Residential Population................................................ 6- 5
Table 6.3 INDEX Indicators Selected................................................................. 6- 9
Table 6.4 Proposed New Land- Use Types.......................................................... 6- 9
Table 6.5 Indicator Score Base Case vs. Scenario 1.......................................... 6- 14
Table 6.6 Indicator Scores Scenario 1 to 3........................................................ 6- 15
Table 6.7 Indicator Scores Scenario 4 and 5...................................................... 6- 17
Table 6.8 INDEX Indicator Scores for Modified Scenarios 1 to 3.................... 6- 19
Table 6.9 Indicator Scores Modified Scenario 4 and 5...................................... 6- 20
Chapter 7 – Conclusions and Recommendations
Table 7.1 Summary of 4D and UTMS Sensitivity to Smart- Growth Strategies.. 7- 2
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page TOC- 5
Final Report
( This page intentionally blank)
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page TOC- 6
Final Report
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies
Executive Summary
Overview
There is a growing interest in communities across California and much of the rest of the nation in what is referred to as “ smart- growth” - land development methods that can help reduce the amount of auto travel required to meet the needs of the people who live, work, shop or play in the development. By concentrating new development in existing urban areas where transit services are available or where more urban services are within walking or bicycling distance, smart- growth strategies seek to reduce the amount of automobile travel required by making it possible for more trips to be made by transit, bicycling, or by walking.
Smart- growth has been identified as a priority in Go California, the Mobility Action Plan of the California Transportation Plan 2025, and local communities are encouraged to explore smart- growth strategies in their land- use planning and development approval processes. To support the consideration of smart- growth strategies, the California Department of Transportation ( Caltrans) funded this research to explore whether there are adequate travel- forecasting tools available to local jurisdictions to use in evaluating the potential vehicle trip reducing potential of smart- growth strategies.
The specific objectives of this study were as follows:
• To review the general adequacy of conventional travel demand models used at the local ( city and county) level for sensitivity to smart- growth strategies
• To identify methods or tools that are available for use by cities and counties to add sensitivity for analyzing smart- growth strategies
• To review the current state- of- the- practice in travel- forecasting practice by local jurisdictions in California
• To produce recommendations for travel- forecasting practice to enhance smart- growth sensitivity
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page E- 1
Final Report
• To recommend additional research, development and training activities to improve the state- of- the- practice for travel forecasting for local land- use planning
Although there are different opinions about what constitutes smart- growth, the following principles of a smart- growth community as articulated by the U. S. Environmental Protection Agency ( U. S. EPA) 1 capture the strategies most commonly included:
1. Mix land- uses
2. Take advantage of compact building design
3. Create a range of housing opportunities and choices
4. Create walkable neighborhoods
5. Foster distinctive, attractive communities with a strong sense of place
6. Preserve open space, farmland, natural beauty and critical environmental areas
7. Strengthen and direct development towards existing communities
8. Provide a variety of transportation choices
9. Make development decisions predictable, fair and cost- effective
10. Encourage community and stakeholder collaboration in development decisions
Smart- growth strategies can have an effect on travel behavior in a variety of ways. This study has investigated whether and how travel demand models and other assessment tools that local jurisdictions in California currently use to assess land- use plans and development projects may be “ sensitive” to smart- growth strategies. This report also suggests types of improvements that could be made to the models and assessment tools to improve the evaluation of smart- growth strategies in local land- use planning and development processes.
The research team identified four key intended effects of smart- growth strategies as follows:
Providing opportunities to satisfy travel needs at nearby destinations with shorter vehicle trips, trip chaining, and/ or non- motorized travel
• Clustering of potential non- home destinations such as daycare, cleaners, restaurants, stores, etc. near work sites
• Providing a higher level of diversity in mixed- use clusters
• Developing neighborhoods with more self- sufficient land- uses
• Providing more jobs- housing balance within sub- areas of regions that allows shorter commutes
1 U. S. EPA’s Smart- growth Network, http:// www. epa. gov/ smartgrowth/ about_ sg. htm Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page E- 2
Final Report
• Providing a more complete range of housing options and pricing near employment centers
Using land- use to create trips with origin- destination pairs that are more easily traveled by alternative modes
• Providing higher density residential and work sites near transit
• Providing higher density residential and work sites along bicycle routes and trails
• Location of schools along bicycle routes and trails
• Clustering potential destinations such as daycare, cleaners, restaurants, and stores near work sites and high density residential areas
Providing better and more attractive conditions for travel by alternative modes
• Locating business entrances as close as possible to transit stops or stations
• Locating entrances to higher density residential buildings as close as possible to transit stops or stations
• Providing good pedestrian and bicycle access to transit stops or station
• Providing bicycle storage facilities at transit stops and stations
• Providing bicycle storage facilities at high density residential developments, work places, schools, and shopping areas
• Locating development on a grid street network
• Providing a high level of sidewalk coverage
Providing economic incentives for use of alternative modes
• Providing a limited supply of parking
• Charging separately for parking at multi- family residential, employment and shopping sites
These intended effects were used to develop a framework for assessing the sensitivity of alternative tools for evaluating smart- growth strategies.
Challenges with Current Travel Modeling Practice
A review of the conventional travel- forecasting process used in California and throughout the U. S. identified a variety of limitations in the model systems regarding smart- growth analysis. A majority of local jurisdictions in California use a version of the Urban Transportation Modeling System ( UTMS) - or “ four- step” travel demand model - in its most basic form: a weekday travel model that forecasts only vehicle trips based on fixed vehicle trips rates
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page E- 3
Final Report
by land- use type. Models of this basic type typically cannot reflect changes in mode or vehicle occupancy that can result from smart- growth strategies or the possibility that trips will be made by bicycle, walking, or public transit instead of by automobile. This study’s review of typical UTMS applications identified issues in all areas of current modeling practice that could potentially limit sensitivity to smart- growth strategies. The most significant limitations are:
• Trips not related ( e. g., doesn’t recognize “ trip chaining”)
• Consideration of only vehicle trips
• Limited or no transit modeling capability
• Limited or no modeling of walking and bicycling
• Fixed vehicle trip rates by land- use type
• Development design ( building, street and sidewalk layout) not reflected in traveler choices
• Zonal aggregation of decision- maker characteristics
• Focus on travel during peak- periods
• Travel analysis zones often too large
• Land- use not affected by travel patterns
The time frame in which smart- growth strategies can be implemented or show benefit is also often beyond the ten- or twenty– year time frame of most local plans or models. This makes testing of long- range smart- growth strategies difficult. In addition, the amount of smart- growth development being tested in a model may be small in comparison to the quantity of other existing and future land- uses also represented in the model. As a result, the effects of the smart- growth may be un- noticeable in the aggregate vehicle trip and VMT output of the model.
Because of these and other limitations, it is generally very difficult for a local jurisdiction to adequately evaluate the potential benefits of smart- growth land- use practices regarding transportation efficiency. Therefore, those who may wish to implement smart- growth strategies often have no way to adequately assess or demonstrate the potential for reduced vehicle traffic volumes that may result from smart- growth implementation practices.
Options for Improving Travel Modeling Practice to Gain Smart- Growth Sensitivity
This study has identified numerous options for improving on the basic UTMS practice, and in most cases identified at least one or more agencies in California that are implementing each type of improvement. A summary of these options is presented in Figure E- 1, which illustrates a progression in model improvement practice. Figure E- 1 roughly defines three ranges Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page E- 4
Final Report
of modeling improvement regarding sensitivity to smart- growth strategies: low, moderate, and high. Most of the modeling in the “ moderate- sensitivity” and “ high- sensitivity” ranges is currently done by Metropolitan Planning Organizations ( MPOs) and/ or Congestion Management Agencies ( CMAs) located in the four major metropolitan areas of the state. When local jurisdictions are able to use focused versions of the MPO or CMA model, they also may have medium or high sensitivity. But the most common practice for local jurisdictions in the state is in the “ low- sensitivity” range.
Figure E- 1 Logical Progression of Steps to Improve UTMS Sensitivity to Smart- Growth Strategies
High- Sensitivity ModelsModerate- Sensitivity ModelsLow- Sensitivity ModelsIncome Stratification in Distribution and Mode ChoiceAuto Ownership Modeling Sensitive to Land- Use CharacteristicsDegree of Sensitivity to Smart- Growth StrategiesModeling Mode of Multiple Modes of Access to TransitDistribution Sensitive to Multi- Modal OptionsDisaggregate Simulation of HouseholdsDaily Vehicle Trip ModelSteps to Improve UTMS Sensitivity to Smart- Growth StrategiesTravel Time Feedback Non- Motorized Modes in Mode ChoiceModeling Peak as well as Daily TravelSimple Mode ChoiceTransit Network and Daily AssignmentSupply and Demand EquilibrationIntegrated Land- Use/ Transportation ModelingActivity- and Tour- Based ModelingExplicit Representation of Pedestrian and Bicycle Networks Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page E- 5
Final Report
New Methods for Gaining Smart- growth Sensitivity
Because of the current lack of smart- growth sensitivity in many models, research has been conducted to develop supplemental tools to provide the missing sensitivity. Over the past 15 years, a series of studies have used cross- sectional analyses of variations in travel patterns for zones in major metropolitan areas. 2,3 These research efforts have documented how four key factors influence the rate of vehicle use per capita.
The four key factors4 are often referred to as the “ 4Ds.” They include:
• Density – population and employment per square mile
• Diversity – the ratio of jobs to population
• Design – pedestrian environment variables including street grid density, sidewalk completeness, and route directness
• Destinations – accessibility to other activity concentrations expressed as the mean travel time to all other destinations in the region
Research that resulted in the 4Ds characteristics also produced estimations of “ elasticities” regarding vehicle travel per capita with respect to changes in each of the 4D variables. 5 These elasticities have been used in a variety of application tools to assess the potential vehicle travel reduction benefits of smart- growth land- use strategies.
Two GIS- based programs - INDEX and I- PLACE3S - have incorporated the 4D elasticities and have been used in land- use planning exercises to assess or demonstrate the transportation benefits of alternative smart- growth strategies. The 4D elasticities have also been applied as a “ post- processor” with conventional travel- forecasting models, and also with other sources of “ baseline” travel data ( such as ITE trip generation rates).
2 Robert Cervero: “ Travel Demand and the 3 Ds: Density, Diversity, and Design,” Transportation Research D, 2, 3: 199- 219, 1997; with K. Kockelmann. “ Travel and the Built Environment: A Synthesis,” Transportation Research Record 1780, pp. 87- 113, 2001; with R. Ewing. “ Built Environments and Mode Choice: Toward a Normative Framework,” Transportation Research D, Vol. 7, 2002, pp. 265- 284.
3 INDEX 4D METHOD A Quick- Response Method of Estimating Travel Impacts from Land- Use Changes, Technical Memorandum, October 2001, Prepared for the U. S. Environmental Protection Agency. By Criterion Planners/ Engineers and Fehr & Peers Associates.
4 A 5th “ D,” “ distance from heavy rail transit,” has been developed and applied as a direct ridership model for predicting transit use associated with transit- oriented development. The 5th D is designed to respond to micro- scale influences around transit stations, such as higher density land uses around stations, station access modes, and parking availability.
5 “ Elasticity” is defined as the percentage change in one variable that results from a one percent change in another variable.
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page E- 6
Final Report
In California, I- PLACE3S has been used in the Sacramento area as an integral part of the regional “ Blueprint” transportation and land- use planning effort. The City of Sacramento used the program for land- use planning around a light rail station and to assist in the City’s recent General Plan update. The San Luis Obispo Council of Governments is using I- PLACE3S for regional land- use and transportation visioning and policy development. The San Diego Association of Governments began using I- PLACE3S in 2005 to assess various smart- growth planning options. The program is also being used by the County of Sacramento, Cities of Rancho Cordova and Ventura, as well as in several locations outside California. 6
INDEX has been used by the City of Sacramento for pedestrian planning, by the County of Sacramento for comprehensive land- use/ transportation planning, and by the Sacramento Metropolitan Air Quality Management District ( SMAQD) for analysis of the benefits of alternative urban design strategies for reducing vehicle air pollutant emissions. INDEX has also been used by the Fresno and Madera Councils of Government as part of the San Joaquin Valley Growth Response Study.
The use of the 4D elasticities as a post- processor with a conventional UTMS model has been undertaken in several locations within California, including the following:
• Sacramento Region ( SACOG) – for testing of alternative future land- use and growth scenarios
• San Luis Obispo ( SLOCOG) – for testing of alternative future land- use and growth scenarios
• Contra Costa County ( CCTA) – for long- range visions process “ Shaping Our Future”
• Humboldt County – for County General Plan development
• Fresno and Madera Councils of Government – as part of the San Joaquin Valley Growth Response Study
( Chapter 5 provides additional information about these efforts).
In addition, a 5th D, Distance to Rail Transit, has been used for analysis of transit- oriented land- use designs by the Bay Area Rapid Transit ( BART) and Caltrain rail transit systems that operate in the San Francisco Bay Area. The 5th D is designed to estimate transit use, but does not estimate changes in vehicle trips or VMT.
The application of the 4D elasticities in these locations has demonstrated their usefulness as a planning aid in visioning or long- range planning processes. However, while the use of the 4D elasticities has added “ sensitivity” for analysis of smart- growth strategies, a variety of issues have been identified that may limit the accuracy of the 4D methods, including the following:
6 Per email from Nancy McKeever, California Energy Commission, July 17, 2007.
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page E- 7
Final Report
• They are based on the aggregate characteristics of urban traffic analysis zones, and therefore the elasticities may reflect other unmeasured factors, such as income or cultural groupings that may be correlated with the 4D variables in those areas.
• The 4D elasticities capture some - but not all - of the potential influences of smart- growth strategies.
• Most 4D elasticities tools are not sensitive to the level of transit service or the availability of other “ alternative” travel modes ( such as bicycling) or demand management strategies ( such as parking pricing) that could influence sensitivity of travel to urban design, density, and diversity.
• When used in conjunction with a local travel demand model that already has moderate or high sensitivity to smart- growth strategies, using the 4D elasticities may double- count some of the benefits of the smart- growth strategies, unless the 4D elasticities are calibrated to reflect sensitivity that is already provided by the travel model.
• The 4D elasticities are generally developed for daily vehicle trips and VMT and are not trip- purpose specific. As a result, it is difficult to relate the results to peak- periods of travel. There have been 4D elasticities developed for specific trip purposes, including a set developed for SACOG’s Blueprint project, 7 which improved the capability to estimate changes in peak- period vehicle trips and VMT in that situation. However, most applications of the 4D elasticities have been for daily trips for all purposes.
Table E- 1 provides a summary comparison of how well the potential UTMS improvements and the 4D elasticities are able to address smart- growth travel effects ( that were identified above). This chart illustrates that increased sensitivity to more of the potential effects of smart- growth strategies can be gained through enhancement of UTMS models as compared to applying the 4D elasticities. However, upcoming research on a “ 5th D” ( in another study) will likely increase the capability of the 4D elasticities to estimate benefits associated with a larger variety of transit service. This improvement will likely further increase the capabilities of 4D elasticities methodologies in the near future to estimate travel demand resulting from smart- growth strategies.
7 Don Hubbard and Gerald Walters, Fehr & Peers, “ Making Travel Models Sensitive to Smart- growth Characteristics,” prepared for the ITE District 6 Conference, Honolulu, HI. July 2006.
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page E- 8
Final Report
Table E- 1 Summary of 4D and UTMS Sensitivity to Smart- Growth Strategies
Potential Options to Address UTMS Deficiencies4D Sensitivity11.1Clustering of potential non- home destinations such as daycare, cleaners, restaurants, stores, etc. near work sitesSmall Zones, More Purposes, Non- motorized Modes, Tour- based ModelingDensity, Diversity 1.2Providing a higher level of diversity in mixed- use clustersSmall Zones, More Purposes, Non- motorized ModesDensity, Diversity 1.3Developing neighborhoods with more self- sufficient land usesSmall Zones, More Purposes, Non- motorized ModesDensity, Diversity 1.4Providing more jobs- housing balance within sub- areas of regions that allows shorter commutesSmall Zones, Feedback to DistributionDiversity, Destination1.5Providing a more complete range of housing options and pricing near employment centersIncome Stratification in DistributionDestination22.1Providing higher density residential and work sites near transitSmall Zones, Transit Modeling, Transit Access ModelingDestination, Distance to a heavy rail station ( not applicable for buses, and light rails) 2.2Providing higher density residential and work sites along bike routes and trailsSmall Zones, Non- motorized Modes2.3Location of schools along bicycle routes and trailsSmall Zones, Non- motorized Modes2.4Clustering potential destinations such as daycare, cleaners, restaurants, stores near work sites and high density residential areasSmall Zones, More Purposes, Non- motorized Modes33.1Locating business entrances as close as possible to transit stops or stationsSmall Zones, Transit Modeling, Transit Access ModelingDistance to a heavy rail station ( not applicable for buses, and light rails) 3.2Locating entrances to higher density residential buildings as close as possible to transit stops or stationsSmall Zones, Transit Modeling, Transit Access ModelingDistance to a heavy rail station ( not applicable for buses, and light rails) 3.3Providing good pedestrian and bicycle access to transit stops or stationSmall Zones, Transit Modeling, Transit Access ModelingDesign3.4Providing bicycle storage facilities at transit stops and stations3.5Providing bicycle storage facilities at high density residential developments, work places, schools, and shopping areas3.6Locating development on a grid street networkSmall Zones, More Purposes, Non- motorized ModesDesign3.7Providing a high level of sidewalk coverageSmall Zones, More Purposes, Non- motorized ModesDesign4Provide economic incentives for use of alternative modes4.1Providing a limited supply of parkingAuto Ownership, Parking Constraint, Multimodal, Non- motorized Modes4.2Charging separately for parking at multi- family residential, employment and shopping sitesIncorporate Price in all Steps, Auto OwnershipProviding opportunities to satisfy travel needs at nearby destinations with shorter vehicle trips, trip chaining or non- motorized travelUsing land use to create trips with origin- destination pairs that are more easily traveled by alternative modesProviding better and more attractive conditions for travel by alternative modesSmart Growth Effect
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page E- 9
Final Report
Conclusions and Recommendations
This study has led to a set of findings that can help guide choices of tools for analyzing smart- growth strategies by local jurisdictions ( the cities and county agencies that are responsible for making local land- use decisions), and focus additional research and development activities to improve the tools currently available. The findings include conclusions in two areas:
• Local Model Sensitivity to Smart- Growth Strategies
• Supplemental Methods
Study recommendations are provided in three areas:
• Local Jurisdiction Practice Regarding Local Travel Modeling
• Local Jurisdiction Practice Regarding 4D Elasticity Tools
• Research, Development, and Training
The conclusions and recommendations are products of a cooperative effort by the research team and several participants in the study’s Technical Advisory Committee.
Conclusions about Local Model Sensitivity to Smart- Growth Strategies
1. Few local jurisdictions in California use models that have sensitivity to smart- growth strategies. Most jurisdictions use models that: ( a) lack the capability to estimate transit use or carpooling; ( b) do not include representation of walking or bicycling trips; and/ or ( c) do not allow for variation in vehicle trip rates based on land- use density, mix, or design.
2. Local jurisdictions using Metropolitan Planning Organization ( MPO) or Congestion Management Agency ( CMA) travel demand models that have “ moderate- to high- sensitivity” ( Figure E- 1) can capture some of the smart- growth sensitivity delineated in Table E- 1, but to what degree is not clear.
3. GIS systems for local jurisdiction land- use and transportation system characteristics are making it possible to bring more information into the UTMS modeling process, and that has the potential to increase smart- growth sensitivity. This includes parcel- level land- uses and GIS layers for street systems, bicycle routes, sidewalks, topography, environmentally sensitive areas, etc. GIS systems are also facilitating the application of supplemental methods such as I- PLACE3S and INDEX.
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page E- 10
Final Report
Conclusions about Supplemental Methods
1. Local jurisdictions with low- sensitivity travel models ( Figure E- 1) can benefit from applying a 4D elasticities post- processor either as a spreadsheet supplement to the local model or applied in sketch- planning software, such as INDEX or I- PLACE3S, if used appropriately. It is also possible to integrate the 4Ds within the local jurisdiction model, but this effort requires more effort and should include calibration to local conditions.
2. For the 4D elasticities to function properly, it is necessary to follow the guidelines developed for their use ( Chapter 4), and to calibrate them to local conditions.
3. The 4D elasticities are able to capture some - but not all - smart- growth sensitivity.
4. When the 4D elasticities are applied in conjunction with a travel model that already has “ moderate” or “ high” sensitivity to smart- growth, there may be double- counting of the smart- growth benefits -- unless the 4D elasticities are adjusted to reflect the local model’s sensitivity. Therefore, it is recommended that the “ moderate” or “ high” model be tested to determine its actual degree of sensitivity, and that the 4D elasticities be calibrated, based on local data, to account only for the sensitivity unaccounted for in the travel model.
5. The 4D elasticities ( or any “ correction factors” that are based on aggregate cross- sectional data) most likely capture some unknown trip or VMT reduction effects as a result of correlations between smart- growth variables of interest ( e. g., the 4Ds) and other factors not listed in the formula but related to how an area is developed. These factors may include:
• Income
• Race and cultural characteristics
• Complementary land- uses
• Quality and frequency of transit service
• Parking costs and availability
• Auto ownership
However, developing locally estimated 4D elasticities can be done in a manner that controls for many of these variables. Doing so allows the 4D adjustments to predict trip reducing effects of smart- growth independent of, for example, income and race.
6. The 4D elasticities estimate reduced VT and VMT assumed to result from the use of transit, walking, or bicycling, with the assumption that basic transit and bicycling facilities are available. The 4D adjustments directly account for the presence or absence of sidewalks and pedestrian route connectivity, but do not explicitly account for bicycling facilities or bus or rail service. 8 If the study area
8 While the 4Ds do not account for the presence of rail transit, if the smart- growth study area is expected to offer rail service, the 5th D ( Distance to Rail Transit) or Direct Transit Ridership Modeling, can be used to assess the effect of rail proximity on the amount of transit ridership generated in an area.
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page E- 11
Final Report
has less than basic bus or bicycle facilities, the elasticities may overestimate the reduction in VT and VMT and assume a level of bus ridership that could not be accommodated by the planned bus service. However, if the smart- growth study area plans to offer basic bus service ( similar to the service in other areas of the region with similar densities), and basic bicycle facilities ( consistent with other areas of the region with similar densities and route connectivity), the 4Ds provide a reasonable approximation of the VT and VMT reductions resulting from pedestrian, bicycle, and bus availability.
7. It is possible to calibrate the 4D elasticities to account for complementary destinations ( e. g., land- uses that provide opportunities for individual or household activity needs away from home, such as at work, to be met by non- motorized modes rather than solely by automobile) and their effect on VT and VMT reduction. This may be accomplished through developing locally validated 4D elasticities for non- home- based trip purposes, as several 4D studies have done.
Recommendations for Local Jurisdiction Practice Regarding Local Travel Modeling
1. Local jurisdictions that implement models that already have “ moderate” to “ high” smart- growth sensitivity ( Figure E- 1) should strive to continue to enhance their models regarding smart- growth sensitivity rather than to supplement them with 4D elasticities or other post- processing approaches. A model should be tested for its sensitivity to smart- growth, however, because the presence of the desirable features listed in Figure E- 1 does not guarantee sensitivity. The 4D elasticities research and other research on smart- growth effectiveness provide evidence of the expected range of sensitivity a model should have to smart- growth and can provide a benchmark for travel model testing. A model can be tested to determine whether it captures the expected range of sensitivity before a decision is made about how to add sensitivity. To perform this type of sensitivity testing, users need full access to travel demand models.
2. Due to the need to better understand and balance regional benefits associated with smart- growth strategies with localized traffic impacts, local jurisdictions that have access to a moderate- to high- sensitivity regional agency model should consider using it to assess proposed land- use plans and projects if such a model provides sufficient detail.
3. Local jurisdictions with low- sensitivity models should consider using a supplemental tool such as one of the 4D elasticities post- processors to evaluate smart- growth strategies in land- use planning efforts.
4. Methods used to capture smart- growth sensitivity ( either improvements in the travel model or supplemental tools) should be calibrated with local data and tested for reasonableness before being used to assess land- use plans or projects.
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page E- 12
Final Report
Recommendations for Local Jurisdiction Practice Regarding 4D Elasticities Tools
1. There should be testing of an existing travel model to assess whether it already has smart- growth sensitivity and whether it estimates travel activity consistent with local travel survey results in order to determine whether a post- processor ( such as the 4D elasticities) should also be used.
2. Local jurisdictions with low- sensitivity models should consider using a 4Ds methodology to gain some sensitivity to smart- growth strategies, either applied in sketch- planning software such as I- PLACE3S, INDEX, or as a spreadsheet post- processor to a local travel model.
3. It is recommended that 4Ds processes ( whether in I- PLACE3S, INDEX, or as a spreadsheet post- process to a local travel model) can appropriately be used as part of local planning, public participation, and decision- making processes, such as:
• Developing and/ or updating city and county general plans and specific area community plans
• Creating and communicating various land- use/ transportation “ scenarios” to workshop participants as part of these processes, and providing feedback to them regarding various potential benefits and impacts
• Assessing land- use projects and plans regarding air quality benefits and impacts
• As part of regional “ visioning” processes ( such as, for example, the SACOG Regional Blueprint Project) to gather input from participants and provide feedback to them regarding estimated benefits and impacts of their choices
It is not recommended that 4D elasticities processes be used for conducting corridor planning of streets or highways ( regarding numbers of lanes or other specific project- level details).
4. For transportation impact studies of proposed land- use development projects, for traffic impact fee programs, or for any CEQA or NEPA documentation, the 4Ds may be used but only if the following requirements are adequately met:
• the 4Ds elasticities are applied in conjunction with a local travel model,
• the 4Ds elasticities have been calibrated to local conditions using a local travel survey,
• the 4Ds elasticities have been calibrated to reflect smart- growth effects and trip purposes that are captured directly by the local travel model ( for models with moderate or high sensitivity), and
• the project is at least 200 acres in size.
5. For the 4D elasticities to function properly, it is necessary to apply them according to the guidelines established by the developers of the elasticities and in Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page E- 13
Final Report
a way that reflects the conditions for which they were developed ( Chapter 4). These include the following guidelines:
• Set minimum and maximum boundaries on the size of areas to be analyzed to reflect the general size of the analysis zones used in the estimation of the elasticities
• Limit the possible percentage change in the 4Ds to the range observed in the estimation data
• Calibrate to local conditions
• Use household travel surveys, if/ when they are available, to determine actual elasticities appropriate for an area before conducting analyses of land- uses using a 4D elasticities post- processor
• Follow recommendations regarding the proper use of each tool ( Chapter 4)
Recommendations for Research, Development, and Training
1. More research, development, and training should be conducted to support the use of more sophisticated modeling tools by local jurisdictions.
2. The diversity of case studies in this report indicates that " best practices" are emerging regarding use of models and tools to analyze smart- growth strategies. Training and education is needed in the form of documentation and technology transfer targeting the majority of local jurisdictions and smaller MPOs.
3. Procedures and standards should be developed for testing a travel model’s sensitivity to smart- growth conditions and judging whether the model is within an acceptable range, or the degree to which adjustment is needed.
4. The most advanced model systems, including activity- based and tour- based models, should be used to conduct research on elasticities for post- processing or correcting less sensitive models, especially to capture the benefits of modeling all modes of travel, short and long trips, and the inter- relationship between trips.
5. Better documentation and explanation of supplemental methods such as the 4Ds methodologies ( including, I- PLACE3S, INDEX, and 4D post- processors) should be developed and provided, along with parameters and recommendations for their appropriate use. Guidelines should also be provided that describe a calibration process for these tools.
6. An assessment should be undertaken of the benefits that improved regional modeling may have in assisting local governments’ abilities to analyze smart- growth land use and transportation strategies at local and site- specific levels.
7. Additional research should be conducted to further support 4D elasticities and other post- processing methods to provide more direct sensitivity to smart- growth effects and to reduce correlation with other factors. There should also be research conducted on the elasticities for a broader range of area types. 9
9 Research currently underway includes: NCHRP Project 08- 51, “ Enhancing Internal Trip Capture Estimation for Mixed- Use Developments,” is currently assembling data on vehicle trip generation rates in mixed- use developments. NCHRP Project 08- 66, “ Trip- Generation Rates for Infill Land Use Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page E- 14
Final Report
8. The 4Ds elasticities, outside of proprietary and copyrighted software, should evolve as “ open architecture” freely available via the Internet.
9. The elasticities in proprietary and open source software should be tested periodically to verify their evolution over time and, most importantly, their transferability across California.
10. Additional research should be conducted with models from one or more case- study areas to assess how much sensitivity is added by different levels of improvement of UTMS modeling and by activity- based modeling. Comparison of results should be made with results from 4D methods to assess the effectiveness of 4D calibration to local model sensitivity. Sensitivity testing should also be used to provide insights regarding which smart- growth strategies are most effective in different types of locations and settings.
Developments in Metropolitan Areas” was recently approved. In addition, U. S. EPA is initiating a study that may provide the opportunity to update the 4D elasticities with more recent national data.
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page E- 15
Final Report
( This page intentionally blank)
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page E- 16
Final Report
Chapter 1
Introduction
1.1 Project Purpose and Objectives
In the past decade, frustration with increasing congestion, air pollution, and suburban sprawl has led to a resurgence of interest in land development patterns, often labeled as “ smart- growth,” including: mixed land- uses, urban and suburban infill, pedestrian and bicycle- oriented design, and transit- oriented developments. The features of smart- growth are generally designed to allow residents to be less dependent upon travel by automobiles. The purpose of this project has been to review the travel modeling methods used by local jurisdictions ( e. g., cities and counties) in California to determine whether there is adequate sensitivity to smart- growth strategies to evaluate the potential impact on trip making and vehicular travel.
Interest in smart- growth strategies has been demonstrated in California by policy statements included in Go California, the Mobility Action Plan of the California Transportation Plan 2025. The document identifies as some of the key strategies to promote more efficient development patterns:
• Increasing densities and using design to facilitate effective transit service
• Promoting street and urban design to encourage walking and bicycling
• Providing information and technical assistance on transit- oriented design
• Encouraging localities to foster “ smart- growth” development practices
• Promoting the revision of local zoning regulations to allow for higher density and mixed- use developments
Along with the increasing interest in new community design have come questions about whether the conventional Urban Transportation Modeling System ( UTMS), or “ four- step” travel demand model as it is commonly known, has the capability to effectively quantify the impacts and benefits associated with smart- growth characteristics, such as those listed below:
• Land- use location
• Land- use density
• Land- use diversity
• Transportation network configuration
• Non- motorized mode facilities ( such as pedestrian and bicycle paths)
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page 1- 1
Final Report
For example, clustering of services such as dry cleaning, day care, restaurants, and stores near major employment sites can provide the opportunity for workers to take care of personal errands on foot from work and possibly avoid unnecessary motor vehicle trips. Most travel models used by local jurisdictions in California do not reflect the differences in vehicle trip generation that result from such clustering of mixed uses. Transit ridership can also vary as a function of the difficulty in crossing streets at bus stops and the presence of waiting shelters and sidewalks, but these micro- scale design features are not recognized in most regional or local models. Building an ideal travel model to address these smart- growth issues would require the collection and interpretation of more data than has been used in current travel forecasting activities. The level of detail required for models of non- motorized modes is much finer than typically encountered in travel forecasting models in use today.
This report provides a review of current modeling practice in California and identifies applications that are designed to quantify the effects of smart- growth on local travel demand. In Chapter 2, the review begins with a brief overview of travel demand models and their use in local land- use decision- making. It is followed in Chapter 3 by a detailed review of the conventional modeling process used by most local jurisdictions in California and the limitations of the approach for smart- growth sensitivity. Chapter 3 also identifies methods for improving the sensitivity of conventional UTMS modeling and provides examples of where innovative practices have been implemented in California.
Chapter 4 provides a review of several existing supplemental tools that are currently in use for gaining smart- growth sensitivity through the application of what are commonly called the “ 4D elasticities:” I- PLACE3S, INDEX, and a 4Ds Post- Processor. Chapter 5 provides a review of current modeling practice in California. The review is intended to be a general overview of how travel models are used by local jurisdictions to support local land- use decision- making. Specific attention is given to the extent to which travel models have been used to make decisions about smart- growth strategies. Six case studies are included to illustrate the range of practice in California.
Chapter 6 provides the results of a sensitivity test of one of the 4Ds- based supplemental tools ( INDEX) designed to increase smart- growth analysis sensitivity. The results from INDEX application are compared with the results from the baseline travel model. Chapter 7 summarizes the conclusions and recommendations from the study and identifies directions for additional research.
Appendix 1 of this report provides a list of the members of the Technical Advisory Committee that provided guidance for the study, and of the research team. Appendix 2 provides definitions for the acronyms used in the report, and Appendix 3 is a glossary of terms used in transportation, modeling, and related topics.
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page 1- 2
Final Report
1.2 Smart- Growth Strategies
Although there are different opinions about what constitutes smart- growth, the following design principles of a smart- growth community as articulated by the U. S. Environmental Protection Agency ( U. S. EPA) 10 capture the elements most commonly included:
1. Mix land- uses
2. Take advantage of compact building design
3. Create a range of housing opportunities and choices
4. Create walkable neighborhoods
5. Foster distinctive, attractive communities with a strong sense of place
6. Preserve open space, farmland, natural beauty and critical environmental areas
7. Strengthen and direct development towards existing communities
8. Provide a variety of transportation choices
9. Make development decisions predictable, fair and cost- effective
10. Encourage community and stakeholder collaboration in development decisions
Transit- oriented development refers to land development patterns that place the development of various commercial and residential activities around a transit station. The design principles of transit- oriented development can be seen as a subset of those of smart- growth. Transit- oriented neighborhood design features typically include:
• Mixed land- use
• Compact development
• Destination within easy walking distance of transit
• Neighborhood focal point
• Pedestrian orientation
In the remainder of this report the term “ smart- growth” is used to refer to all of the strategies identified above.
Smart- growth strategies can have an effect on travel behavior in a variety of ways. The ways in which they affect travel behavior have direct implications for whether travel models used by local jurisdictions are sensitive to the smart- growth strategies. They also have direct implications for what kinds of improvements to the models or supplemental methods might improve the local jurisdictions’ ability to evaluate smart- growth strategies in their land- use planning processes. The research team identified four key intended objectives of smart- growth strategies as follows:
Providing opportunities to satisfy travel needs at nearby destinations with shorter vehicle trips, trip chaining, or non- motorized travel.
10 U. S. EPA’s Smart- growth Network: http:// www. epa. gov/ smartgrowth/ about_ sg. htm Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page 1- 3
Final Report
• Using land- use to create trips with origin- destination pairs that are more easily traveled by “ alternative” modes such as transit, walking, and/ or bicycling.
• Providing better and more attractive conditions for travel by alternative modes.
• Providing economic incentives for the use of alternative modes.
The research team also identified examples of specific ways in which smart- growth strategies can produce these effects, and these are provided in Table 1.1. The assessment of local jurisdiction modeling practice and supplemental methods for their smart- growth sensitivity was conducted with these potential effects as the frame of reference.
1.3 Research Approach
This study was conducted through a combination of literature review, survey, case study analysis, and sensitivity testing of models. A Technical Advisory Committee ( TAC) was formed to provide guidance and quality control for the project and also to provide technical input on the state of modeling practice in the state. A list of the TAC members and the other study participants is available in Appendix 1.
The research team performed a thorough review of conventional UTMS travel models that are used by most local jurisdictions to determine what limitations in the model influence sensitivity to smart- growth. Each major component of the four- step model was reviewed. Suggestions were generated regarding how the sensitivity of the conventional model could be improved.
The current state- of- the- practice of travel modeling for land- use planning and decision- making in California was characterized by conducting a survey of the TAC members and the professional experience of the research team. The review was designed to provide a profile of the range of travel- forecasting tools used, the applications of tools for land- use planning, and efforts made to gain smart- growth sensitivity. The range of practice is illustrated in more detail by a review of six case- study cities:
• Fresno
• Irvine
• San Diego
• San Jose
• San Luis Obispo
• West Sacramento
These case studies illustrate different local approaches to travel modeling and various approaches to analyzing land- use plans and projects, especially regarding smart- growth strategies.
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page 1- 4
Final Report
Table 1.1 Intended Effects from Smart- Growth Strategies on Travel Behavior
11.1Clustering of potential non- home destinations such as daycare, cleaners, restaurants, stores, etc. near work sites1.2Providing a higher level of diversity in mixed- use clusters1.3Developing neighborhoods with more self- sufficient land uses1.4Providing more jobs- housing balance within sub- areas of regions that allows shorter commutes1.5Providing a more complete range of housing options and pricing near employment centers22.1Providing higher density residential and work sites near transit2.2Providing higher density residential and work sites along bike routes and trails2.3Location of schools along bicycle routes and trails2.4Clustering potential destinations such as daycare, cleaners, restaurants, stores near work sites and high density residential areas33.1Locating business entrances as close as possible to transit stops or stations3.2Locating entrances to higher density residential buildings as close as possible to transit stops or stations3.3Providing good pedestrian and bicycle access to transit stops or station3.4Providing bicycle storage facilities at transit stops and stations3.5Providing bicycle storage facilities at high density residential developments, work places, schools, and shopping areas3.6Locating development on a grid street network3.7Providing a high level of sidewalk coverage4Provide economic incentives for use of alternative modes4.1Providing a limited supply of parking4.2Charging separately for parking at multi- family residential, employment and shopping sitesProviding opportunities to satisfy travel needs at nearby destinations with shorter vehicle trips, trip chaining or non- motorized travelUsing land use to create trips with origin- destination pairs that are more easily traveled by alternative modesProviding better and more attractive conditions for travel by alternative modesSmart- Growth Effect and Smart- Growth Strategies Designed to Achieve the Effect
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page 1- 5
Final Report
Researchers also conducted a review of existing tools for supplementing conventional models to gain smart- growth sensitivity by examining documentation of the tools. The review focused on how each of three 4D- based tools - I- PLACE3S, INDEX, and 4D post- processors - captured the additional sensitivity and the data used to provide that sensitivity. This report describes the structure of each of these tools, along with the equipment, data, and other resources and guidelines required for their appropriate application.
To gain a better understanding of how the existing tools for supplementing travel models work and the differences they produce for a sample urban environment, a “ sensitivity test” was conducted using the 4D elasticities. The tests were conducted using the INDEX software applied to travel data available from West Sacramento. 11 The sensitivity tests were designed to assess how much reduction in travel demand that INDEX predicts would result from a variety of strategies. The sensitivity test also provided an assessment of the data and effort necessary to use the 4D elasticities in INDEX.
The research team and TAC members generated a set of conclusions and recommendations from the study based on the results of the activities described above. The focus of the conclusions and recommendations ( Chapter 7) is on how local jurisdictions can, in the short run, make the most effective use of available models and tools to gain smart- growth sensitivity. Recommendations were also developed regarding additional steps that could lead to more smart- growth sensitivity in models and tools available to local jurisdictions.
11 Sensitivity tests of I- PLACE3S or a 4D post- processor were not conducted due to insufficient time and other resources. Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page 1- 6
Final Report
Chapter 2
Overview of Travel Models and Their Use in Local Planning
2.1 Uses of Models in Local Land- use and Transportation Planning
In California, as in most states, land- use planning and approval of development projects is the responsibility of the cities in incorporated areas and the counties in un- incorporated areas. Cities and counties in California have the responsibility to prepare a general plan as a statement of development policies setting forth objectives, principles, standards, and plan proposals for the coordination of land- use, circulation, housing, open space, conservation, environmental quality and safety. The general plan is usually developed with the aid of a travel model that can translate alternative land- use forecasts and configurations into travel patterns. Because of the availability of personal computers and fairly standardized software packages for applying travel models, most cities and counties have the ability to develop and use a local travel model for development of the general plan and for other uses.
Cities and counties also have the authority to review and approve land- use development projects. That review typically includes an assessment of the potential impact of the development on the transportation system. Again this review is frequently aided by the application of a travel model to assess the additional travel that could be generated by the development.
At a regional level, transportation planning is required in the United States as a conditional requirement to receive federal transportation funds for larger urban areas. Requirements for urban transportation planning emerged during the early 1960s. The Federal- Aid Highway Act of 1962 created the federal requirement for urban transportation planning largely in response to the construction of the Interstate Highway System and the planning of routes through and around urban areas. The Act required, as a condition attached to federal transportation financial assistance, that transportation projects in urbanized areas of 50,000 or more in population be based on a continuing, comprehensive, urban transportation planning process undertaken cooperatively by the state and local governments -- the birth of the so- called 3Cs, “ continuing, comprehensive and cooperative” planning process.
Throughout the years, the requirements have been expanded and modified in subsequent legislation, through the Intermodal Surface Transportation Efficiency Act of 1991 ( ISTEA), the Transportation Efficiency Act ( TEA- 21), and the Safe, Accountable,
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page 2- 1
Final Report
Flexible, Efficient Transportation Equity Act - A Legacy for Users ( SAFETEA- LU) in 2006. ISTEA listed 15 specific factors that must be considered in urban transportation planning. These factors have led to regulations that require planning agencies to deal more directly with air quality issues, multi- modal planning, and better management of existing systems, expanded public input, and financial analysis requirements. Generally, they have led to a greater role for transportation planning in urban areas, and to the consideration of a wider range of alternatives and consequences of transportation investment choices.
In addition to national laws and regulations, California requires urban counties to develop and maintain travel models for use in the Congestion Management Program. This requirement originated from Proposition 111, passed by California voters in 1990. Proposition 111 added nine cents per gallon to the state fuel tax to fund local, regional, and state transportation projects and services. It also required 32 “ urban counties” to designate a “ Congestion Management Agency”, whose primary responsibility is to develop and maintain a “ countywide transportation computer model: to coordinate transportation planning, funding and other activities in a congestion management program.” The codified task is in California Government Code Section 65089 ( c):
The agency, in consultation with the regional agency, cities, and the county, shall develop a uniform data base on traffic impacts for use in a countywide transportation computer model and shall approve transportation computer models of specific areas within the county that will be used by local jurisdictions to determine the quantitative impacts of development on the circulation system that are based on the countywide model and standardized modeling assumptions and conventions. The computer models shall be consistent with the modeling methodology adopted by the regional planning agency. The data bases used in the models shall be consistent with the databases used by the regional planning agency. Where the regional agency has jurisdiction over two or more counties, the databases used by the agency shall be consistent with the databases used by the regional agency.
The requirement for a Congestion Management Program does not apply in a county in which a majority of local governments that represent a majority of the population in the county adopt resolutions electing to be exempt from the congestion management program.
2.1.1 Policy Development ( Sketch Planning)
Policy development often involves exploring potential outcomes in a broad- based way as a way of screening down options to identify strategies that are worthy of more investigation. Travel models can provide important information regarding some benefits and costs of various options and scenarios.
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page 2- 2
Final Report
Policy studies often examine model results from prior studies as a point where trends and potential issues can be identified. If further system alternatives are to be considered, models can be used to test the effects of system changes. Some ways that travel models can be used vary depending on the policy choices being considered and also the model design.
Examples of the types of options and questions that travel models are typically used to assess include: whether and where traffic congestion levels may get worse, whether specific roadways will reach congested conditions, and the direct effects of land- use growth patterns on the transportation system. For example, if a travel model has sensitivity to transit service, that same model can be used to examine whether or not increases in transit service ( resulting in increased transit service frequencies) or changes in transit fares may result in mode shifts. If the travel model has sensitivity to vehicle occupancy with HOV lanes, then different lane assumptions can be tested. Finally, area- wide measures such as aggregate vehicle miles of travel ( VMT) or vehicle hours of travel ( VHT) can be estimated to describe system performance.
2.1.2 General Plan
California communities must have an adopted General Plan, as defined in California Government Code 65300. A General Plan is a set of policies and maps designed to establish how the community will change should the community continue to experience development. General plans address various aspects of community planning including circulation, which is one of the core elements required by state law.
Travel models are used in General Plans, both in plan development as well as in the assessment of potential environmental impacts resulting from General Plan implementation. The procedure is to examine system performance and compare the consequences of leaving an existing General Plan intact or adopting an updated document.
2.1.3 Specific Plan
A Specific Plan is similar to a General Plan, but for a portion of the jurisdiction rather then an entire city or county. This planning concept is intended to set a series of area- wide improvements into motion, including possible set- asides for rights- of- way, exactions, and programming for new transportation facilities. This planning process is governed by California Government Code 65450 to 65457. A Specific Plan includes a text and a diagram or diagrams that specify all of the following in detail:
• The distribution, location, and extent of the uses of land, including open space, within the area covered by the plan.
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page 2- 3
Final Report
• The proposed distribution, location, and extent and intensity of major components of public and private transportation, sewage, water, drainage, solid waste disposal, energy, and other essential facilities proposed to be located within the area covered by the plan and needed to support the land- uses described in the plan.
• Standards and criteria by which development will proceed, and standards for the conservation, development, and utilization of natural resources, where applicable.
• A program of implementation measures including regulations, programs, public works projects, and financing measures necessary to carry out the Plan.
• A statement of the relationship of the Specific Plan to the General Plan.
Travel models are used in Specific Plans to assess the potential consequences of various proposed actions. Traffic impact analyses ( TIAs) are often conducted for Specific Plans as part of California Environmental Quality Act ( CEQA) requirements.
2.1.4 Transportation Investment Study/ Corridor Study
Studies and strategies are often performed to define potential transportation investments in major corridors. Special studies are often needed to reduce the number of alternative strategies, and/ or to refine the content of alternatives. These studies then are used to inform decision- makers regarding more detailed environmental studies and design- related questions.
One key use of travel demand models is to assist in the development of investment strategies for transportation corridors. Depending on the type of model that is used and the alternatives being proposed, a travel model can provide responsive information on the demand that would result from different alternatives, providing one key piece of information in helping decision- makers reduce the number of alternatives. Travel models also provide input to micro- level traffic simulation models that are used in defining the geometric requirements of the roadway or intersection design based on an analysis of intersection “ levels of service” and related queue lengths, or on segment level of service and related technical performance of merging, diverging, and weaving analysis.
2.1.5 Traffic Impact or Development Fee Program
Some jurisdictions have enacted traffic impact or development fee programs. Developer fees are dedicated assessments that are applied to new development in a district for the purpose of funding new transportation projects that would be needed as a result of growth. Such assessments help ensure that a community’s transportation performance standards would continue to be met. Developer fees provide a “ fair share” mechanism for funding transportation improvements on a proportional basis rather than requiring that a particular transportation project be funded through a single land- use development. In
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page 2- 4
Final Report
California, development fees are enabled by California Government Code 66000 through 66008, which establishes the authority and procedures for creating and operating a program.
Travel models are often used as tools in developing and updating assessment fee programs. They represent one of the most defensible tools available for addressing many technical questions involved in fee studies. Travel models typically are used to estimate the proportion of traffic growth attributable to new development, identify the origins or destinations of the new traffic, determine an average forecasted trip length as a basis for the size of the fee district, and assess whether the proposed program to be funded by the fee will address the anticipated system deficiencies adequately.
2.1.6 Traffic Impact Analysis/ CEQA Analysis for New Development
One current standard use of travel models is to analyze traffic impacts of new development, as required by the California Environmental Quality Act ( CEQA), a California statute that became law in 1970. CEQA requires state, regional, and local agencies to identify and assess the significant environmental impacts of their actions and to avoid or mitigate those impacts, if feasible. The current CEQA law is found in the California Public Resources Code Division 13: Environmental Protection.
Each “ lead agency” accepts an Environmental Impact Report ( EIR), Negative Declaration, or Categorical Exemption regarding proposed new plans and development projects. Other communities or government agencies – and the public - can provide feedback during the initial stages of document preparation (“ Notice of Preparation”) or through a review of the draft EIR. The CEQA process includes a requirement to examine circulation issues. Forecast traffic volumes are also used in analysis of air quality and noise effects related to the proposed project ( these are also studied through the CEQA process).
Travel models often provide a technical resource for preparation of CEQA studies. For example, travel models can be a source of background volumes, of trip and/ or distribution of traffic generated by the development proposal, and of the aggregate impacts of new roadways or other improvements that may be contained in the development proposal. Typically, a travel model will provide traffic volume forecasts for cumulative “ no project” and “ cumulative plus project” conditions. These traffic volumes have a direct influence on the need and extent of mitigation.
Given this reliance on travel models by local agencies that control land- use decisions, clearly defining the “ state- of- the- practice” for local modeling is an important first- step before recommending that local agencies invest in new or improved features that will increase the sensitivity of their models to smart- growth strategies.
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page 2- 5
Final Report
2.1.7 Transportation Project EIS/ EIR under NEPA/ CEQA
Transportation projects that require construction and obtain federal funding must have an Environmental Impact Statement ( EIS) as required by the National Environmental Policy Act ( NEPA), passed in 1969. The adoption of the related CEQA in 1970 established a set of more specific rules that, if applied, typically also satisfy the NEPA process. Minor projects may be exempted from NEPA and CEQA depending on the urgency, nature and size of the project.
Often, transportation projects funded with Federal Highway Administration ( FHWA) resources must be supported by an analysis of anticipated traffic conditions 20 years after project completion. Regional travel models are typically used to provide the necessary travel forecast. Forecast traffic volumes are also used in analysis of air quality and noise impacts, which are also studied through the NEPA/ CEQA process.
Travel models are most often used to forecast future traffic volumes on area roadways. While models can be used to forecast some operational conditions on the roadways, they typically are not used in this way because models are not typically calibrated to operational attributes such as delay or travel time.
2.1.8 Transit New Starts Project Analysis
Federal funding for transit projects began in the 1960s. The popularity of transit projects began to rise in the 1970s, and a need emerged at that time for a better process to determine the relative benefits of making transit capital investments from the competitive Federal Transit Administration’s ( FTA) New Starts grant program. The appropriation of New Starts funding is now tied to a rating system established by FTA that includes existing and planned land- uses.
The adoption of TEA- 21 in 1998 began to institutionalize the New Starts funding reports in a more comprehensive way. This federal act requires FTA to:
• Develop a rating for each criterion as well as an overall rating of “ highly recommended,” “ recommended,” or “ not recommended” and use these evaluations and ratings in approving projects’ advancement toward obtaining grant agreements; and
• Issue regulations on the evaluation and rating process.
TEA- 21 directs FTA to use these evaluations and ratings to decide which projects to recommend to Congress for funding in a report due each February. These funding recommendations are also reflected in the U. S. Department of Transportation’s ( USDOT) annual budget proposal. In the annual appropriations act for USDOT, Congress specifies the amounts of funding for individual New Starts Program projects.
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page 2- 6
Final Report
Travel model data are a key source of information for evaluating New Starts project proposals. Many calculations are based upon reports on rider demand, congestion, and impacts and benefits to other transit and transportation systems.
Because many travel models have not been adequately sensitive to transit demand, FTA has received many grant applications with potentially inaccurate transit rider forecasts. Consequently, the FTA has developed an evaluation process to closely review inputs, land- uses, and behavioral assumptions in travel models to determine whether New Starts program grant applicants have properly developed forecasts of rider demand.
2.2 Types of Transportation Planning Models
Travel demand models are used in the regional transportation planning process, which involves modeling and forecasting of the influences that various policies, programs and projects may have on travel in a region. The modeling and forecasting process also provides fairly detailed information, such as traffic volumes, transit ridership, and turning movements, to be used by engineers and planners in their designs. Travel demand forecasts typically include estimates of the number of cars on a future freeway or the number of passengers using a transit service. When properly designed and implemented, a regional travel model might also be able to predict the amount of reduction in auto use that could occur in response to central- area parking fee programs.
To decide which actions to implement, decision- makers need to understand how each potential improvement measure could affect the transportation system and the region as a whole. Models are used to estimate the number and types of trips that will be made on transportation system alternatives at future dates. These estimates are the basis for regional transportation planning and are used in major investment analyses, environmental impact analyses, and in setting priorities for infrastructure improvements. An understanding of modeling processes is therefore important to better understand how they are used in decision- making processes.
Several different techniques and models for travel demand forecasting are available depending on the requirements of the analysis. These techniques differ in complexity, cost, level of effort, sophistication and accuracy, but each has its place in travel forecasting. Each modeling technique is explained briefly below.
2.2.1 Sketch Planning Tools
Sketch planning involves the preliminary screening of possible configurations or concepts. It is used to compare a large number of proposed policies in enough analytical detail to support broad policy decisions. Useful in both long- range and short- range planning and in preliminary corridor analyses, sketch planning – that has minimal data costs - yields rough aggregate estimates of capital and operating costs, patronage,
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page 2- 7
Final Report
corridor traffic flows, service levels, energy consumption, and air pollution. The planning process usually remains in the sketch- planning mode until comparisons of possibilities are completed or a strategic plan worthy of consideration at a finer level of detail is obtained.
Sketch- planning tools designed for smart- growth sensitivity have been used in California for charrette or workshop- style visioning exercises to assess the potential benefits of various strategies in a city, county, or region. The quick turnaround provided by the sketch planning models allows a group to test many options in a short period of time.
2.2.2 Conventional Models ( 4- Step Models)
Conventional models deal with many fewer alternatives than sketch planning tools, but in much greater detail. Inputs typically include demographic data, the location of principal roadway facilities, and delineated transit routes. At this level of analysis, the outputs are detailed estimates of number of lanes of a highway, transit fleet size and operating requirements for specific service areas, refined cost and patronage forecasts, and level- of- service measures for specific geographical areas. The cost of examining an alternative at the traditional level could be 10- 20 times its cost in sketch planning, although default models - which dispense with many data requirements - can be used for a less expensive “ first look.” Potentially promising plans can be analyzed in detail, and problems uncovered at this stage may suggest a return to sketch planning to accommodate new constraints.
2.2.3 Activity- Based Models
Activity- based models represent a significant restructuring of modeling of travel demand. Instead of structuring the modeling around the trip as is done in UTMS, activity- based models structure the modeling around the activities that a household wishes to pursue during a day and how travel can occur to satisfy the activity desires. Travel is modeled in “ tours” rather than trips and the decision- making unit is the household rather than all the households in a zone. Activity- based modeling is an emerging method that holds promise for improving smart- growth sensitivity because it recognizes that trips made by a household are not independent of each other but are often connected for efficiency or convenience. Many smart- growth strategies are designed to reduce vehicular travel by making it easier for individuals or households to chain trips together. Only two activity- based models have been developed to date in California: by the San Francisco County Transportation Authority and by the Sacramento Area Council of Governments. A brief overview of how these models can address some of the common deficiencies in UTMS models is provided in Chapter 3.
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page 2- 8
Final Report
2.2.4 Micro- level Traffic Models
Micro- level or post- processing traffic models are applicable when actual implementation of a project grows near. They are the most detailed of all transportation planning tools. At this level of analysis, it is possible to make a detailed evaluation of the congestion levels of passenger and vehicle flows through a particular intersection, transportation terminal, or activity center. Final analysis may draw upon conventional traffic operations analysis using deterministic software programs such as HCS, TRAFFIX, or SYNCHO, or more complex stochastic micro- simulation traffic operations software programs such as CORSIM, SIMTRAFFIC, PARAMICS, or VISSIM.
Micro- level traffic operations analyses usually draw upon traffic volume output from a relevant travel demand model as direct inputs to the traffic operations models. This may take the form of trip tables, link volumes, or intersection turning movement volumes. Near- term planning is most effective when traffic volumes from actual counts can be used for the micro- simulation inputs, but it is sometimes necessary to use the traditional longer- range planning model to forecast future count data.
2.3 The Conventional ( UTMS) Transportation Planning Model
The history of demand modeling for passenger travel has been dominated by the modeling approach, which has come to be referred to as the Urban Transportation Modeling System ( UTMS). Travel has always been viewed in theory as derived from the demand for activity participation, but in past practice has been modeled with trip- based rather than activity- based methods. Trip origin/ destination ( OD) surveys, rather than activity surveys, form the principle database. As the sequence of modeling steps in the conventional forecasting process proceeds, there is less attention to the activities that the travel satisfies and more attention to the point- to- point trips that are made. The application of this modeling approach is currently nearly universal.
UTMS might best be viewed in two stages. In the first stage, various characteristics of the traveler and the land- use activity system ( and to a varying degree, the transportation system) are " evaluated, calibrated, and validated" to produce a non- equilibrated measure of travel demand ( or trip tables). In the second stage, this demand is loaded onto the transportation network in a process that amounts to formal equilibration of route choice only, not of other choice dimensions - such as destination, mode, time- of- day, or whether to travel at all ( feedback to prior stages has often been introduced, but not in a consistent and convergent manner). Although this approach has been moderately successful in the aggregate, it has failed to perform in most relevant policy tests, whether on the demand or supply side.
Transportation modeling developed as a component of the process of transportation analysis, which came to be established in the United States during the era of post- war development and economic growth. Initial application of analytical methods began in the
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page 2- 9
Final Report
1950s. The initial development of models of trip generation, distribution, and diversion in the early 1950s led to the first comprehensive application of the four- step model system in the Chicago Area Transportation Study. The focus was decidedly highway- oriented with new facilities being evaluated versus traffic engineering improvements.
The 1960s brought federal legislation requiring " continuous, comprehensive, and cooperative" urban transportation planning, fully institutionalizing the UTMS. Further legislation in the 1970s brought environmental concerns to planning and modeling, as well as the need for multimodal planning. It was recognized that the existing model system might not be appropriate for application to these emerging policy concerns. In what might be referred to as the " first travel model improvement program," a call for improved models led to research and the development of disaggregate travel demand forecasting and equilibrium assignment methods that integrated well with the UTMS and have directed modeling approaches for most of the last 25 years. The late 1970s brought " quick response" approaches to travel forecasting and independently the start of what has grown to become the activity- based approach.
A growing recognition of the misfit of UTMS regarding relevant policy questions in the 1980s led to the Federal Travel Model Improvement Program in 1991. As a result, much of the last decade has been directed at improving the state- of- the- practice relative to the conventional model, while also fostering research and development regarding new methodologies to further the state- of- the- art, such as disaggregate simulation of households and activity- based models. ( Many of the limitations of UTMS specifically for modeling smart- growth strategies are identified in a review of the conventional UTMS model in Chapter 3. The chapter also identifies some innovations in practice that can increase sensitivity of UTMS models to smart- growth strategies and provides examples of applications in California where such innovations have been incorporated.)
2.3.1 Limitations of Travel Demand Models
Travel demand modeling was developed primarily for highway planning. As the need to examine other issues such as transit, land- use planning, and air quality analysis has arisen, the modeling process has been modified to add additional techniques to attempt to deal with these needs. Travel models provide forecasts only for those factors and alternatives that are explicitly included in the equations and data of the models. If the models are not sensitive to certain polices or programs, the models’ outputs will not include the effect of these policies or programs. More specifically, these policies and programs cannot be formulated as input variables into the models. For example, travel- forecasting models usually do not include pedestrian and bicycle trips; therefore, plans or programs that include bicycle or pedestrian system improvements cannot be evaluated with the conventional modeling procedure if the models ignore these types of trips. However, it would not be correct to conclude that pedestrian or bicycle improvements are ineffective. The actual impact is unknown. Therefore it is critical that the assumptions used in the modeling process and the model limitations be explicitly stated and considered before decisions are made based on their results.
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page 2- 10
Final Report
One concern in modeling of smart- growth strategies by local jurisdictions with available travel models is the time it takes for many of the strategies to have a significant impact within an area. In older parts of urban areas where some of the best opportunities exist for in- fill development and development near transit services, the time required to achieve a significant amount of smart- growth development may be long. In some cases this may be beyond the forecast time frame of the local model and beyond the time frame of the jurisdictions general plan. Even when the smart- growth is occurring in more suburban areas where the developments may be larger, full build- out of the developments may be staged over a long period of time and the effects from the smart- growth of the developments may not be present in the earlier stages of the development.
The amount of new development in higher density urban areas may also be small compared to the existing land- use in an area. As a result, the vehicle trip and VMT rates per capita for the new development may be lower in the high- density area than in a corresponding development in a less dense suburban area, but the impact on an area- wide scale may be virtually un- noticeable when only the area- wide vehicle trip or VMT is used as the measure. Using a travel model to test smart- growth strategies in a development can mask the potential benefits of the strategies unless care is taken to examine the vehicle trip and VMT reduction benefits to, from and within the proposed smart- growth development.
2.4 New Methods of Reflecting Smart- growth
A variety of new methods have been developed in recent years to add sensitivity to the conventional UTMS model, and the methods span a broad spectrum in terms of complexity, resources required for implementation, and resources required for maintenance. There is also significant variation in how the different methods can be used in support of land- use planning for local jurisdictions. These methods can be categorized in four general approaches:
• Post- processor to UTMS for application of smart- growth trip and VMT elasticities
• Stand- alone tools for aggregate application of smart- growth trip and VMT elasticities
• Enhancement of UTMS models
• Integrated land- use/ economic/ and transportation models
Methods in the first two categories involve the application of vehicle trip and VMT “ elasticities” for smart- growth strategies estimated on the basis of cross- sectional comparison of areas with smart- growth characteristics to areas without these characteristics. In both of the first two categories, the elasticities are applied to baseline travel data provided by a travel model. A progression of research efforts have contributed to the development of what are referred to as the “ 4D Elasticities” because they reflect
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page 2- 11
Final Report
the potential reduction in vehicle trips and VMT associated with changes in land- use characteristics that reflect smart- growth strategies.
In the first category – 4D elasticities post- processor to UTMS - methods are designed to directly supplement the UTMS model by factoring trip ends in the model to account for the effects of smart- growth strategies with the capability to produce assignments that reflect the factored trip ends. Methods in the second category – stand- alone tools - apply the elasticities to aggregate measures of travel to estimate what the area- wide effect of smart- growth strategies may be. These methods are designed primarily for interactive planning in a workshop or charrette setting during which alternative land- use strategies can be tested by participants. Two of the specific tools that have been used in California for this purpose are I- PLACE3S and INDEX. The results of a detailed review of the methods in these first two categories are provided in Chapters 4 and 6.
The final category - integration of land- use, economic, and travel data and models - provides more direct linkages between these complex systems and how they interactively affect one another. In a fully integrated modeling process, travel demand is a function of existing and future land- uses and economic activities. In turn, future land- uses and economic activities are also functions of the transportation system as well as demand on the system. These interactive analytical processes are replicated through numerous iterations. This interactive analysis system provides smart- growth sensitivity because it recognizes the synergistic effects that such strategies can have over time. For example, the economic and travel response to the implementation of smart- growth strategies can result in greater market demand for smart- growth projects and programs. The state- of- the- practice and advancements in this category are the subject of another Caltrans- funded study, Assessment of Integrated Land- use/ Transportation Models. 12
Assessment of Local Models and Tools for Analyzing Smart- Growth Strategies Page 2- 12
12 “ Assessment of Integrated Transportation/ Land Use Models Final Report,” Robert Johnston & Michael McCoy, UC Davis, May 31, 2006. http:// www. ice. ucdavis. edu/ um/ ( Final Report) Final Report
Chapter 3
Review of the Conventional Transportation Planning Model: Characteristics, Sensitivity to Smart- Growth Strategies, and Areas for Possible Improvement
3.1 General Characteristics
The Urban Transportation Modeling System ( UTMS), commonly known as the travel demand model, is the primary tool used for forecasting future demand and performance of a transportation system, typically defined at a regional or sub- regional scale. This chapter provides a review of UTMS, including a description of its features and the process by which travel forecasts are produced. The chapter also provides an assessment of some of the limitations of UTMS, as it is commonly applied, for assessment of smart- growth strategies. A summary of the limitations of UTMS for smart- growth analysis and the improvement options is provided in Table 3.1 at the end of this chapter.
There are several examples of UTMS applications in California that have addressed one or more of the limitations with an approach that increases the smart- growth sensitivity, and some of these examples are provided. The most sophisticated applications of UTMS are generally those by Metropolitan Planning Organizations ( MPOs) for large urban areas, and so many of the examples provided in this report for improvement options come from the large MPOs in the state. Because it is becoming common for local jurisdictions within a major metropolitan area to use a focused version of an MPO model, advanced practices are ( or could be) available to the local jurisdictions in the region as well.
For UTMS to be optimally useful, models must be suitably policy- sensitive to allow for the comparison of alternative programs, policies, and projects to influence future travel demand and performance. However, the model system was developed primarily for evaluating large- scale infrastructure projects, and not for more subtle and complex policies involving management and control of existing infrastructure or introduction of programs that directly influence travel behavior.
Application of travel- forecasting models is a continuous process. The period required for data collection, model estimation, and subsequent forecasting exercises may take years, during which time the activity and transportation systems change, as do policies of interest - often requiring new data collection efforts and a new modeling effort.
Assessment of Local Models and Tools for Analyzing Smart- growth Strategies Page 3- 13
Final Report
A study area can be defined to encompass the area of expected policy impact; a cordon line defines this area. The area within the cordon is composed of Traffic Analysis Zones ( TAZs) and is subject to explicit modeling and analysis. Interaction with areas outside the cordon is defined via external “ stations” which effectively serve as gateways for trips into, out of, and through the study area. The Activity System for these external stations is defined directly in terms of trips that pass through them, and the models that represent this interaction are separate from and less complex than those that represent interactions within the study area ( typically, growth factor models are used to forecast future external traffic).
The internal Activity System is typically represented by socio- economic, demographic, and land- use data defined for TAZs or other convenient spatial units. The number of TAZs ( usually based on purpose for the model, size of analysis area, data availability, and model vintage) can vary significantly from a few hundred to several thousand. The unit of analysis, however, can vary over stages of the UTMS and might be at the level of individual persons, households, TAZs, or some larger aggregation for different steps. In the majority of models, TAZs are derived from US Census geographical subdivisions. Data releases follow the Decennial Census lagged by a few years for data packaging to develop TAZs in a form known as Census Transportation Planning Package ( CTPP).
The Transportation System is typically represented via network graphs defined by links ( one- way homogeneous sections of transportation infrastructure or service) and nodes ( link endpoints, typically intersections or points representing changes in link attributes). Both links and nodes have associated attributes ( for example, length, speed, and capacity for links and turn prohibitions or penalties for nodes). The activity system is interfaced with the Transportation System via centroid connectors which are abstract links connecting TAZ centroids to realistic access points on the physical network ( typically mid- block or at points where minor collector streets meet the arterial streets represented in the model, usually not connected to nodes representing roadway intersections). Different networks may be used to represent different modes. If a transit network is included, it will define routes, stops, schedules and fares for service as well as the links that the service can use.
The UTMS provides a mechanism to determine capacity- constrained flows. For elementary networks, direct demand functions can be estimated and, together with standard link performance functions and path enumeration, can provide the desired flows ( i. e., traffic volumes on roadway segments represented by links in the modeling network). For any realistic regional application, an alternative model is required due to the complexity of the network. The UTMS was developed to deal with this complexity by formulating the process as a sequential four- step model.
First, in Trip Generation, measures of trip frequency are developed providing the propensity to travel for different reasons or purposes. Trips are represented as trip ends: the production trip end and the attraction trip end are estimated separately but their totals must eventually match.
Assessment of Local Models and Tools for Analyzing Smart- growth Strategies Page 3- 14
Final Report
Second, in Trip Distribution, the trip productions are distributed across the trip attractions whereby each trip production is matched to a trip attraction. The distribution ( or linkage) of the productions to attractions is modeled using empirically obtained travel impedance relationships ( connecting the likelihood of making a trip to the travel time and/ or cost associated with the trip). The result is a set of trip tables ( person- trips or vehicle- trips, depending on the model) that satisfy the demand for travel given travel options and costs.
Third, in Mode Choice, logit mode choice models developed and calibrated from household survey data are used to determine trip mode ( i. e. drive alone, carpool, transit, bicycle or walk). These calibrated model parameters are assumed to hold constant over time – that is, the same model parameters are used in both the existing conditions models and in the 20 and 30- year horizon models. However, in many of the locally developed travel demand models, the trip tables are essentially factored ( using the mode split and auto occupancy factors from a regional model, if one is available) to reflect relative proportions of trips by alternative modes.
Fourth, in Route Choice, modal trip tables are assigned to mode- specific networks ( if provided in the model) incrementally or via a multi- iteration equilibrium assignment scheme.
The time dimension ( time- of- day) is typically introduced after trip distribution or mode choice where the production- attraction tables are factored to reflect observed distributions of trips in defined periods ( such as the AM or PM travel peaks). Performance characteristics of the transportation system are first introduced in route choice and so UTMS in its most basic form only equilibrates route choices. Total " demand" as specified through generation, distribution, mode choice, and time- of- day models, is fixed with only the route decision to be determined. Many applications of UTMS now include feedback of equilibrated link travel times from route choice to the mode choice and/ or trip distribution models for a second pass ( and occasionally more) through the last three steps, but no formal convergence of the travel times used in the different steps is guaranteed in most applications. Because integrated activity- location procedures ( combined land- use and transportation models) are absent in most U. S. applications, the future activity system is forecast independently with no feedback from the UTMS.
The UTMS has significant data demands in addition to those required to define the activity and transportation systems. The primary need is data that defines travel behavior, and this is gathered via a variety of survey efforts. Household travel surveys with travel/ activity diaries provide much of the data that is required to calibrate the UTMS. These data and observed traffic studies ( counts and speeds) provide much of the data needed for model calibration and validation.
Household travel surveys provide:
• household and person- level socio- economic data ( typically including income and the number of household members, workers, and cars);
Assessment of Local Models and Tools for Analyzing Smart- growth Strategies Page 3- 15
Final Report
• activity/ travel data ( typically including activity type, location, start time, and duration and, if travel was involved, mode, departure time, and arrival time for each activity performed over a 24- hour period); and
• household vehicle- ownership data.
The survey data are used to validate the sample's ability to represent the resident population, to develop and estimate trip generation, trip distribution, and mode choice, and time- of- travel models.
3.2 Representation of the Traveler/ Decision Maker and the Unit of Travel
3.2.1 General Approach
UTMS applications generally use aggregate characteristics for populations within a Traffic Analysis Zone ( TAZ) rather than the characteristics for actual decision- making units, such as an individual or a household. As a result, the travel choice behavior represented in a UTMS model must be based on correlation between observed aggregate travel patterns and average characteristics for the aggregated population within a zone. While this method has proven to be an efficient method for developing approximate forecasts of travel activity for a large area, it has limited the ability of models to represent the influence of how individual or household characteristics can influence travel choices or how different individuals or households within a zone would be influenced by differences in the nature of the transportation system or land- use within the various parts of the zone.
UTMS is also designed to predict the decisions about travel on the basis of a trip, with each trip independent of any other. This method works fairly well for trips that are simple round trips from one zone to another and back, but does not work well for trips that are part of a tour that includes multiple stops.
3.2.2 Common Limitations and Improvement Options
Aggregation of zonal characteristics
The loss of sensitivity brought on by aggregation of the characteristics of the population within a zone is particularly troublesome when there are non- linear relationships between traveler characteristics and how the traveling populations respond to characteristics of the transportation system. This non- linearity is common in how income affects travelers’ responses to changes in travel costs.
Assessment of Local Models and Tools for Analyzing Smart- growth Strategies Page 3- 16
Final Report
Numerous efforts have been made to reduce the biases that are introduced by the aggregation of decision makers into zones. Sample enumeration is one method for “ synthesizing” households in a zone based on the aggregate characteristics and then predicting travel behavior for each of these synthesized households. The results are then aggregated after the forecasts are produced. This avoids the bias introduced by non- linearity, and by representing all travelers in a TAZ as a homogenous group ( e. g., all having the same value of time, and the same propensity toward walking versus driving). The Metropolitan Transportation Commission of the San Francisco Bay Area ( MTC) and the Sacramento Area Council of Governments ( SACOG) use stratification of households by household characteristics including income, number of autos owned and number of workers. MTC has also used sample enumeration as a technique for simulation of individual households based on aggregate zonal characteristics. The newly developed SacSim model, which is designed to work with I- PLACE3S, is the first synthetic population generator that reproduces the resident population at a fine parcel level of spatial resolution.
Trip- based methods do not recognize the linkage between trips
Travelers may often combine a variety of purposes into a sequence of trips as they run errands and link together activities. This is called trip chaining and is a complex process. The standard UTMS trip- based modeling process treats such trip combinations in a very limited way. For example, non- home- based trips are calculated based only on employment characteristics of zones and do not consider how members of a household coordinate their errands. Because many of the smart- growth concepts are designed to group activities so that multiple functions ( work, daycare, shopping, dry cleaning, workout, etc.) can be satisfied in single tour rather than multiple trips, the deficiency inherent in the trip- based method of the UTMS makes analysis of smart- growth strategies difficult, at best.
Travel models are now being developed that consider the activities that a household typically undertakes during a day and then predict “ tours” to achieve the desired activities. These activity- or tour- based models provide greater sensitivity to strategies that encourage trip chaining or satisfying multiple activity goals in a single location. For example, activity- based models have been developed by the San Francisco County Transportation Authority ( SFCTA). MTC and the Southern California Association of Governments ( SCAG) have recently embarked on the development of activity- based models. One of the most complete and sophisticated tour- based models that incorporates synthetic population generation is the " SacSim" model currently being developed for the Sacramento Area Council of Governments ( SACOG). The SacSim model also targets smart- growth and transit policies.
Assessment of Local Models and Tools for Analyzing Smart- growth Strategies Page 3- 17
Final Report
3.3 Representation of Land- uses
3.3.1 General Approach
Before travel demand forecasts are made, it is necessary to develop forecasts of future population and/ or households, economic activity, and land- uses. Forecasted transportation demand is directly linked to projected land- uses. Trips are assumed to follow future land- use patterns; if land- use forecasts are changed, travel demand and travel patterns will likewise change. Local land- use plans, however, typically only project to 10 years, while regional transportation plans are required to project out 20 years. As a result, there is often at least a ten- year period for which transportation planning is not linked to local land- use planning. In the absence of local land- use plans for the period, regional agencies develop land- use forecasts based on extrapolation of development and economic trends.
Planning agencies may prepare study area population and/ or household forecasts, or they may rely on forecasts prepared by others ( such as a state or regional agency). Forecasts of economic activity ( commercial development) are done in conjunction with the population forecasts, since the two are highly interrelated. Subsequently, population and economic growth have to be distributed to different locations in order to conduct travel forecasts because it is necessary to know where people will live, work, shop and go to school in the future to estimate future trip- making.
Land- use plans prepared by cities and counties establish quantities, types, amounts, and locations of land for various uses to meet projections of population and employment as part of the General Plan and Specific Plan development processes. These plans are then also reflected in regional travel demand forecasts. Alternative plans can be developed to reflect different goals, land- use policies and assumptions. For example, land- use plans could be developed to continue current trends; to reduce low- density urban development; or to concentrate development along major corridors, in satellite communities, or in undeveloped portions of existing urban areas. Different assumptions could be made regarding the extent to which environmentally sensitive areas and prime agricultural land will be protected.
Once the quantities and types of land are estimated for the future, those uses must be allocated to specific locations for transportation modeling. A regional allocation is important since local communities often overestimate their growth. For example, individual community zoning often allocates far more commercial and industrial land- use than may actually be demanded when examined from a regional marketplace perspective. Regional allocation addresses situations in which co