MORPC Model Comparison Project Trip vs. Tour Model

1
MORPC Model Comparison ProjectTrip vs. Tour Model
2
Research Project

• Led by Ohio DOT and initiated in 2008
• Main objective examine the performance of the
  MORPC trip-based and tour-based frameworks in the
  context of a before-and-after project analysis
• ODOT, MORPC, OKI and NOACA are looking to obtain
  a clearer picture of the potential practical
  benefits of tour-based models in the context of
  assessing projects and policies

3
Research Tasks

1. Understand model differences
2. Determine analysis methodologies and data
   requirements
3. Select study projects for before/after analysis
4. Determine data collection projects
5. Prepare models and model data
6. Run models, analyze output and observed conditions

4
Requirements for an Analogous Comparison

• Common analysis years
• Using 1990, 2000, 2005 (due to better 1990 SE
  data than 1995)
• Identical estimation datasets
• Isolate supply-side differences
• Isolate demand-side differences
• Borrowed a Trip Model from OMS

5
New Trip Model Formulation
6
Estimation Datasets

• Estimate new Trip Generation and Gravity
  Distribution Models with the 1999 HIS
• Trip model will use mostly identical SE data as
  the tour model
• Update mode choice model to use IVT, OVT and wait
  coefficients from tour model
• Other coefficients will be scaled

7
Mode Choice

• Mode choice
• Trip model uses nested logit structure based on
  1993 on-board survey
• Tour model uses mostly multinomial structures
  based on 1999 HIS 1993 on-board survey - Also
  adheres to FTA New Starts parameter guidelines

8
Model Areas
9
Demand-side Differences

• 4-period assignment
• External and CMV models are based on SE data and
  network impedances, so they would change with
  different assignments
• Solution hold trip tables constant across the
  models and alternatives
• Equilibrium assignment closure rates can vary
  mode choice impedances and final highway volumes
• Solution apply very high closure rate to both
  models

10
Validation - VMT
11
Validation - RMSE
12
Other Considerations

• Trip Model is fairly simplistic
• No peak spreading
• No vehicle ownership
• Daily level generation and distribution
• Gravity distribution model
• 1 iteration of feedback to mode choice

13
Proposed Before/After Projects

• Spring-Sandusky interchange
• Large-scale freeway project
• Project is completed and subsequent land-use
  development has stabilized
• Polaris
• Medium-scale freeway interchange project
• New and subsequently modified interchange in
  rapid growth area

14
Spring-Sandusky
15
Spring-Sandusky
16
Polaris - 1988
17
Polaris - 2008
18
Proposed Before/After Projects

• Systemwide transit analysis
• 35 decline in transit service 2001-2005
• Trunk routes virtually unchanged, with suburban
  service reduced
• Hilliard-Rome Road Area
• High growth area, but no substantial
  transportation changes
• Land use changes have now largely subsided
• Control Site IR 71 South of the CBD

19
Traffic Volumes

• Why we care about traffic volumes
• 100-200 projects a year that use the models
  traffic volumes

20
Contact Information

• Rebekah Anderson ODOT
• 614-752-5735
• rebekah.anderson_at_dot.state.oh.us
• Greg Giaimo ODOT
• 614-752-5738
• greg.giaimo_at_dot.state.oh.us
• David Schmitt AECOM
• 614-901-6026
• david.schmitt_at_aecom.com