Class 6: Disaggregate models and networks

1
Class 6 Disaggregate models and networks

• Disaggregate or behavorial models
• Travel diaries
• Jigsaw on modeling
• Networks accessibility and flow
• Transportation and natural hazards

2
Review of aggregate models

• UTMS was the standard method in 1960s
• Trip generation, distribution, modal split,
  route assignment
• Vulnerable to criticisms
• Assumes homogeneity (ecological fallacy)
• Plans for places, not people
• No way to include environmental concerns
• Only commuting trips

3
Disaggregate models background

• 1975 Transportation System Management
• Creating capacity for people, not vehicles
• HOV lanes, paratransit
• Also, social sciences interested in behavior
• But how do you forecast demand for new users and
  modes?

4
Disaggregate models background

• Behavorial (disaggregate) models
• Explain process behind travel decisions
• Can then re-aggregate data
• Can be people- or place-oriented
• Consider potential constraints on movement
• Gender, age, family status
• Income, employment
• Environmental/physical conditions

5
Disaggregate models data collection

• Data collection one disadvantage
• Needs to be tailored to place and purpose
• Time-intensive and expensive
• Methods of collection
• During the trip (on-board, cordon)
• Household surveys (recall or diary)

6
Disaggregate models basic ideas

• Distance decay
• Distance minimization as major motivation
• Tradeoff between distance and utility

7
Logit models

• AKA logistic regression analysis
• Same concept as regression, but variables are in
  categories, not continuous
• Modeling the behavior of individuals, so
  probability comes into play

8
Logit models

• Include observable and random utility aspects
• Used for mode choice, route choice
• Also used to show the effects of a policy
• Criticisms
• Do reported responses match behavior?
• What about constraints on choice?

9
(No Transcript)
10
Activity-based models

• How are travel decisions derived from everyday
  life?
• Demand for activity participation
• Activity scheduling in space and time
• Constraints (personal, space and time, etc.)
• Interactions among decisions
• Interactions among individuals
• Household structure and roles

11
Activity-based models

• Even more data required
• Only possible with computers
• No main example of such a model
• Remember tradeoff between simplicity and
  inclusiveness

12
Transportation and natural hazards

• What is a natural hazard?
• Impact on transportation systems
• Evacuation procedures
• Examples hurricanes, earthquakes, fire, severe
  weather, avalanche, tsunami
• Also relevant for chemical spills, nuclear power
  plants, etc.

13
Impact on transportation systems

• Emergency access
• Network redundancy
• Critical points
• Cost of rebuilding the network
• Impacts to region in the meantime
• Economic losses
• Detours, inaccessibility

14

• North Carolina State Hwy. 12
• Costs 4x the maintainence of other state roads
• 32 million since 1987 (5 mill. from Isabel)

15
(No Transcript)
16
Evacuation and natural hazards

• Regional vs. local scale
• Differences from 4-step model?
• Minimum cost flow or transportation problem
• Aggregate and disaggregate measures
• Consider more than engineering

17
Example wildfire evacuation

• More houses in danger zone transportation tends
  to follow residential
• Evacuation generally managed locally
• What do you want to minimize?
• What variables do you need to consider?

18
Case study Emigration Oaks, UT

• 250-unit development, one access road
• Fire-prone area outside of Salt Lake City
• Second road being considered, but there are
  ecological costs
• Use microsimulator to model evacuation

19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
Case study Conclusions

• Results are place-specific
• Consider time spent in car vs. total evacuation
  time
• Aggregate and disaggregate indicators are
  important