Presentation - UrbanSim

urbansim.org
  • No tags were found...

Presentation - UrbanSim

Land Use and Travel Model Integration

Testing The PSRC Land Use Model Response

to Transportation Strategies

Originally presented at the Innovations in

Travel Modeling Conference, Tempe, AZ, May

10-12, 2010


Presentation Overview

2

• Land Use Model Background

• Integra5on with Travel Model

• Transporta5on Scenarios Tested

• Results

• Future Direc5ons


Land Use Model Background


PSRC Land Use Model - UrbanSim

4

• Micro-­‐simula5on of ac5ons of actors on parcels and buildings:

– Households and Workers

– Jobs

– Developers / Landowners

• Primary Inputs include:

– Allowable development (comp plans)

– Transporta5on system

– Major planned developments (pipeline developments)

– Regional economic forecasts

• Many opera5ng assump5ons:

– Reloca5on rates

– SQFT needed per job by sector

– Construc5on costs

– Vacancy rates

• Simulates each year from 2001-­‐2040


Land Use Model Elements

5

Land

Development

Models

Process Pipeline Events

Real Estate Price Model

Household

Location

Models

Expected Sale Price Model

Development Proposal Choice Model

Building Construction Model

Employment

Location

Models

Household Transition Model

Household Relocation Model

Household Location Choice Model

Employment Transition Model

Employment Relocation Model

Employment Location Choice Model

Workplace

Location

Models

Economic Transition Model

Home-based Job Choice Model

Workplace Location Choice Model

5


Integration With Travel Model


PSRC Analysis Framework

7

Regional Economic

Forecasts

Transport

System

URBANSIM

Land Use

Forecasts

Travel

Forecasts

Air Quality

Analysis

Benefit-Cost

Analysis


Model Handshake – Current Setup

8

Model Inputs and

Integration

Analysis Year

2006 (base) 2015 2025 2035 2040

Land Use Model

Runs, using

accessibilities from:

a previous travel

model run for land

use model run 2006

2006 travel model

for land use model

runs 2007 through

2015

2015 travel model

for land use model

runs 2016 through

2025

2025 travel model

for land use model

runs 2026 through

2035

2035 for land use

model runs 2036

through 2040

Travel Model Runs,

using population and

employment from:

2006 land use

model run

2015 land use

model run

2025 land use

model run

2035 land use

model run

2040 land use

model run


Accessibility Measures – passed to UrbanSim

9

Zone-­‐based, measured to a downtown loca3on

• Generalized Cost to SeaXle CBD, HBW AM SOV

• Generalized Cost to Bellevue CBD, HBW AM SOV

Zone-­‐based

• Average Travel Time, Trip-­‐weighted, AM, SOV, HBW

• Jobs within 30 minutes travel 5me, AM, SOV

Person-­‐based, Home to Work Zones

• Network distance from Home to Work

• Log Sum, HBW AM from Home to Work


Mode Choice Logsum Diagram

Exp U3li3es

from (O-­‐D)

matrices

Exp U3li3es

from (O-­‐D)

matrices

Exp U3li3es

from (O-­‐D)

matrices

Exp U3li3es

from (O-­‐D)

matrices

Sum of Exp U3li3es

Ln(

– all four modes

)

this calcula*on does not yield the sum of the

individual unexponen*ated u*li*es; rather, it yields

an unexponen*ated value that is, in effect, a raw

u*lity that is weighted by mode u*li*es and is thus

reflec*ve of the modes with higher u*li*es

Ln(Sum of Exp

U3li3es)

X

HHs in Car-­‐Worker Category 1 for this Origin Zone

HHs in Car-­‐Worker Category 1 for the All Zones

Car-Worker Group 1(No cars available)

Car-Worker Group 2(2+ Workers, more workers than cars)

Car-Worker Group 3(2+ Workers, at least 1 car per worker)

Car-Worker Group 4(1 Worker, at least one car)

Income Group 1( Under $30K)


Mode Choice Logsum Diagram (Income Segment)

Income Group 1 ( Under $30K)

Income Group 2 ($30-$55K)

Income Group 3 ($55-90)

Income Group 4 ($90K+)


Accessibility – Model Blocks

12


Relative Influence of Variables - WLCM

13


Relative Influence of Variables - HLCM

14


Transportation Scenarios


16

Sensitivity Tests

Base Case Scenario

• Transporta5on Networks (2020, 2040)

• Modest investments in roads and road-­‐based transit

• Near-­‐term voter-­‐approved rail transit extensions

• Very limited tolling (two bridge crossings)

• No real growth in vehicle opera5ng costs

• Modest real growth in parking costs

Alterna3ve Scenarios

• Lower parking costs in selected neighborhoods (zones)

• Higher vehicle opera5ng costs forecast

• Major extensions of rail transit

• Major investments in highway capacity


Alternatives

17

Light Rail

Commuter Rail


Results


Expectations

19

• Short-­‐run subs5tu5on will minimize the magnitude of cost

changes reflected in long-­‐run (loca5on) choices

• Some modest correla5on between a composite measure of

zonal accessibility and the outputs of the land use model

(popula5on, households, employment, work trip loca5ons)

• Higher transporta5on costs should result in lower site values,

and vice versa

• A resor5ng by willingness to pay for sites may dominate the

loca5on choices


Selected Travel Model Statistics

20


VMT

21


Transit Trips

22


Vehicle Trips

23


HBW Average Trip Lengths

24


Changes in Access Costs – AM Productions

25

Access Improvement

• A drop in generalized

costs of auto travel

• Trip weighted average

from each zone to all

other zones


Changes in Access Costs – AM Attractions

26

Access Improvement

• A drop in generalized

costs of auto travel

• Trip weighted average

from each zone to all

other zones


Lower Parking Charges

27

Workplace Loca3on Choice

• Trip aXrac5ons increase in zones with lower parking costs

• Income sensi5vity


Rail Transit Extensions

28

Change in Population and

Employment

red-gains, blue-losses


Increased Highway Capacity

29

Changes in Employment

Change in Population

blue-gains, red-losses


30

Findings

Land Use Response to Transporta3on Scenarios

• A modest response is in line with theore5cal expecta5ons

• Accessibility measures from the travel model do change across

scenarios and reflect route and des5na5on choices (and to a

more limited degree mode choice).

• Short-­‐run subs5tu5on and ac5vity sor5ng across sites likely

limits the effects on development capital

• The influence of access on site values is probably a central

feature in proper simula5ons. We have not explicitly evaluated

site values


31

Some Additional Tests

Influence of Developer Models

• Hypothesis – development dynamics may impose constraints

that limit the influence of accessibility on loca5on choices

• Test 1:

‒ higher threshold vacancy rates for mul5-­‐family

developments

‒ higher redevelopment threshold (improvement value/total

value)

• Test 2: changes as per Test 1 above plus 100% household

reloca5on rates (tested for a single year)


32

Test 1: Highway Example

Highway Capacity Scenario

Highway compared to

Base Case

• Test 1:

‒ higher threshold vacancy rates

for mul5-­‐family developments

‒ higher redevelopment threshold

(improvement value/total value)

• Greater degree of household

response to accessibility


Test 2: Highway Example

33

Highway Capacity Scenario

• Test 2: changes as per Test 1 above

plus 100% household reloca5on rates

(tested for a single year)

• Compared Highway Scenario with

100% household reloca5on rate with

the same scenario with default

loca5on rate – single year analysis

• Greater degree of household

response to accessibility


Future Directions


Future Directions

35

Accessibili3es Variables

• Revisit the zonal composite variables used in the real estate

price and employment loca5on choice models

• Changes to real estate price model to more fully reflect scale of

demand and accessibility

• A revised zone structure (from 938 to over 3,500) should reduce

aggrega5on problems

• Ac5vity-­‐based travel model development will open up

numerous opportuni5es for disaggregate access measures

Revisit Integra3on Structure

• Frequency of travel model runs (currently every 10 forecast

years)

• Ac5vity-­‐based model development will necessitate a different

approach (interac5on between long-­‐run and short-­‐run choices)


Puget Sound Regional Council:

Matthew Kitchen, Chris Johnson,

Peter Caballero, Mark Simonson,

and Stefan Coe

Maren L. Outwater, Resource Systems Group Inc

More magazines by this user
Similar magazines