THE EFFECTS OF URBAN FORM ON WALKING TO TRANSIT Ayşe ÖZBİL 1, John PEPONIS 2 8 th International...

THE EFFECTS OF URBAN FORM ON WALKING TO TRANSIT

Ayşe ÖZBİL1, John PEPONIS2

8th International Space Syntax Symposium 3-6 January, 2012 Santiago, Chile

1 Okan University Department of Architecture, İstanbul 2 GaTech College of Architecture, Atlanta

• introduction

• research background

• analytical framework

• street connectivity and transit-related walking

• conclusions/implications

2

8th International Space Syntax Symposium 3-6 January Santiago, Chile

The Effects of Urban Form on Walking to Transit Ayşe Özbil, John Peponis

• introduction





3



purpose:

the impacts of urban form, in general, and street network, in

particular, on public transport-related walking urban form

. population densities . land-use patterns . street networks

aim:

how far street connectivity is related to transit walk-mode

shares after controlling for socio-demographic attributes and

transit service features

hypothesis:

connected environments supporting different kinds of walking

support public transportation and promote walking as transit

access/egress mode choice

objectives

4



question:

to what extent local conditions of station environments

contribute to an explanation of variations in

transit-access/egress walking shares (number of riders

walking from within a range as a proportion of total ridership)?

contribution:

to better understand how far street network connectivity

influences the choice to walk for transit

implication:

relevant in assessing the sensitivity of transit access/egress

walk mode choice to changes in urban form

contribution

5



• introduction





6



7



source: Belmont 2002

8



source: Quade and Douglas Inc. 1996

source: Belmont 2002

9



RIDERSHIP WALKING for TRANSIT

(transit/auto trips-VMT per person/household & station boardings) ▪ (walking trips per person/household (proportion of trips by transit) ▪ (walk mode share)

10





11



probability commute by walk/bicyclesource: Krizek 2003

many businesses (black dots) indicating high land use mix.

fewer businesses (black dots) indicating lower land use mix.

source: Cervero 1996

probability commute by walk/bicycle

12



13





source: Southworth and Owens 1993

14





source: Cervero and Radisch 1996

traditional neighborhood (TND)

planned unit development (PND)

source: Kulkarni and McNally 1997

traditional neighborhood contemporary neighborhood

15



source: Ozbil and Peponis 2007

R2=0.30; p<0.0003; n=55

lowest global integration

Integration rn

logp

ed/2

0min

highest global integration

axial lines analysis of Virginia Highland according to global integration

correlation between movement counts and global integration in Virginia Highland

16



• introduction





17



MARTA system overlaid on the map of Atlanta. Gray lines represent roads; red lines denote the freeway system.

2001-2002 Regional On-Board

Transit Survey

. trip origin

. final destination

. boarding (bus/train on) location

. alighting (bus/train off) location

. access/egress mode (walk, bus, car, etc.)

. household size

. household income

. auto availability

walk-mode share :

total walk trips / total ridership by station

MARTA rail transit stations

10.13 – 12.09 2001 & February 2002

trip origins

case context & data

18



transit service features (park&ride facilities, service frequency, feeder bus services, station structures)

parcel-based land-use data (based on 2004 data acquired by SMARTRAQ program)

population density (based on 2000 Census Data)

socio-demographics (acquired from transit survey)

street connectivity measures (based on ESRI Streetmap 2003)

case context & data

19



*

. Accessibility:

sidewalk availability measuring the percentage of streets with sidewalk

. Density: population density (people in gross acres) within 1, 0.5, and 0.25 mile radii of stations established using US 2000 census data.

. Land-Use: mixed-use entropy index* computed using parcel-based land-use data. Separate entropy indices computed for each range.

. Transit service characteristics: number of station-area parking supplies; service frequency (# of inbound trains in am peak

hour); availability of feederbus services at station; types of station structure (underground, elevated, at-grade)

. Composite socio-demographic variable:

auto-ownership relativized by per-capita income

. Measures of street connectivity:

k

pp

entropyuseMixed

k

iii

ln

ln

1 1

independent variables

* Cervero and Kockelman (1997); Cervero (2006); Greenwald (2006)

20



Segment-based Measures of Connectivity

definition of line segments and road segments

source: Peponis et al. 2008


21





Metric Reach

definition of Metric Reach


22





Directional Reach

23



definition of Directional Reach


metric reach (1 mile, 0.5 mile, 0.25 mile)

(total linear footage within each

buffer)

2-directional reach (10o) (every

street segment and portion of street segment)

directional distance (average

# of turns needed to access all portions of streets within

metric reach)

composite connectivity measure (metric reach / directional distance, 10o)



24



• introduction





25



sum of squares F ratio prob>F





0.051 8.844 0.006 0.002 0.567 0.458 0.008 2.748 0.109 0.0095 3.0113 0.094 0.0094 2.8784 0.1017

0.023 1.994 0.154 0.016 1.931 0.163 0.020 3.220 0.055 0.0199 3.1619 0.0584 0.020 3.030 0.066

0.002 0.395 0.535 0.003 0.805 0.377 0.031 10.215 0.003 0.032 10.168 0.0036 0.028 8.637 0.007

0.111 19.360 0.000 0.044 10.458 0.003 0.030 9.641 0.004 0.0303 9.6295 0.0045 0.027 8.255 0.008

0.053 9.346 0.005 0.009 2.133 0.155 0.000 0.060 0.809 0.000 0.0061 0.9381 0.000 0.008 0.928

0.048 11.350 0.002 0.046 14.967 0.001 0.044 13.965 0.001 0.039 11.942 0.002

0.037 12.201 0.002 0.029 9.307 0.005 0.018 5.512 0.027

0.001 0.338 0.566 0.001 0.304 0.586

0.000 0.004 0.949

N=37

R 2

R2 adjusted

Notes:

coefficient of residual variability,V e

std. error,S e 0.08 0.07 0.06

0.62 0.72 0.80

+ Density

0.06 0.06

0.79 0.78

+ Land Use + Connectivity +Accessibility

auto ownership relativized by per-capita income*

total riders walked / total ridership per station

Controls

station structure typeb

service frequency‡

feederbus services (no)

parking supplies

avg. Reach (1mile)

sidewalk availabilitya

population density: persons per gross acre within 1 mile of station

mixed-land use index†

0.84 0.84

b types of station structure: at-grade, elevated, underground ‡ number of inbound trains in am peak hour (7am-9am)* ratio of average auto-ownership to average per-capita income calculated per station

Numbers in bold = p< 0.05; numbers in italics = p<0.10

a proportion of roads with sidewalk

†

0.69 0.77 0.84

39.17

k

pp

entropyuseMixed

k

iii

ln

ln

1 1

26



F ratio prob>F F ratio prob>F F ratio prob>F F ratio prob>F F ratio prob>F

0.051 8.844 0.006 0.002 0.567 0.458 0.008 2.748 0.109 0.0095 3.0113 0.094 0.0094 2.8784 0.1017

0.023 1.994 0.154 0.016 1.931 0.163 0.020 3.220 0.055 0.0199 3.1619 0.0584 0.020 3.030 0.066

0.002 0.395 0.535 0.003 0.805 0.377 0.031 10.215 0.003 0.032 10.168 0.0036 0.028 8.637 0.007

0.111 19.360 0.000 0.044 10.458 0.003 0.030 9.641 0.004 0.0303 9.6295 0.0045 0.027 8.255 0.008

0.053 9.346 0.005 0.009 2.133 0.155 0.000 0.060 0.809 0.000 0.0061 0.9381 0.000 0.008 0.928

0.048 11.350 0.002 0.046 14.967 0.001 0.044 13.965 0.001 0.039 11.942 0.002

0.037 12.201 0.002 0.029 9.307 0.005 0.018 5.512 0.027

0.001 0.338 0.566 0.001 0.304 0.586

0.000 0.004 0.949

N=37

R2 adjusted

Notes:




†

k

pp

entropyuseMixed

k

iii

ln

ln

1 1

27



B t std β B t std β B t std β

0.15 0.22 4.20

-0.002 -2.471 -0.256 0.049 3.156 0.317 -0.127 -3.480 -0.442

0.062 3.334 0.280 0.404 5.915 0.592 0.027 3.299 0.429

0.773 6.800 0.605 0.014 2.214 0.197 0.025 2.361 0.304

0.016 3.908 0.421N N

N N N

R 2 R 2 R 2

Prob>F Prob>F Prob>F

Notes:

0.00

avg.metric reach(1mile)/ directional distance(10°)

37

0.81

R 2 adjusted 0.79

std. error,S e 0.06

k

pp

entropyuseMixed

k

iii

ln

ln

1 1


0.15 0.22 4.20

-0.002 -2.471 -0.256 0.049 3.156 0.317 -0.127 -3.480 -0.442

0.062 3.334 0.280 0.404 5.915 0.592 0.027 3.299 0.429

0.773 6.800 0.605 0.014 2.214 0.197 0.025 2.361 0.304

0.016 3.908 0.421

N N N


Notes:

Reduced Model

total riders walked within 1 mile / total ridership per station

constant


Feederbus services (no)

avg.metric reach(1mile)/ directional distance(10°)


k

pp

entropyuseMixed

k

iii

ln

ln

1 1


constant -0.936 -0.389 3.906

-0.031 -2.254 -0.214 0.043 2.621 0.278 -0.117 -3.869 -0.408

-0.003 -2.882 -0.324 0.398 5.741 0.583 -0.001 -3.526 -0.495

0.052 2.725 0.235 0.006 2.059 0.202 0.022 3.185 0.355

0.765 6.365 0.599 0.036 5.142 0.751

0.005 3.480 0.457

N N N


Notes:

avg. Reach (1mile)

k

pp

entropyuseMixed

k

iii

ln

ln

1 1


constant -0.936 -0.389 3.906

-0.031 -2.254 -0.214 0.043 2.621 0.278 -0.117 -3.869 -0.408

-0.003 -2.882 -0.324 0.398 5.741 0.583 -0.001 -3.526 -0.495

0.052 2.725 0.235 0.006 2.059 0.202 0.022 3.185 0.355

0.765 6.365 0.599 0.036 5.142 0.751

0.005 3.480 0.457

N N N

R 2 R 2 R 2

R 2 adjusted R 2 adjusted R 2 adjusted


Notes:

std. error,S e

total riders walked / total ridership per station

Reduced Model

0.00

0.79

0.06

0.8337


avg. Reach (1mile)


feederbus services (no)

station structure typeb

(elevated)

k

pp

entropyuseMixed

k

iii

ln

ln

1 1

F ratio prob>F F ratio prob>F F ratio prob>F F ratio prob>F F ratio prob>F

0.051 8.844 0.006 0.002 0.567 0.458 0.008 2.748 0.109 0.0095 3.0113 0.094 0.0094 2.8784 0.1017

0.023 1.994 0.154 0.016 1.931 0.163 0.020 3.220 0.055 0.0199 3.1619 0.0584 0.020 3.030 0.066

0.002 0.395 0.535 0.003 0.805 0.377 0.031 10.215 0.003 0.032 10.168 0.0036 0.028 8.637 0.007

0.111 19.360 0.000 0.044 10.458 0.003 0.030 9.641 0.004 0.0303 9.6295 0.0045 0.027 8.255 0.008

0.053 9.346 0.005 0.009 2.133 0.155 0.000 0.060 0.809 0.000 0.0061 0.9381 0.000 0.008 0.928

0.048 11.350 0.002 0.046 14.967 0.001 0.044 13.965 0.001 0.039 11.942 0.002

0.037 12.201 0.002 0.029 9.307 0.005 0.018 5.512 0.027

0.001 0.338 0.566 0.001 0.304 0.586

0.000 0.004 0.949

N=37

R2 adjusted

Notes:




†

k

pp

entropyuseMixed

k

iii

ln

ln

1 1

28



• introduction





29



confirm hypothesis:

street network design and land-use mix around stations are

significantly associated with walk-mode shares, controlling

for population density, personal attributes, and transit

service characteristics

theoretical implications

30



mixed-use developments increase the odds of walking

to/from transit ; street networks with denser connections

significantly affect walking shares

directional accessibility affects people's willingness and

capacity to walk for transit as significantly as metric

accessibility

takeaway: incorporating measures of street density and

directional accessibility in transit-oriented studies can lead

to enhanced models of urban form & function

takeaway for practice:

guide future efforts to integrate subdivision with zoning

regulations in developing sparse suburban areas towards

dense transit-oriented urban hubs

transit-oriented policies are compatible with policies aimed

at the enhancement of health and the reduction of obesity

through daily physical activity (walking to/from the station)

effective policies developed to encourage new designs with

the option to walk will support more sustainable cities

practical implications

31



thank you.

32



THE EFFECTS OF URBAN FORM ON WALKING TO TRANSIT Ayşe ÖZBİL 1, John PEPONIS 2 8 th International...

Documents

Transcript of THE EFFECTS OF URBAN FORM ON WALKING TO TRANSIT Ayşe ÖZBİL 1, John PEPONIS 2 8 th International...