Directionality Influences in Spatial Processes by JD Hunt, University of Calgary M Thériault,...
-
Upload
emerald-todd -
Category
Documents
-
view
217 -
download
2
Transcript of Directionality Influences in Spatial Processes by JD Hunt, University of Calgary M Thériault,...
Directionality Influences in Spatial Processes
by
JD Hunt, University of CalgaryM Thériault, Université LavalP Villeneuve, Université Laval
PROCESSUS Second International Colloquium
Toronto ON, CanadaJune 2005
Overview• Introduction
• Context• Motivations• Approach
• Evidence• Disaggregate Observations of Choice
Behaviour• Aggregate Patterns of System Behaviour
• Conclusions• Modelling• Expectations regarding urban form• Planning and design
Introduction• Context
• Modelling spatial decisions, representation relative location
• Travel time components• Travel cost components• Comfort and convenience
• What of ‘directionality’?• ‘together’, ‘on the way’ rather than ‘out of the way’• Anchored relative to reference locations: work, CBD, etc• Nature of perception beyond times and costs• System reinforcing directional tendencies
• Motivation• Adding representation of directionality – simple
form• Seeing in results in various forms• Improved understanding• Higher fidelity• Increased accuracy?• Faster processing
Introduction
• Approach• Draw on previous work in several locations
• Evidence of directionality effects• Representation of these effects
• Disaggregate behavioural evidence• Parking location choice in Edmonton• Commercial vehicle stop location choice in Calgary• Intermediate shopping stop choice in Quebec City
• Aggregate system behavioural evidence• Trip durations in Quebec City• Historical development patterns in Quebec City• Recent employment dynamics in Montréal
Parking Location ChoiceEdmonton
• 1983: Hunt• Develop mode choice model for regional
travel demand model• Include parking location choice for CBD-
destined auto driver alternative• composite utility for parking• parking demand allocation• support mode and parking policy analysis
Data
• 1983 Morning Commuter Survey• 80 employers (business & government
establishments)• 1702 travellers• 468 drivers, selecting parking locations
• Employer parking policy regarding travellers• 124 publicly available off-street parking facilities• Unmetered on-street parking areas, aggregated
into 12 areas
1 km
107 Ave
104 Ave
Jasper Ave
97 S
t
N
Jasp
er A
ve
10
5 S
t
10
1 S
t
109
St
A
I
J
C
H
E
F
G
B
D
K
L
area of on-street parking
CBD boundary
Legend:
Parking Location Choice Model
attributes in utility functions:• Walking distance to final destination• Parking charge per day• Number of stalls• Surface treatment – if paved or not• Adjacent land use – if residential or not• Security• Cleanliness• Angle relative to CBD and home, ANG
Parking Location Choice Model Nesting Structure
employerarranged
on-street off-street
individualoff-streetlocations
individualon-streetlocations
…. ….
ANG Measurehome
workplace
parkinglocation
ANG
ANG Measurehome
workplace
parkinglocation
ANG
and ANG = 90° when• Walk distance < 450 m for off-street• Walk distance < 700 m for on-street
Parking Location ChoiceEdmonton
Parking Location Choice Model Coefficients t-ratio < 1.96 ; 1.96 < t-ratio < 3.29 Choice Level Money
Cost Walk
Distance Res land use
Number Stalls
Surface Type
ANG On-street Nest
Off-street Nest
On-street Constant
Emp Arranged Constant
Money Cost for
Emp Arranged
Walk Distance for Emp
Arranged Individual on-street locations
-.01300 -4.12 -.02760
Individual off-street locations
-1.18 -.00955 .00313 1.31 -.00678
Parking types 0.352 0.380 2.18 3.85 -.762 -.000494
Parking Location ChoiceEdmonton
Parking Location Choice Model Coefficients t-ratio < 1.96 ; 1.96 < t-ratio < 3.29 Choice Level Money
Cost Walk
Distance Res land use
Number Stalls
Surface Type
ANG On-street Nest
Off-street Nest
On-street Constant
Emp Arranged Constant
Money Cost for
Emp Arranged
Walk Distance for Emp
Arranged Individual on-street locations
-.01300 -4.12 -.02760
Individual off-street locations
-1.18 -.00955 .00313 1.31 -.00678
Parking types 0.352 0.380 2.18 3.85 -.762 -.000494
Parking Location ChoiceEdmonton
• Findings:• ANG appears to have influence• driving time differences not included,
potential for bias• consider in future location choice modelling
Commercial Stop Location ChoiceCalgary
commercial vehicle movements
• 2005: Hunt, Stefan, McMillan, Abraham, et al
• Develop model of ehicles operated for commercial purposes• As opposed to household, personal
movements• Includes ‘non-commercial’ non-household
purposes (government, not-for-profit)• Comprise 10-15% of total urban traffic
Commercial Vehicle Movements
• Vehicles operated for commercial purposes• As opposed to household, personal
movements• Includes ‘non-commercial’ non-household
purposes (government, not-for-profit)• Comprise 10-15% of total urban traffic
Some Examples
Commercial• Hauling freight for
a company• Service workers
visiting clients• Sales meetings• Mail• Delivering parcels
Personal• Travel to work• Travel to school• Shopping• Leisure trips• Social visits
Data
• 2001 Commercial Movement Study • All commercial movements
• Not just freight• Not just trucks
• 3,100 establishments in Calgary• 4,300 establishments in Edmonton• 24 hour stop diary• Firmographics
• Employment structure• Vehicle fleet
Next Stop Location• Assigns location for each subsequent, non-
establishment stop on each tour in list• by 13 commercial model segments
(industry, vehicle and tour purpose categories)
• Monte Carlo, probabilities based on Logit• Single-level Logit among locations (zones)
for next stop, total of 1,447 zones in model
Next Stop Location Choice Model
attributes in utility function:• Travel gen cost to potential stop location• Travel gen cost for return to establishment
from potential stop• Population and employment accessibilities• Land use coefficients (5 land uses)• Average income for households at potential
stop• Population and employment size terms• Enclosed Angle
Angle Measure
CURRENTSTOP
ESTABLISHMENT
NEXTSTOP
ANGLE
CURRENTSTOP
ESTABLISHMENT
NEXTSTOP
ANGLE
Angle Measure
CURRENTSTOP
ESTABLISHMENT
NEXTSTOP
ANGLE
CURRENTSTOP
ESTABLISHMENT
NEXTSTOP
ANGLE
and ANGLE = 0 when• starting tour• NEXT STOP = CURRENT STOP
Commercial Stop Location ChoiceCalgary
Next Stop Location Choice Model Coefficients t-ratio < 1.96 ; 1.96 < t-ratio < 3.29 Firm-Tour-Veh
Types Low Den land use
CommRet land use
Ind land use
EmpNode land use
AveInc
(10-6)
GenCost travel to
next stop
GenCost return to
base
Pop Access (10-6)
Emp Access (10-6)
Enclosed Angle
(10-3)
Size Term
Emp Size Term ratio
All-Other-LMH -.7902 .0270 -.1595 -.6126 -11.490 .3039 .1310 -7.651 -9.696 -2.346 .2800 6.779 PS-S-L -.0898 -.2755 .2152 -.4623 1.676 .3283 -10.83 -2.653 -1.884 .3094 .087 PS-S-MH .7250 -.1057 .4655 -.7546 9.476 .0848 .1229 -44.65 9.296 3.684 .2219 PS-G-LMH -.3327 .5674 .4926 .2062 .5688 5.717 -16.54 -6.348 .1588 R-S-LMH -.9676 -.2310 .1547 -.5132 .3601 .03662 -17.35 0 -1.241 .2841 .633 R-G-LMH -.1707 -.0256 .8014 -.1840 .3734 .09158 -13.32 -1.682 1.914 .2067 1.633 I-S-L -1.144 -.2361 .0503 -.4182 .2869 -24.98 5.477 -3.067 .2371 1.231 I-S-MH -.3231 .2789 -.8438 .1627 .1279 -13.25 -16.96 2.934 .1205 1.012 I-G-LMH -.1497 .5575 -.2042 .2581 .09615 -10.68 -5.139 -2.146 .2722 W-S-LMH -.9340 -.2130 .1440 -.4410 2.367 .3849 .04300 -11.81 -27.71 1.761 .2426 2.138 W-G-L -.6668 .9271 -.2688 .4495 .1075 -32.84 -67.74 .892 .2248 W-G-MH -.1226 .1445 -.1183 .3123 .03430 -31.84 5.950 -1.431 .3021 2.313 T-X-LMH -.5279 .1004 .6275 .0267 -4.691 .3792 -11.72 -5.984 3.109 .0087
Commercial Stop Location ChoiceCalgary
Next Stop Location Choice Model Coefficients t-ratio < 1.96 ; 1.96 < t-ratio < 3.29 Firm-Tour-Veh
Types Low Den land use
CommRet land use
Ind land use
EmpNode land use
AveInc
(10-6)
GenCost travel to
next stop
GenCost return to
base
Pop Access (10-6)
Emp Access (10-6)
Enclosed Angle
(10-3)
Size Term
Emp Size Term ratio
All-Other-LMH -.7902 .0270 -.1595 -.6126 -11.490 .3039 .1310 -7.651 -9.696 -2.346 .2800 6.779 PS-S-L -.0898 -.2755 .2152 -.4623 1.676 .3283 -10.83 -2.653 -1.884 .3094 .087 PS-S-MH .7250 -.1057 .4655 -.7546 9.476 .0848 .1229 -44.65 9.296 3.684 .2219 PS-G-LMH -.3327 .5674 .4926 .2062 .5688 5.717 -16.54 -6.348 .1588 R-S-LMH -.9676 -.2310 .1547 -.5132 .3601 .03662 -17.35 0 -1.241 .2841 .633 R-G-LMH -.1707 -.0256 .8014 -.1840 .3734 .09158 -13.32 -1.682 1.914 .2067 1.633 I-S-L -1.144 -.2361 .0503 -.4182 .2869 -24.98 5.477 -3.067 .2371 1.231 I-S-MH -.3231 .2789 -.8438 .1627 .1279 -13.25 -16.96 2.934 .1205 1.012 I-G-LMH -.1497 .5575 -.2042 .2581 .09615 -10.68 -5.139 -2.146 .2722 W-S-LMH -.9340 -.2130 .1440 -.4410 2.367 .3849 .04300 -11.81 -27.71 1.761 .2426 2.138 W-G-L -.6668 .9271 -.2688 .4495 .1075 -32.84 -67.74 .892 .2248 W-G-MH -.1226 .1445 -.1183 .3123 .03430 -31.84 5.950 -1.431 .3021 2.313 T-X-LMH -.5279 .1004 .6275 .0267 -4.691 .3792 -11.72 -5.984 3.109 .0087
Commercial Stop Location ChoiceCalgary
base
1
2
34
5
+ve
base
1
2
3
4
5
-ve
• Findings:• Enclosed angle has strong influence• Sign changes for different segments
• Displaying different spatial patterns
• Driving gen cost differences included for next location and base location
• Stronger case for directionality influence• Include in future location choice modelling• Expect to find aggregate impacts
Commercial Stop Location ChoiceCalgary
HBW Intermediate Shopping Choice
Quebec City• 2005: Thériault• Model decision to make an intermediate
shopping stop on trip from work• Consider influence of locations relative to
work - home axis• ‘on the way’ vs ‘out of the way’
from work to home
Data
• 2001 OD Survey• 29,249 workers with fixed workplace
(not working at home)• 825 intermediate shopping stops made
during HBW trips• 323 to large store• 222 to small shop• 270 to grocery
Stop to Shop Choice Model
attributes in utility function:
• Gender• Age• Household size• Household auto ownership• Distance from home to central axis (Grande
Allee)• Distances from home to workplace
• Straight-line• North-south and east-west components separately
distances from home to work with directionality
components
workplace
home
east-westcomponent
distanceX-axis
N
straight-li
ne
distance
nort
h-so
uth
com
pone
ntdi
stan
ceY
-axi
s
Stop to Shop ChoiceQuebec City
Variables in the Equation
,806 ,076 113,565 1 ,000 2,240
,739 ,142 27,085 1 ,000 2,093
-,196 ,035 31,955 1 ,000 ,822
-,108 ,052 4,331 1 ,037 ,897
,013 ,005 5,931 1 ,015 1,013
,076 ,031 5,998 1 ,014 1,079
-6,269 ,561 125,013 1 ,000 ,002
Woman versus Man
Natural logarithm of Age (Years)
Number of persons in the household
Number of cars owned by the household
Euclidean distance from home to central axis (Km)
Log of Euclidean distance from home to Workplace (Km)
Constant
Step1
B S.E. Wald df Sig. Exp(B)
Stop to Shop ChoiceQuebec City
Variables in the Equation
,806 ,076 113,565 1 ,000 2,240
,739 ,142 27,085 1 ,000 2,093
-,196 ,035 31,955 1 ,000 ,822
-,108 ,052 4,331 1 ,037 ,897
,013 ,005 5,931 1 ,015 1,013
,076 ,031 5,998 1 ,014 1,079
-6,269 ,561 125,013 1 ,000 ,002
Woman versus Man
Natural logarithm of Age (Years)
Number of persons in the household
Number of cars owned by the household
Euclidean distance from home to central axis (Km)
Log of Euclidean distance from home to Workplace (Km)
Constant
Step1
B S.E. Wald df Sig. Exp(B)
Stop to Shop ChoiceQuebec City
Variables in the Equation
,791 ,076 109,399 1 ,000 2,207
,728 ,142 26,273 1 ,000 2,072
-,199 ,035 32,737 1 ,000 ,820
-,103 ,052 3,899 1 ,048 ,903
,017 ,005 9,371 1 ,002 1,017
-,014 ,005 7,182 1 ,007 ,986
,095 ,028 11,172 1 ,001 1,100
-6,143 ,559 120,575 1 ,000 ,002
Woman versus Man
Natural logarithm of Age (Years)
Number of persons in the household
Number of cars owned by the household
Euclidean distance from home to central axis (Km)
Absolute distance between Home and Workplace on X axis of the Map (Km)
Log of Absolute distance between Home and Workplace on Y axis of the Map (Km)
Constant
Step1
a
B S.E. Wald df Sig. Exp(B)
Variable(s) entered on step 1: GENDER, LNAGE, NBPERS, NBAUTO, DEUCREAX, ADX, LNADY.a.
Stop to Shop ChoiceQuebec City
Variables in the Equation
,791 ,076 109,399 1 ,000 2,207
,728 ,142 26,273 1 ,000 2,072
-,199 ,035 32,737 1 ,000 ,820
-,103 ,052 3,899 1 ,048 ,903
,017 ,005 9,371 1 ,002 1,017
-,014 ,005 7,182 1 ,007 ,986
,095 ,028 11,172 1 ,001 1,100
-6,143 ,559 120,575 1 ,000 ,002
Woman versus Man
Natural logarithm of Age (Years)
Number of persons in the household
Number of cars owned by the household
Euclidean distance from home to central axis (Km)
Absolute distance between Home and Workplace on X axis of the Map (Km)
Log of Absolute distance between Home and Workplace on Y axis of the Map (Km)
Constant
Step1
a
B S.E. Wald df Sig. Exp(B)
Variable(s) entered on step 1: GENDER, LNAGE, NBPERS, NBAUTO, DEUCREAX, ADX, LNADY.a.
• Findings:• Home location relative to central axis has +ve
impact – more chaining of shopping with work travel when home location has relatively less nearby
• Home to work distance has +ve impact – more ‘on the way’ intermediate opportunities
• Directionality components of home to work distance have different impacts
• Y-Axis (N-S) +ve linear impact• X-Axis (E-W) –ve logrithmic impact• Perhaps related to highway network, with more high-speed
capacity N-S
• Travel times & costs not included, potential for bias?
HBW Intermediate Shopping Choice
Quebec City
Trip Duration InfluencesQuebec City
• 2003: Vandermissen, Villeneuve and Thériault• Examine how trip duration is influenced by
alignment of trip origin and destination with CBD• Hypothesis 1: Trip duration will decrease as
alignment of origin and destination with CBD increases
• Hypothesis 2: Influence of alignment with CBD on trip duration will decrease as city becomes less monocentric
• ‘directionality’ here is relative to CBD
Data• 1991 OD Survey (n=29,046); 2001 OD Survey
(n=46,664)
• Congested network travel times - from model• Trip purposes
• Work• Study• Shopping• Leisure• Other
• Traveller characteristics• Gender• Age
• Trip Characteristics• Mode• Time of Travel (peak vs off-peak)• Distance from origin to CBD• Distance from destination to CBD
Measure of DirectionalityCBD
origin
destination
Path of trip in street network
Dperpen
Dperpen is the length of theperpendicular between thedestination point and the locusof the straight line passing throughthe origin of the trip and the CBD
Measure of DirectionalityCBD
origin
destination
Path of trip in street network
Dperpen
Dperpen is the length of theperpendicular between thedestination point and the locusof the straight line passing throughthe origin of the trip and the CBD
Origin and destination are aligned with CBD when Dperpen = 0;
As Dperpen increases alignment decreases
Measure of DirectionalityCBD
origin
destination
Path of trip in street network
Dperpen
Hypothesis 1: Trip duration increases as Dperpen increases
Dperpen is the length of theperpendicular between thedestination point and the locusof the straight line passing throughthe origin of the trip and the CBD
Origin and destination are aligned with CBD when Dperpen = 0;
As Dperpen increases alignment decreases
Trip Duration Model
independent variables in regression:
• Gender• Age• Mode• Time of travel (peak vs off-peak)• Distance from origin to CBD• Distance from destination to CBD• Dperpen
• Measure of directionality of O-D relative to CBD• Increase in Dperpen less aligned with CBD; increase in trip
duration• +ve coefficient
1991 2001 1991 2001 1991 2001 1991 2001 1991 2001 1991 2001
adjR2 0.518 0.493 0.586 0.608 0.540 0.523 0.529 0.452 0.493 0.448 0.562 0.549
Sex 0,041 0.029 0.034 0.026 0.024 0.029Age -0.016 -0.065 -0.039 -0.070 -0.083 -0.056 -0.022 -0.065
Mode 0.461 0.382 0.648 0.602 0.612 0.377 0.541 0.430 0.493 0.248 0.543 0.414Peak 0.172 0.218 0.050 0.064 0.065 0.095 0.021 0.093 0.093 0.126 0.135 0.214
Dorigcen 0.363 0.342 0.302 0.394 0.315 0.256 0.367 0.289 0.262 0.212 0.335 0.316
Ddestcen -0.398 -0.359 -0.164 -0.113 -0.247 -0.210 -0.330 -0.336 -0.276 -0.301 -0.328 -0.315
Dperpen 0.509 0.545 0.408 0.400 0.442 0.601 0.467 0.524 0.542 0.621 0.459 0.508
N 14031 17730 3402 3067 3869 5402 1214 2987 6530 17478 29046 46664
Dependent variable : ln of trip duration Directionality is measured through Dperpen
Only coefficients significant at the 0,001 level are shown Dorigcen : Distance between origin of trip and CBD
Return home trips are not included Ddestcen : Distance between dest. of trip and CBCSex: male = 1; female = 0 Age in yearsMode: car driver = 0; bus rider = 1 Peak: 7h00-9h00&16h00-18h00 = 1; other = 0
Standardized regression coefficients (betas)
Isolating the effect of directionality on trip duration
Work Study Shopping Leasure Other Total
Trip Duration InfluencesQuebec City
1991 2001 1991 2001 1991 2001 1991 2001 1991 2001 1991 2001
adjR2 0.518 0.493 0.586 0.608 0.540 0.523 0.529 0.452 0.493 0.448 0.562 0.549
Sex 0,041 0.029 0.034 0.026 0.024 0.029Age -0.016 -0.065 -0.039 -0.070 -0.083 -0.056 -0.022 -0.065
Mode 0.461 0.382 0.648 0.602 0.612 0.377 0.541 0.430 0.493 0.248 0.543 0.414Peak 0.172 0.218 0.050 0.064 0.065 0.095 0.021 0.093 0.093 0.126 0.135 0.214
Dorigcen 0.363 0.342 0.302 0.394 0.315 0.256 0.367 0.289 0.262 0.212 0.335 0.316
Ddestcen -0.398 -0.359 -0.164 -0.113 -0.247 -0.210 -0.330 -0.336 -0.276 -0.301 -0.328 -0.315
Dperpen 0.509 0.545 0.408 0.400 0.442 0.601 0.467 0.524 0.542 0.621 0.459 0.508
N 14031 17730 3402 3067 3869 5402 1214 2987 6530 17478 29046 46664
Dependent variable : ln of trip duration Directionality is measured through Dperpen
Only coefficients significant at the 0,001 level are shown Dorigcen : Distance between origin of trip and CBD
Return home trips are not included Ddestcen : Distance between dest. of trip and CBCSex: male = 1; female = 0 Age in yearsMode: car driver = 0; bus rider = 1 Peak: 7h00-9h00&16h00-18h00 = 1; other = 0
Standardized regression coefficients (betas)
Isolating the effect of directionality on trip duration
Work Study Shopping Leasure Other Total
Trip Duration InfluencesQuebec City
• Findings:• Trip duration decreases as alignment of origin
and destination with CBD increases • Supports Hypothesis 1• For full range of trip purposes• More of a network effect, supply vs demand
• Combined with other work, this influence is increasing
• Potential for reinforcing directionality aspects of choice behaviour
Trip Duration ModelQuebec City
Historical Development Patterns
Quebec City• 2004: Thériault and Bourel• Examine growth patterns
• Axes of development, impact of ‘on the way’ increasing activity at intermediate locations
• Role of CBD as reference
Data
• History of development• Various points starting in 1830
• Range of modal influences• Rail encourages linear patterns in Quebec City Auto
encourages more radial expansion in all directions with some clustering along high-speed roads
• Configuration of land parcels (normal to river)
0 5 10
kilometres
High DensityUrbanised Areas
1 8301 8801 9201 9451 9611 9711 9781 9852 000
Old QuebecOld QuebecOld QuebecOld QuebecOld QuebecOld QuebecOld QuebecOld QuebecOld Quebec
CharlesbourgCharlesbourgCharlesbourgCharlesbourgCharlesbourgCharlesbourgCharlesbourgCharlesbourgCharlesbourg
BeauportBeauportBeauportBeauportBeauportBeauportBeauportBeauportBeauport
LorettevilleLorettevilleLorettevilleLorettevilleLorettevilleLorettevilleLorettevilleLorettevilleLoretteville
SillerySillerySillerySillerySillerySillerySillerySillerySillery
Sainte-FoySainte-FoySainte-FoySainte-FoySainte-FoySainte-FoySainte-FoySainte-FoySainte-Foy
LévisLévisLévisLévisLévisLévisLévisLévisLévis
CharnyCharnyCharnyCharnyCharnyCharnyCharnyCharnyCharny
Ancienne LoretteAncienne LoretteAncienne LoretteAncienne LoretteAncienne LoretteAncienne LoretteAncienne LoretteAncienne LoretteAncienne Lorette
Historical Development of Quebec City
(1) Early 19th CenturyOld City Core and Villages
(2) Mid 19th CenturyDevelopment of “Faubourgs”
(3) Turn of 20th CenturyExtension of “Faubourgs”
(4) 1920 -1945 – Axes appearNew Neighbourhoods
(5) 1945 -1960 – Axes confirmBeginning of urban sprawl
(6) 1960-1975 – Axes consolidatePeak of urban sprawl – Remote towns
(7) 1975 - 2000 – Filling gapsSome extension of axes – Suburbanization
• Findings:• Distinct axes appear
• Radial out from CBD• Before sprawl
• Consistent with linear impact of ‘on the way’ directionality influence
• Also consistent with rail then auto impacts• Parcel orientation also a factor• Difficult to separate influences
Historical Development Patterns
Quebec City
Recent Employment DynamicsMontréal
• 2005: Barbonne• Examine employment patterns 1981 and
2001• Axes of development, impact of ‘on the way’
increasing activity at intermediate locations• Role of CBD as reference
Source: Barbonne,2005
Centrographic analyis suggests elongation of local labour markets
• Findings:• Distinct axes appear
• Radial out from CBD• Before sprawl
• Consistent with linear impact of ‘on the way’ directionality influence
• Following roads oriented out form CBD
• Aggregate emergent behaviour, combining• disaggregate choice behaviour• system supply characteristics
Recent Employment DynamicsMontréal
Conclusions• For modelling:
• There is a directionality component to spatial choice behaviour
• Relevant demand and choice models should include representation of ‘on the way’ vs ‘out of the way’ relative to various reference locations along with usual time and cost elements
• Expectations regarding urban form:• Axes of development apparent
• Planning and design:• Acknowledge potential for CBD-based directionality in
provision of transportation supply and in resulting travel times
• Consider longer term impacts of such directionality working in combination with demand and associated spatial choice behaviour
• Build on existing axes orientation
Further Material
Modelling the effect of directionality on trip duration
• In this section of the paper, directionality is defined with regard to CBD
• Hypothesis 1: Alignment of origin and destination of a trip with CBD should decrease duration
• Hypothesis 2: As cities become less monocentric, the effect of alignment should decrease
• First results, using an angular measure with small travel zones as the georeference, support Hyp 1 & 2 for work trips in Quebec City, comparing 1977 and 1996 (Vandersmissen, Villeneuve, Thériault, 2003).
• Now, we look at trip purposes other than work, using what appears to be a simpler measure of directionality, with a better georeference (6-character postal codes)
Are there planning implications?
• The effect of directionality seems to have increased during de 1990s.
• What happens in the aggregate?• Jobs along corridors• Elongated local labour markets
around the CMA• Elongated urban form, mixed-use and
transit
Lévis
Couronnes Travail Étude Com-loisir Résidence Autre Total Entropie C1 920 0 4649 1176 475 7220 0,443238 C2 346 347 239 1288 241 2461 0,584086 C3 558 768 486 1917 485 4214 0,622878 C4 282 367 469 1490 507 3115 0,609198 C5 321 0 59 901 127 1408 0,422429
Total 2427 1482 5902 6772 1835 18418 0,621981 Québec-Sainte-Foy
Segments Travail Étude Com-loisir Résidence Autre Total Entropie S1 2169 197 3828 1437 1233 8864 0,591087 S2 2082 3409 2606 1061 1522 10680 0,666422 S3 768 448 359 5463 933 7971 0,450315 S4 3110 1234 5050 11469 3049 23912 0,591368 S5 5770 513 3385 7110 2785 19563 0,609972
Total 13899 5801 15228 26540 9522 70990 0,647725
-Lévis: concentric rings around CBD-Qc-Ste_Foy: segments of an urban corridor-Entropy measures diversity within rings and segments-Greater diversity in segments than in rings-Perhaps because of land price gradients?-But, here the segments are much more urbanized than the rings
Does an elongated urban form favour mixed land use?
Concluding questions
• Is the directionality effect the same in the center as it is in the periphery? (center: 0.438 in 1991; 0.523 in 2001) (periphery: 0.434 in 91; 0,486 in
01)
• Are there alignments on other poles than the CBD? (or alignments of trips without poles?)
• Are the examples of Curitiba and Ottawa reproducible?
Acknowledgments to:
• Céline Bourel for computing the trip durations• Rémy Barbonne for analyzing the local labour markets
and mapping the jobs• Simon Faucher for compiling data on poles and corridors• RTC (réseau de transport de la Capitale and MTQ
(ministère des transports du Québec) for giving access to the OD surveys