Measuring the Effects of Land Use on Measuring the Effects of Land Use on Travel Behavior and Climate ChangeTravel Behavior and Climate Change
Jerry Walters, Fehr & PeersJerry Walters, Fehr & Peers
HaagenHaagen--SmitSmit SymposiumSymposium““Challenge to ChangeChallenge to Change””
April 2008April 2008
AgendaAgenda
1.1. Smart Growth and Climate ChangeSmart Growth and Climate Change
2.2. Measuring Effects of Smart Growth on TravelMeasuring Effects of Smart Growth on Travel
3.3. Getting the Models to Get it RightGetting the Models to Get it Right
4.4. Focusing on MultiFocusing on Multi--Modalism and MobilityModalism and Mobility
Growth in COGrowth in CO22 Emissions assuming more Emissions assuming more Stringent Vehicle and Fuel StandardStringent Vehicle and Fuel Standard(45 mpg CAFE in 2030) + ((45 mpg CAFE in 2030) + (--15% Fuel 15% Fuel GHGsGHGs) = (24% above 1990 in 2030)) = (24% above 1990 in 2030)
70%
80%
90%100%
110%
120%
130%
140%150%
160%
170%
2005 2010 2015 2020 2025 2030
2005
= 1
00%
Sources: VM T: EIA with 10% rebound, M PG & Fuel: Trend Ext rapolat io n
CO2
1990 CO2
Fuel GHG
MPG
VMT
Neighborhood comparison: Neighborhood comparison: 2/3rd VMT Reduction2/3rd VMT Reduction
Daily Vehicle Miles per Person vs. Residential DensitySource: Baltimore Metropolitan Council, 2001 Travel Survey
0
10
20
30
40
50
60
0 2 4 6 8 10 12 14 16 18
Households per Acre
Dai
ly V
ehic
le M
iles
per P
erso
n
Charles Street
Hampstead
Odenton
Owings Mills
Dundalk
Reservoir Hill
Butcher's Hill
Brewer's HillBolton Hill
Canton
Federal Hill
Taneytown
Westminster Greens
Westminster Downtown Havre de Grace
Land useLand use--transportation scenario planning transportation scenario planning studies in the U.Sstudies in the U.S (Bartholomew 2007)(Bartholomew 2007)
VMT vs. Density for 62 Planning Scenarios VMT vs. Density for 62 Planning Scenarios Relative to TrendRelative to Trend
Density & VMT
R2 = 0.5575
-35%
-30%
-25%
-20%
-15%
-10%
-5%
0%
5%
10%
-20% -10% 0% 10% 20% 30% 40% 50% 60% 70%
Percent Difference in Density (trend vs. non-trend)
Perc
ent D
iffer
ence
in V
MT
(tren
d vs
. non
-tren
d)
n = 62
Site Design & Location Studies in US and CanadaSite Design & Location Studies in US and Canada
Site Design Studies
Regional Location Studies
Effect on VMT of Placing Development at Higher Density Infill Location
-70%
-60%
-50%
-40%
-30%
-20%
-10%
0%0% 200% 400% 600% 800% 1000% 1200% 1400% 1600%
Increase in Density
Red
uctio
n in
VM
T
% Reduction in Site Density vs % Change in VMT per Capita(density reduction accompanied by relocation of development from infill to greenfield)
SACOG Travel Generation by Density of PlaceSACOG Travel Generation by Density of Place
Transit + Walk + Bike Trips per HH
00.51
1.52
2.53
3.54
0 20 40 60 80 100 120 140
VMT per Household
0
10
20
30
40
50
60
0 20 40 60 80 100 120 140
Jobs + Households within ¼ Mile of Place of Residence
AgendaAgenda
1.1. Smart Growth and Climate ChangeSmart Growth and Climate Change
2.2. Measuring Effects of Smart Growth on TravelMeasuring Effects of Smart Growth on Travel
3.3. Getting the Models to Get it RightGetting the Models to Get it Right
4.4. Focusing on MultiFocusing on Multi--Modalism and Quality of LifeModalism and Quality of Life
Trip generation is directly related to D’s:
DensityDensity dwellings, jobs per acredwellings, jobs per acre
DiversityDiversity mix of housing, jobs, retailmix of housing, jobs, retail
DesignDesign connectivity, walkability connectivity, walkability
DestinationsDestinations regional accessibilityregional accessibility
Distance to TransitDistance to Transit rail proximityrail proximity
Shortens trip lengthsShortens trip lengths
More walking/bikingMore walking/biking
Supports quality transitSupports quality transit
Density (jobs and dwellings per acre)
Links trips, shortens distancesLinks trips, shortens distances
More walking/ bikingMore walking/ biking
Allows shared parkingAllows shared parking
Diversity (mix of housing, jobs, retail)
Design (connectivity, walkability)
Destinations (accessibility to regional activities)
Development at infill or closeDevelopment at infill or close--in locations reduces in locations reduces vehicle trips and milesvehicle trips and miles
Transit shares higher within ¼ mile and ½ mile of station
Distance to Transit
20% to 51%4. Destinations
2% to 13%3. Design
1% to 13%2. Diversity
1% to 17%1. Density
Reductions in VMT per 100% increase in 4D’s
4D4D’’s s (Land Use Clustering, Mixing, Traditional Design)(Land Use Clustering, Mixing, Traditional Design) ––All Reduce TravelAll Reduce Travel
Sources: National Syntheses, Twin Cities, Sacramento, Holtzclaw
55thth D D -- Distance from TransitDistance from Transit
Vehicle-miles traveled, compared with regional average:
• 42% reduction for households within ½ mile of transit
• 21% reduction for households between ½ and 1 mile
Emerging research:Emerging research:Other Other ““DD”” factors that affect VMTfactors that affect VMT
6. Development scale
7. Demographics
8. Demand management• parking management • pricing policies • traveler information • neighborhood electric vehicles
varies8. Demand Management
11% to 23%7. Demographics
15% +/-6. Development Scale
Reduction in VMT per 100% increase
in “D”
Effects of Other Effects of Other ““DD”” FactorsFactors
Source: EPA study on effects of mixed use development – Portland case study
Smart Growth Trip GenerationSmart Growth Trip Generation
National studies of Mixed Use, National studies of Mixed Use, TOD and Infill developmentTOD and Infill development
Statistical analysis, empirical Statistical analysis, empirical validationvalidation
Examples: San Diego, Seattle, Portland, Sacramento, Houston, Atlanta, Boston
36%44%35%Trip Discount
InfillTODMXD
Direct Transit Ridership ModelsDirect Transit Ridership Models
Examples: BART, Caltrain, Sacramento LRT, Salt Lake LRT, Denver RTD
TOD Population
TOD Employment
Catchment Population
Parking Supply
Train Frequency
Feeder Bus Frequency
Development Mix
Walk Connections
Bike Parking
Model 1- Relationship Between PM Peak Boardings and 1/2 mile Non-Retail Employment, 1/2 mile Population, and Downtown SF Indicator, R2=.985
02,0004,0006,0008,000
10,00012,00014,00016,00018,00020,000
PredictedActual
AgendaAgenda
1.1. Smart Growth and Climate ChangeSmart Growth and Climate Change
2.2. Measuring Effects of Smart Growth on TravelMeasuring Effects of Smart Growth on Travel
3.3. Getting the Models to Get it RightGetting the Models to Get it Right
4.4. Focusing on MultiFocusing on Multi--Modalism and QualityModalism and Quality of Lifeof Life
Shortcomings of Conventional Travel Models Shortcomings of Conventional Travel Models in Assessing Smart Growthin Assessing Smart Growth
•• Primary use is to forecast longPrimary use is to forecast long--distance auto travel on distance auto travel on freeways and major roadsfreeways and major roads
•• Secondary use is to forecast systemSecondary use is to forecast system--level transit uselevel transit use
•• ShortShort--distance travel, local roads, nondistance travel, local roads, non--motorized travel motorized travel modes are not addressed in model validationmodes are not addressed in model validation
Levels of Model SophisticationLevels of Model Sophistication
Caltrans Assessment of Local Models and Tools for Analyzing Smart-Growth Strategies, 2007
High-Sensitivity Models
Moderate-Sensitivity Models
Low-Sensitivity Models
Inco
me
Stra
tific
atio
n in
Dis
trib
utio
n an
d M
ode
Cho
ice
Aut
o O
wne
rshi
p M
odel
ing
Sens
itive
to L
and-
Use
Cha
ract
eris
tics
Deg
ree
of S
ensi
tivity
to S
mar
t-Gro
wth
Strat
egie
s
Mod
elin
g M
ode
of M
ultip
le M
odes
of
Acc
ess
to T
rans
it
Dis
trib
utio
n Se
nsiti
ve to
Mul
ti-M
odal
Opt
ions
Dis
aggr
egat
e Si
mul
atio
n of
Hou
seho
lds
Dai
ly V
ehic
le T
rip M
odel
Steps to Improve UTMS Sensitivity to Smart-Growth Strategies
Trav
el T
ime
Feed
back
Non
-Mot
oriz
ed M
odes
in M
ode
Cho
ice
Mod
elin
g Pe
ak a
s w
ell a
s Dai
ly
Trav
el Sim
ple
Mod
e Cho
ice
Tran
sit N
etw
ork
and
Dai
ly
Ass
ignm
ent
Supp
ly a
nd D
eman
d Eq
uilib
ratio
n
Inte
grat
ed L
and-
Use
/Tra
nspo
rtat
ion
Mod
elin
g
Act
ivity
- and
Tou
r-Bas
ed M
odel
ing
Expl
icit
Rep
rese
ntat
ion
of
Pede
stria
n an
d Bic
ycle
Net
wor
ks
Typical Model Typical Model ““Blind SpotsBlind Spots””
• Abstract consideration of distances between land uses within a given TAZ or among neighboring TAZ’s
• Limited or no consideration intra-zonal or neighbor-zone transit connections
Network in ModelNetwork in Model Network in FieldNetwork in Field
Typical Model Typical Model ““Blind SpotsBlind Spots””
•• Sidewalk completeness, route directness, block Sidewalk completeness, route directness, block size generally not considered.size generally not considered.
Typical Model Typical Model ““Blind SpotsBlind Spots””
•• Little consideration is given to spatial relationship Little consideration is given to spatial relationship between land uses within a given TAZ (density)between land uses within a given TAZ (density)
•• Interactions between different nonInteractions between different non--residential land residential land uses (e.g. offices and restaurants) not well uses (e.g. offices and restaurants) not well representedrepresented
Conventional Ridership Modeling
Screen for TravelersScreen for Origin / DestinationScreen for Mode
Law of Small Numbers
FTA Report on Conventional Forecasting
•“… ridership projections for New Starts are often highly inaccurate in terms of both total ridership and the characteristics of the markets that are actually served.”
Caltrans Study RecommendationCaltrans Study Recommendation
Source: Assessment of Local Models and Tools for Analyzing Smart-Growth Strategies,2007
4DElasticities
4DPost Processor
PLACE3S
INDEX
Research Results
Planning ToolsUse 4D’s to compensate for any lack of sensitivity in travel models
AgendaAgenda
1.1. Smart Growth and Climate ChangeSmart Growth and Climate Change
2.2. Measuring Effects of Smart Growth on TravelMeasuring Effects of Smart Growth on Travel
3.3. Getting the Models to Get it RightGetting the Models to Get it Right
4.4. Focusing on MultiFocusing on Multi--Modalism and MobilityModalism and Mobility
Traffic LOS Traffic LOS Person MobilityPerson Mobility
• Person accessibility and safety
• Travel time mobility for all modes
• Comfort and convenience for all users
Van Ness Ave BRT Alternatives
Source: Van Ness BRT Feasibility Study, Public Workshop, October 19, 2006, San Francisco County Transportation Authority.Alt. 2 reduces total traveler delay by 8% with no increase in vehicle delay.
Alt. 3 increases vehicle delay by 8% but reduces delay for all travelers 5%.
Intersection LOS Improvement StudyAlternative 1 -- Conventional Treatment
Alternative 2 – All Bike/Pedestrian Phase
Alternative 3 – Ped/ Bike Head-Start Phase (balanced LOS for all modes)
AgendaAgenda
1.1. Smart Growth and Climate ChangeSmart Growth and Climate Change
2.2. Measuring Effects of Smart Growth on TravelMeasuring Effects of Smart Growth on Travel
3.3. Getting the Models to Get it RightGetting the Models to Get it Right
4.4. Focusing on MultiFocusing on Multi--Modalism and MobilityModalism and Mobility
5.5. Case StudyCase Study
Contra Costa: Shaping Our FutureContra Costa: Shaping Our Future
Integrated Land Use/ Transportation Integrated Land Use/ Transportation Visioning and Planning StrategyVisioning and Planning Strategy
Concentrate land use around Concentrate land use around potential transit nodespotential transit nodes
Prioritize transportation system Prioritize transportation system expansions that work best with expansions that work best with compact, transit oriented compact, transit oriented development.development.
Emphasize development forms known to reduce travel perEmphasize development forms known to reduce travel percapita: density, mix, transitcapita: density, mix, transit--oriented design, infill and oriented design, infill and closeclose--in locationsin locations
ScenarioScenario Overlay Merged Environmental Constraints MapOverlay Merged Environmental Constraints Map
Modeling Future Development ScenariosModeling Future Development Scenarios
Virtual Land Use Virtual Land Use Future, 2030Future, 2030
Available LandAvailable Land
Future Transportation Future Transportation & Land Use Model, & Land Use Model, 20302030
Measurements and Metrics:
•Economic Analysis
•Environmental Impact
•Land Conversion
•Social/Demographic Impacts
•Other Metrics
Transportation Transportation NetworkNetwork
Transportation Transportation PoliciesPolicies
Transportation ModelingLand Use Modeling
Jobs & Population Jobs & Population ForecastForecast
Development Development Policy ScenarioPolicy Scenario
““Vision ScenarioVision Scenario”” Smart Growth ScorecardSmart Growth Scorecard
Increased development at Increased development at infill locationsinfill locations
DestinationsDestinations
25% greater potential for 25% greater potential for traditional designtraditional design
DesignDesign
23% increase in mixing at 23% increase in mixing at local levellocal level
DiversityDiversity
11% increase for new growth11% increase for new growthDensityDensity
Vision Scenario Vision Scenario improvements over Trendimprovements over Trend
5. Distance to Transit 5. Distance to Transit
8%8% of new residents live within of new residents live within ½½ mile of transit (1mile of transit (1%% under Trend Case)under Trend Case)
11%11% of new jobs are within of new jobs are within ½½ mile of transit (8mile of transit (8%% under Trend Case)under Trend Case)
The Smart Growth Scenario reduces VMT and The Smart Growth Scenario reduces VMT and improves levels of congestion on major roadsimproves levels of congestion on major roads
Countywide VMTCountywide VMT --7%7%
% of Arterial Miles Congested % of Arterial Miles Congested -- 42%42%(Peak hour LOS E or F)
% of Freeway Miles Congested% of Freeway Miles Congested -- 15%15%(Peak hr LOS E or F in at least 1 direction)
Measuring the Effects of Land Use on Measuring the Effects of Land Use on Travel Behavior and Climate ChangeTravel Behavior and Climate Change
Jerry Walters, Fehr & PeersJerry Walters, Fehr & Peers
HaagenHaagen--SmitSmit SymposiumSymposium““Challenge to ChangeChallenge to Change””
Questions?
