Estimating Pedestrian Volumes Robert J. Schneider Safe Transportation Education & Research Center...

Estimating Pedestrian Volumes

Robert J. SchneiderSafe Transportation Education & Research Center (SafeTREC)—January

2010

0.00%

0.20%

0.40%

0.60%

0.80%

1.00%

1.20%

1.40%

12 A

M

4 A

M

8 A

M

12 P

M

4 PM

8 PM

12 A

M

4 A

M

8 A

M

12 P

M

4 PM

8 PM

12 A

M

4 A

M

8 A

M

12 P

M

4 PM

8 PM

12 A

M

4 A

M

8 A

M

12 P

M

4 PM

8 PM

12 A

M

4 A

M

8 A

M

12 P

M

4 PM

8 PM

12 A

M

4 A

M

8 A

M

12 P

M

4 PM

8 PM

12 A

M

4 A

M

8 A

M

12 P

M

4 PM

8 PM

Perc

ent o

f Wee

kly

Pede

stri

an V

olum

e pe

r H

our

M T W Th F Sa Su

Overview

• Why are pedestrian volumes important?• Alameda County pedestrian counting

methods• Extrapolating short counts to weeks and

years• Estimating volumes from a model

1) Why are Pedestrian Volumes Important?

• Track pedestrian volume over time

• Quantify exposure to calculate pedestrian crash risk

• See where & when pedestrian activity occurs

City of Portland, OR

Pedestrian Crash Analysis

Mainline Roadway

Intersecting Roadway

Reported Pedestrian

Crashes (1996-2005)

Mission Boulevard

Torrano Avenue 5

Davis Street Pierce Avenue 4Foothill Boulevard D Street 1Mission Boulevard

Jefferson Street 5

University Avenue Bonar Street 7International Boulevard 107th Avenue 2San Pablo Avenue Harrison Street 2East 14th Street

Hasperian Boulevard 1

International Boulevard 46th Avenue 3

Solano AvenueMasonic Avenue 2

Broadway 12th Street 5

Pedestrian Risk Analysis

Mainline Roadway

Intersecting Roadway

Estimated Total Weekly

Pedestrian Crossings

Annual Pedestrian

Volume Estimate

Ten-Year Pedestrian

Volume Estimate

Reported Pedestrian

Crashes (1996-2005)

Pedestrian Risk (Crashes

per 10,000,000 crossings)

Mission Boulevard

Torrano Avenue 1,169 60,796 607,964 5 82.24

Davis Street Pierce Avenue 1,570 81,619 816,187 4 49.01Foothill Boulevard D Street 632 32,862 328,624 1 30.43Mission Boulevard

Jefferson Street 5,236 272,246 2,722,464 5 18.37

University Avenue Bonar Street 11,175 581,113 5,811,127 7 12.05International Boulevard 107th Avenue 3,985 207,243 2,072,429 2 9.65San Pablo Avenue Harrison Street 4,930 256,357 2,563,572 2 7.80East 14th Street

Hasperian Boulevard 3,777 196,410 1,964,102 1 5.09

International Boulevard 46th Avenue 12,303 639,752 6,397,522 3 4.69

Solano AvenueMasonic Avenue 22,203 1,154,559 11,545,589 2 1.73

Broadway 12th Street 112,896 5,870,590 58,705,898 5 0.85

2) Alameda County Pedestrian Counting Methods

• Manual Counts– Field data collectors & count sheets– Short time periods

• Automated Counts– Sensor technology– Continuous counts

Alameda County, CA

• Key Partner = ACTIA• Population = 1.46 million• Land area = 738 square miles• Largest City = Oakland

(401,000)

San Francisco

Oakland

Alameda County

40 miles

Example: Broadway & 2nd Street

Google Earth—Tele Atlas 2008

Pedestrian Screenline/Segment Counts



Pedestrian Midblock Crossing Counts



Pedestrian Intersection Crossing Counts

Intersection Count Form (Pedestrians)



Bicyclist Intersection Turning Counts

Right

Straight

Left

Intersection Count Form (Bicyclists)

Manual Count Resource Needs

• Site selection (strategic analysis)• $200-$300 per 2-hour pedestrian & bicycle count

– Data collector training & management– Travel to site– Count period– Most sites = 1 data collector (except high volume)

• Data entry & cleaning (15-20 min. per 2-hour count)

• Cost reduction– Use existing count methods and forms– Use interns & volunteers (cautiously!)– Repeat at regular intervals

Automated Counters

www.eco-compteur.com

Percent of Weekly Volume by Hour (Composite of 13 Automated Count Sites)

0.00%

0.20%

0.40%

0.60%

0.80%

1.00%

1.20%

1.40%12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM

Perc

ent o

f Wee

kly

Pede

stri

an V

olum

e pe

r H

our

M T W Th F Sa Su

Raw Counter Data

Automated Count Resource Needs

• $2,000-$2,500 per EcoCounter Infrared Sensor• Budget time for permission & travel to location• 10-15 min. to install (follow instructions)• 5-10 min. to download data in field (need software)

• Data cleaning & analysis– Search data for anomalies– Summarize data in meaningful way– Repeat at regular intervals

• Alternatives: Other sensors, video

3) Extrapolating Short Pedestrian Counts

• Calculated extrapolation factors from continuous pedestrian counts– Time of day, day of week, season of year– Land use– Weather

• Identified “peak” pedestrian activity• Derived from 13 locations in Alameda

County

0.00%

0.20%

0.40%

0.60%

0.80%

1.00%

1.20%

1.40%12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM

Perc

ent o

f Wee

kly

Pede

stri

an V

olum

e pe

r H

our

M T W Th F Sa Su

“Typical” Pedestrian Activity Pattern vs. Employment Centers


0.00%

0.20%

0.40%

0.60%

0.80%

1.00%

1.20%

1.40%12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM

Perc

ent o

f Wee

kly

Vol

ume

per H

our

M T W Th F Sa Su


0.00%

0.20%

0.40%

0.60%

0.80%

1.00%

1.20%

1.40%12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM12

AM

4 A

M8

AM

12 P

M4

PM8

PM

Perc

ent o

f Wee

kly

Vol

ume

per H

our

M T W Th F Sa Su

Composite of 13 Locations12 to 2 p.m., Wednesday=

2.16% of total weekly volume

Employment Centers12 to 2 p.m., Wednesday =

2.63% of total weekly volume

Land Use Adjustment Factors

Land Use Category DefinitionWeekday12-2 p.m.

Weekday2-4 p.m.

Weekday3-5 p.m.

Weekday4-6 p.m.

Saturday9-11 a.m.

Saturday12-2 p.m.

Saturday3-5 p.m.

Employment Center >=2,000 jobs within 0.25 miles (402 m)4 0.83 0.97 0.99 0.99 1.16 1.00 1.07

Residential Area>=500 jobs within 0.25 miles (402 m)4 & no

commercial retail properties within 0.1 miles (161 m)5 1.37 0.96 0.90 0.98 0.86 1.14 1.12

Neighborhood Commercial Area

>=10 commercial retail properties within 0.1 miles

(161 m)5 0.92 1.00 1.00 0.97 1.04 0.77 0.78

Near Multi-Use Trail

>=0.5 centerline miles of multi-use trails within 0.25

miles (402 m)6 1.63 0.79 0.72 0.91 0.69 1.31 1.07

Near School>=1 elementary, middle, or high school within 0.25

miles (402 m)5 0.94 0.77 0.82 1.07 1.20 1.23 1.37

Count Times when Adjustment Factors were Applied

Counts taken at locations with specific types of land uses were multiplied by these factors to match counts taken at “typical” Alameda County Locations

Weather Adjustment Factors

Counts taken under certain weather conditions were multiplied by these factors to match counts taken during “typical” Alameda County weather conditions

Weather Condition Definition

Warm>=80 degrees Fahrenheit (27 degrees Celsius) during

first count hour7

Cool<=50 degrees Fahrenheit (10 degrees Celsius) during

first count hour7

Cloudy<= 0.6 of the expected solar radiation (Langleys per

day) during first count hour7,8

Rain>=0.01 inch (0.254 mm) of precipitation during either

count hour7

Saturday9 a.m.-5 p.m.


1.07

Weekday12-6 p.m.

1.10

1.11

1.27

1.12

1.06

1.11

1.34

1

Effect of Rain on Pedestrian Volumes: Weekend vs. Weekday

-61%

2%

-24%

1%

-70%

-60%

-50%

-40%

-30%

-20%

-10%

0%

10% Rain

No rain

Sunday 12PM

Weekday 8AM

ave

rag

e %

de

via

tio

n f

rom

ty

pic

al

Seasonal Adjustment Factors

Counts taken during the spring were multiplied by these factors to match counts taken in Alameda County during a typical time of the year

Land Use Category Definition

Employment Center >=2,000 jobs within 0.25 miles (402 m)4

Residential Area>=500 jobs within 0.25 miles (402 m)4 & no

commercial retail properties within 0.1 miles (161 m)5

Neighborhood Commercial Area

>=10 commercial retail properties within 0.1 miles

(161 m)5

Near Multi-Use Trail

>=0.5 centerline miles of multi-use trails within 0.25

miles (402 m)6

Near School>=1 elementary, middle, or high school within 0.25

miles (402 m)5 0.93


All Time Periods

0.98

0.97

0.98

0.91

Example: Estimate the Annual Pedestrian Volume at an

Intersection• Telegraph Ave. & 27th St. (Oakland)• 2-hour pedestrian count (Tue., 5/26, 2-4

p.m.)• 65 degrees, cloudy• Total crossings of all legs = 212



p.m.)• 65 degrees, cloudy• Total crossings of all legs = 212– Extrapolate to typical week: Multiply by 42.54



p.m.)• 65 degrees, cloudy• Total crossings of all legs = 212– Extrapolate to typical week: Multiply by 42.54– Extrapolate to typical year: Multiply by 52.18



p.m.)• 65 degrees, cloudy• Total crossings of all legs = 212– Extrapolate to typical week: Multiply by 42.54– Extrapolate to typical year: Multiply by 52.18– Account for spring count: Multiply by 0.981



p.m.)• 65 degrees, cloudy• Total crossings of all legs = 212– Extrapolate to typical week: Multiply by 42.54– Extrapolate to typical year: Multiply by 52.18– Account for spring count: Multiply by 0.981– Account for employment & commercial retail

land uses: Multiply by 0.97 and by 1.002


Intersection• Telegraph Ave. & 27th St. (Oakland)• 2-hour pedestrian count (Tue., 5/26, 2-4 p.m.)• 65 degrees, cloudy• Total crossings of all legs = 212

– Extrapolate to typical week: Multiply by 42.54– Extrapolate to typical year: Multiply by 52.18– Account for spring count: Multiply by 0.981– Account for employment & commercial retail land

uses: Multiply by 0.97 and by 1.002– Account for cloudy weather: Multiply by 1.06


Intersection• Telegraph Ave. & 27th St. (Oakland)• 2-hour pedestrian count (Tue., 5/26, 2-4 p.m.)• 65 degrees, cloudy• Total crossings of all legs = 212

– Extrapolate to typical week: Multiply by 42.54– Extrapolate to typical year: Multiply by 52.18– Account for spring count: Multiply by 0.981– Account for employment & commercial retail land

uses: Multiply by 0.97 and by 1.002– Account for cloudy weather: Multiply by 1.06– Estimated annual pedestrian crossings ~ 475,000

4) Estimating Pedestrian Volumes from Statistical Models

• Developed model from counts at 50 intersections in Alameda County

• Identified factors associated with higher vols.– Total population within 0.5 mi– Total employment within 0.25 mi– Number of commercial retail properties within

0.25 mi– Presence of regional rail station within 0.1 mi

• Created simple spreadsheet for applying model

Pilot Model Formula

Estimated Weekly Pedestrian Crossings =

0.928 * Total population within 0.5-miles of the

intersection+ 2.19 * Total employment within 0.25-

miles of the intersection

+ 98.4 * Number of commercial properties within

0.25-miles of the intersection+54,600 * Number of regional transit stations

within 0.10-miles of the intersection

- 4910 (Constant)

Adjusted R2=0.897Independent variables significant at 95% confidence level

Model Spreadsheet

Model Spreadsheet

ID # Mainline Roadway Intersecting Roadway City

Total population within 1/2-mile

radius3

Total employment within 1/4-mile radius

Total number of commercial properties within 1/4-mile radius

Presence of regional transit station within 1/10 mile (Yes = 1, No = 0)

Estimated Pedestrian Crossings

in a Typical Week5,6,7

1 Telegraph Avenue 16th Street Oakland 10720 16440 86 0 495042 Telegraph Avenue 27th Street Oakland 9780 3100 66 0 174493 Telegraph Avenue 51st Street Oakland 8970 970 41 0 95734 Telegraph Avenue 59th Street Oakland 10270 610 27 0 8613

Model OutputIntersection Identification Model Inputs 4

Pedestrian Intersection Crossing Volume ModelPilot Model--January 20091,2

Developed by Robert Schneider, Lindsay Arnold, and David RaglandUniversity of California-Berkeley Traffi c Safety Center

Pilot Pedestrian Volume Model

Application

Considerations for Applying the Model

• Designed for estimating volumes at neighborhood, corridor, and community levels. Actual pedestrian counts should be used for site-level safety, design, and engineering analyses.

Thank you

• Lindsay Arnold & David Ragland (SafeTREC)

• Alameda County Transportation Improvement Authority

• California Department of Transportation

• Volunteer counters & SafeTREC students

Questions?

EcoCounter Validation Counts

• Prior Studies:– Shawn Turner, et al. (2007), "Testing and Evaluation of Pedestrian

Sensors", http://swutc.tamu.edu/publications/technicalreports/167762-1.pdf

– Ryan Greene-Roesel, et al. (2008), “Effectiveness of a Commercially Available Automated Pedestrian Counting Device in Urban Environments: Comparison with Manual Counts”, http://www.tsc.berkeley.edu/news/08-0503session240ryanposter.pdf

• High and low pedestrian volumes• Different sidewalk widths• Different weather conditions

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

Monday Tuesday Wednesday Thursday Friday Saturday Sunday

Dai

ly P

edes

tria

n Vo

lum

eDaily Pedestrian Volume on Broadway in Downtown Oakland

Daily Pedestrian Volume (Feb. 16-22, 2009) Typical Daily Pedestrian Volume

Rained 0.59" Rained 0.96"No Rain

Rained 1.65"

No Rain

No Rain

No Rain

Rain totals are for Feb. 16-22 April 2008 to April 2009

How Does Weather Affect Pedestrian Volume?*

(Linear model from Attaset, Schneider, Arnold, & Ragland, 2009)

Rain (35 to 57 percent reduction)Pedestrian counts during hours with measurable rain tended to be between 35and 57 percent lower than the average volume for the same hour of the weekover the entire year. The effect of rain may be greater on weekends becausemore trips may be discretionary.

Cloud cover (5 to 24 percent reduction)Pedestrian volumes collected when it was cloudy tended to be between 5 and24 percent lower than typical volumes during the same hour of the week overthe entire year. The effect of clouds may be greater on weekends due todiscretionary trips.

Warm temperatures (slight reduction)Pedestrian counts taken between 12 p.m. and 1 p.m. on Saturdays showedthat each additional degree Fahrenheit was associated with one percent lowerpedestrian volume. Two weekday models showed that pedestrian volumes

maybe 5 to 8 percent lower than average when the temperature is above 80degrees Fahrenheit (27 Celsius).

How Does Weather Affect Pedestrian Volume?*

(Linear model from Attaset, Schneider, Arnold, & Ragland, 2009)

Cool temperatures (slight reduction)The weekday afternoon model showed that temperatures below 50

degreesFahrenheit (10 Celsius) were associated with lower pedestrian

volumes.

High winds (slight reduction)The weekday mid-day model showed that higher winds were

associated withlower pedestrian volumes.

*Results are from Alameda County, CA (very mild climate)

Pilot Pedestrian Volume Model Testing

• Found that some counts were close to predicted values, but others were more than 50% off

• Proposed alternative model specifications based on validation counts

• Other pedestrian volume models– Cameron (1976)—Manhattan– Benham & Patel (1977)—Milwaukee CBD– Desyllas, et al. (2003)—Central London– Raford & Ragland (2004, 2005)— Oakland, CA; Boston, MA– Pulugurtha & Repaka (2008)—Charlotte– Clifton, et al. (2008)—Maryland Cities– Liu & Griswold (2009)—San Francisco

Variation in Pedestrian Volumes

• 5 Control Intersections

ID #

2008 Weekly Pedestrian Volume

based on Counts

2009 Weekly Pedestrian Volume

based on CountsAbsolute Difference

(2009 - 2008) Percent Difference1

50 315 310 -5 1.6%2650 15691 16113 422 2.7%9179 8342 7429 -913 12.3%9436 105297 88118 -17179 19.5%

499 5186 3448 -1738 50.4%1) Percent difference is calculated using the smaller number as the base value. If the model value is greater than the actual value, the percent difference is calculated as (2009 - 2008)/2008. If the actual value is greater than the model value, the percent difference is calculated as (2008 - 2009)/2009.

2009 Observed Volumes vs. Pilot Model Predictions

-5,000

0

5,000

10,000

15,000

20,000

25,000

0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000

Wee

kly

Pede

stri

an V

olum

e Pr

edic

ted

by P

ilot M

odel

in 2

008

Weekly Pedestrian Volume "Observed" in 2009

Trendline forObserved vs.

Predicted Data

Line Representing Perfect Prediction (Observ. = Pred.)

Possible Model Formulations

• 1) New Alameda County Model– Correlation between employment & vehicle ownership– Distance to closest school

• 2) Model with Total Population Squared– Correlation between employment & vehicle ownership– Five factors significant at 99% confidence level– RMSE = 4,470; RMSPE = 7,480

• 3) Revised Pilot Model– Four key factors are significant in most models

estimated

Revised Pilot Model Formula

Estimated Weekly Pedestrian Crossings =

0.987 * Total population within 0.5-miles of the

intersection+ 2.19 * Total employment within 0.25-

miles of the intersection

+ 71.1 * Number of commercial properties within

0.25-miles of the intersection+49,300 * Number of regional transit stations

within 0.10-miles of the intersection

- 4850 (Constant)

Adjusted R2=0.900Independent variables significant at 90% confidence level

Which Intersection Features are Associated with Pedestrian Risk?

Pedestrian Crossings (+) While intersections with more pedestriancrossings have more pedestrian crashes,there may be a “safety in numbers”

effect(i.e., lower crash risk per crossing).

Motor Vehicle Volume (+) There may be a “danger in numbers”

effectwith mainline motor vehicle volume, but

needto explore the influence of congestion andspeed.

(Expected Effect*: 100% more pedestrian crossings, 49% more crashes)

(Expected Effect*: 100% more mainline AADT, >100% more crashes)


Number of Right-Turn-Only Lanes (+) Intersections with more right-turn-only lanesmay have longer crossing distances and

morecomplex interactions between drivers andpedestrians.

Number of Driveway Crossings (+) Intersections with more non-residentialdriveway crossings within 50 ft. may havemore conflict points; drivers may focus onentering or exiting motor vehicle lanes.

Medians (-) Mainline and cross-street legs with medianshave a refuge that allows pedestrians tocross one direction of traffic at a time, whichmay make crossing safer.

(Expected Effect*: 1 more right-turn-only lane, 53% more crashes)

(Expected Effect*: 1 more driveway crossing, 33% more crashes)

(Expected Effect*: Medians on mainline roadway crossings, 75% fewer crashes)


Number of Commercial Properties (+) Intersections with more commercialproperties within 0.1 miles may have

moredrivers looking at signs and for parking;

morepedestrians may cross between cars.

Percentage of Residents Under 18 (+) A greater percentage of young

pedestrianswithin 0.25 miles may indicate that more

ofthe people crossing are less experienced

andhave higher risk crossing busy streets.

(Expected Effect*: 100% more pedestrian crossings, 49% more crashes)

(Expected Effect*: 100% more mainline AADT, >100% more crashes)

Estimating Pedestrian Volumes Robert J. Schneider Safe Transportation Education & Research Center...

Documents

Transcript of Estimating Pedestrian Volumes Robert J. Schneider Safe Transportation Education & Research Center...