Schneider

Document Sample
Schneider Powered By Docstoc
					                                                                                                  Percent of Weekly Pedestrian Volume per Hour
                                                                                  0.00%
                                                                                                                0.60%
                                                                                                                                         1.00%
                                                                                                                                                              1.40%




                                                                                          0.20%
                                                                                                   0.40%
                                                                                                                             0.80%
                                                                                                                                                 1.20%




                                                                          12 AM
                                                                          4 AM
                                                                          8 AM
                                                                                                                                                         M




                                                                          12 PM
                                                                           4 PM
                                                                           8 PM




                            Robert J. Schneider
                                                                          12 AM
                                                                          4 AM
                                                                          8 AM
                                                                                                                                                         T




                                                                          12 PM
                                                                           4 PM
                                                                           8 PM
                                                                          12 AM
                                                                          4 AM
                                                                          8 AM
                                                                          12 PM
                                                                                                                                                         W




                                                                           4 PM
                                                                           8 PM
                                                                          12 AM
                                                                          4 AM
                                                                          8 AM
                                                                                                                                                         Th




                                                                          12 PM
                                                                           4 PM
                                                                           8 PM
                                                                          12 AM
                                                                          4 AM
                                                                          8 AM
                                                                                                                                                         F




                                                                          12 PM
                                                                           4 PM
                                                                           8 PM
                                                                          12 AM
                                                                          4 AM
                                                                          8 AM
                                                                                                                                                         Sa




                                                                          12 PM
                                                                           4 PM
                                                                           8 PM
                                                                          12 AM
                                                                          4 AM
                                                                          8 AM
                                                                                                                                                         Su




                                                                          12 PM
Safe Transportation Education & Research Center (SafeTREC)—January 2010




                                                                           4 PM
                                                                                                                                                                      Estimating Pedestrian Volumes




                                                                           8 PM
                   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       City of Portland, OR
   pedestrian activity
   occurs
                   Pedestrian Crash Analysis
                                          Reported
                                        Pedestrian
 Mainline        Intersecting              Crashes
 Roadway         Roadway               (1996-2005)
Mission         Torrano
Boulevard       Avenue                           5

Davis Street    Pierce Avenue                    4
Foothill
Boulevard       D Street                         1
Mission         Jefferson
Boulevard       Street                           5
University
Avenue          Bonar Street                     7
International
Boulevard       107th Avenue                     2
San Pablo
Avenue          Harrison Street                  2
East 14th       Hasperian
Street          Boulevard                        1
International
Boulevard     46th Avenue                        3
              Masonic
Solano Avenue Avenue                             2

Broadway        12th Street                      5
                     Pedestrian Risk Analysis
                                     Estimated        Annual       Ten-Year       Reported   Pedestrian
                                  Total Weekly     Pedestrian    Pedestrian     Pedestrian Risk (Crashes
 Mainline        Intersecting       Pedestrian        Volume        Volume         Crashes per 10,000,000
 Roadway         Roadway              Crossings      Estimate      Estimate    (1996-2005)      crossings)
Mission         Torrano
Boulevard       Avenue                     1,169        60,796       607,964              5           82.24

Davis Street    Pierce Avenue              1,570        81,619       816,187              4           49.01
Foothill
Boulevard       D Street                    632         32,862       328,624              1           30.43
Mission         Jefferson
Boulevard       Street                     5,236       272,246     2,722,464              5           18.37
University
Avenue          Bonar Street              11,175       581,113     5,811,127              7           12.05
International
Boulevard       107th Avenue               3,985       207,243     2,072,429              2            9.65
San Pablo
Avenue          Harrison Street            4,930       256,357     2,563,572              2            7.80
East 14th       Hasperian
Street          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
              Masonic
Solano Avenue 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
                                                                40 miles




                                                   Oakland



                                                      Alameda
    Key Partner = ACTIA
                                          San
•                                      Francisco
                                                       County


•   Population = 1.46 million
•   Land area = 738 square miles
•   Largest City = Oakland (401,000)
Example: Broadway & 2nd Street
    Pedestrian Screenline/Segment Counts




                                      Google Earth—Tele Atlas 2008
Example: Broadway & 2nd Street
    Pedestrian Midblock Crossing Counts




                                      Google Earth—Tele Atlas 2008
Example: Broadway & 2nd Street
   Pedestrian Intersection Crossing Counts




                                       Google Earth—Tele Atlas 2008
Example: Broadway & 2nd Street
   Pedestrian Intersection Crossing Counts




                                       Google Earth—Tele Atlas 2008
Intersection Count Form (Pedestrians)
Example: Broadway & 2nd Street
      Bicyclist Intersection Turning Counts




                         Straight




   Left

                                        Right




                                          Google Earth—Tele Atlas 2008
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 Pedestrian Volume per Hour




        0.00%
                                      0.60%
                                                               1.00%
                                                                                    1.40%




                0.20%
                         0.40%
                                                   0.80%
                                                                       1.20%
12 AM
4 AM
8 AM
                                                                               M
12 PM
 4 PM
 8 PM
12 AM
4 AM
8 AM
                                                                               T




12 PM
 4 PM
 8 PM
12 AM
4 AM
8 AM
12 PM
                                                                               W




 4 PM
 8 PM
12 AM
4 AM
8 AM
                                                                               Th




12 PM
 4 PM
 8 PM
12 AM
4 AM
8 AM
                                                                               F




12 PM
 4 PM
 8 PM
12 AM
4 AM
8 AM
                                                                               Sa




12 PM
 4 PM
 8 PM
12 AM
4 AM
8 AM
                                                                               Su




12 PM
 4 PM
 8 PM
                                                                                            Percent of Weekly Volume by Hour (Composite of 13 Automated Count Sites)
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
                        Percent of Weekly Pedestrian Volume per Hour




        0.00%
                                      0.60%
                                                               1.00%
                                                                                    1.40%




                0.20%
                         0.40%
                                                   0.80%
                                                                       1.20%
12 AM
4 AM
8 AM
12 PM
                                                                               M
 4 PM
 8 PM
12 AM
4 AM
8 AM
                                                                               T




12 PM
 4 PM
 8 PM
12 AM
4 AM
8 AM
12 PM
                                                                               W




 4 PM
 8 PM
12 AM
4 AM
8 AM
                                                                               Th




12 PM
 4 PM
 8 PM
12 AM
4 AM
8 AM
                                                                               F




12 PM
 4 PM
 8 PM
12 AM
4 AM
8 AM
                                                                               Sa




12 PM
 4 PM
 8 PM
12 AM
4 AM
8 AM
                                                                               Su




12 PM
                                                                                            “Typical” Pedestrian Activity Pattern vs. Employment Centers




 4 PM
 8 PM
                                Percent of Weekly Volume per Hour




        0.00%
                                         0.60%
                                                     0.80%
                                                                            1.20%
                                                                                         1.40%




                0.20%
                        0.40%
                                                                    1.00%
12 AM
 4 AM
 8 AM
                                                                                    M
12 PM
 4 PM
 8 PM
12 AM
 4 AM
 8 AM
                                                                                    T




12 PM
 4 PM
 8 PM
12 AM
 4 AM
 8 AM
                                                                                    W




12 PM
 4 PM
 8 PM
12 AM
 4 AM
 8 AM
                                                                                    Th




12 PM
 4 PM
 8 PM
12 AM
 4 AM
 8 AM
                                                                                    F




12 PM
 4 PM
 8 PM
12 AM
 4 AM
 8 AM
                                                                                    Sa




12 PM
 4 PM
 8 PM
12 AM
 4 AM
 8 AM
                                                                                    Su




12 PM
 4 PM
                                                                                                 “Typical” Pedestrian Activity Pattern vs. Employment Centers




 8 PM
                                      “Typical” Pedestrian Activity Pattern vs. Employment Centers
                                    1.40%
                                            M       T       W              Th              F                Sa   Su
                                                                                       Employment Centers
                                                                                     12 to 2 p.m., Wednesday =
                                    1.20%
                                                                                         2.63% of total
                                                                                         weekly volume

                                    1.00%
Percent of Weekly Volume per Hour




                                    0.80%




                                    0.60%                    Composite of 13 Locations
                                                              12 to 2 p.m., Wednesday=
                                                                2.16% of total
                                                                weekly volume
                                    0.40%




                                    0.20%




                                    0.00%
                                             4 AM
                                             8 AM




                                             4 AM
                                             8 AM




                                             4 AM
                                             8 AM




                                             4 AM
                                             8 AM




                                             4 AM
                                             8 AM




                                             4 AM
                                             8 AM




                                             4 AM
                                             8 AM
                                            12 PM




                                            12 PM




                                            12 PM




                                            12 PM




                                            12 PM




                                            12 PM




                                            12 PM
                                            12 AM




                                            12 AM
                                            12 AM




                                            12 AM




                                            12 AM




                                            12 AM




                                            12 AM
                                             4 PM
                                             8 PM




                                             4 PM
                                             8 PM




                                             4 PM
                                             8 PM




                                             4 PM
                                             8 PM




                                             4 PM
                                             8 PM




                                             4 PM
                                             8 PM




                                             4 PM
                                             8 PM
              Land Use Adjustment Factors
  Counts taken at locations with specific types of land uses were multiplied by these
  factors to match counts taken at “typical” Alameda County Locations
                                                                                    Count Times when Adjustment Factors were Applied
                                                                              Weekday Weekday Weekday Weekday Saturday Saturday Saturday
Land Use Category Definition                                                  12-2 p.m. 2-4 p.m. 3-5 p.m. 4-6 p.m. 9-11 a.m. 12-2 p.m. 3-5 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
                    >=500 jobs within 0.25 miles (402 m) 4 & no
Residential Area    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    >=10 commercial retail properties within 0.1 miles
Commercial Area (161 m)5                                                       0.92      1.00     1.00      0.97      1.04     0.77      0.78
Near Multi-Use      >=0.5 centerline miles of multi-use trails within 0.25
Trail               miles (402 m)6                                             1.63      0.79     0.72      0.91      0.69     1.31      1.07
                    >=1 elementary, middle, or high school within 0.25
Near School         miles (402 m)5                                             0.94      0.77     0.82      1.07      1.20     1.23      1.37
            Weather Adjustment Factors
 Counts taken under certain weather conditions were multiplied by these factors to
 match counts taken during “typical” Alameda County weather conditions
                                                                     Count Times when Adjustment Factors were Applied
Weather                                                                     Weekday                        Saturday
Condition    Definition                                                     12-6 p.m.                    9 a.m.-5 p.m.
             >=80 degrees Fahrenheit (27 degrees Celsius) during
Warm         first count hour 7                                              1.07                           1.12
             <=50 degrees Fahrenheit (10 degrees Celsius) during
Cool         first count hour 7                                              1.10                           1.06
             <= 0.6 of the expected solar radiation (Langleys per
Cloudy       day) during first count hour 7,8                                1.11                           1.11
             >=0.01 inch (0.254 mm) of precipitation during either
Rain         count hour 7                                                    1.27                           1.34
                                    Effect of Rain on Pedestrian Volumes: Weekend vs. Weekday
                                   10%
                                                      Rain                  2%             1%
                                    0%
                                                                                 No rain
average % deviation from typical




                                   -10%


                                   -20%

                                                             -24%
                                   -30%


                                   -40%


                                   -50%                                                Sunday 12PM


                                   -60%                                                Weekday 8AM
                                               -61%

                                   -70%
Condition                Definition                                                                        12-6 p.m.                                 9 a.m.-5 p.m.



               Seasonal Adjustment Factors
                         >=80 degrees Fahrenheit (27 degrees Celsius) during
Warm                     first count hour 7                                                                   1.07                                        1.12
                         <=50 degrees Fahrenheit (10 degrees Celsius) during
Cool                     first count hour 7                                                                   1.10                                        1.06
                         <= 0.6 of the expected solar radiation (Langleys per
Cloudy                   day) during first count hour 7,8                                                     1.11                                        1.11

Counts taken during the spring were multiplied by these factors to match counts
                         >=0.01 inch (0.254 mm) of precipitation during either
Rain            count hour 7                                        1.27              1.34
                                               typical time were multiplied
taken in Alameda County during aApril through Juneof the year by these factors to match counts taken
Seasonal Adjustment Factors (Counts taken from
in Alameda County during a typical time of the year) 3
                                                                                                 Count Times when Adjustment Factors were Applied

Land Use Category Definition                                                                                              All Time Periods

Employment
Center                   >=2,000 jobs within 0.25 miles (402 m) 4                                                                0.98
                                                                   4
                         >=500 jobs within 0.25 miles (402 m) & no
Residential Area         commercial retail properties within 0.1 miles (161 m) 5                                                 0.97
Neighborhood             >=10 commercial retail properties within 0.1 miles
Commercial Area          (161 m)5                                                                                                0.98
Near Multi-Use           >=0.5 centerline miles of multi-use trails within 0.25
Trail                    miles (402 m)6                                                                                          0.91
                         >=1 elementary, middle, or high school within 0.25
Near School              miles (402 m)5                                                                                          0.93
1) Land use adjustment factors based on hourly automated sensor counts taken at 13 locations in Alameda County between April 2008 and June 2009.
2) Weather adjustment factors based on hourly automated sensor counts taken at 13 locations in Alameda County between April 2008 and June 2009.
3) Employment center, residential area, neighborhood commercial area, and multi-use trail seasonal adjustment factors based on hourly automated sensor counts taken at 13
locations in Alameda County from April 2008 to June 2009. School seasonal adjustment factor based on hourly automated sensor counts taken at 3 locations in Alameda County from
May 2009 to June 2009.
4) Source = Traffic Analysis Zones from San Francisco Bay Area Metropolitan Transportation Commission, 2005
5) Source = Land Use Parcels from Alameda County Tax Assessor's Office, 2007
6) Source = Bay Area Multi‐Use Trail Centerlines from San Francisco Bay Area Metropolitan Transportation Commission, 2007
7) Source = California Irrigation Management Information System, 2008-2009 (Mills College, Union City, and Pleasanton weather stations).
       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
       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
     – Extrapolate to typical week: Multiply by 42.54
       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
     – Extrapolate to typical week: Multiply by 42.54
     – Extrapolate to typical year: Multiply by 52.18
       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
     – Extrapolate to typical week: Multiply by 42.54
     – Extrapolate to typical year: Multiply by 52.18
     – Account for spring count: Multiply by 0.981
       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
     –   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
       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
     – 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
       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
     – 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=0.897
Independent variables significant at 95% confidence level
Model Spreadsheet
                                                                 Model Spreadsheet

                                                                        Pedestrian Intersection Crossing Volume Model
                                                                                       Pilot Model--January 20091,2
                                                                    Developed by Robert Schneider, Lindsay Arnold, and David Ragland
                                                                          University of California-Berkeley Traffic Safety Center
                                                                                                                                                                       4
                                          Intersection Identification                                                                                   Model Inputs                                                Model Output
                                                                                                                                                                 Total number of          Presence of regional
                                                                                                                  Total population        Total employment       commercial               transit station within             Estimated
                                                                                                                  within 1/2-mile         within 1/4-mile        properties within        1/10 mile              Pedestrian Crossings
     ID # Mainline Roadway                       Intersecting Roadway                     City                    radius 3                radius                 1/4-mile radius          (Yes = 1, No = 0)      in a Typical Week5,6,7

        1 Telegraph Avenue                       16th Street                              Oakland                                10720                  16440                        86                       0                49504
        2 Telegraph Avenue                       27th Street                              Oakland                                 9780                   3100                        66                       0                17449
        3 Telegraph Avenue                       51st Street                              Oakland                                 8970                    970                        41                       0                 9573
        4 Telegraph Avenue                       59th Street                              Oakland                                10270                    610                        27                       0                 8613
Notes :
1. Thi s i s a pi l ot model ba s ed on a s tudy of weekl y pedes tra i n vol umes a t 50 i nters ections i n Al a meda County, CA. The model ha s a good fi t for the Al a meda County s tudy da ta (a djus ted-
 2
R =0.897). Si nce the a na l ys i s wa s conducted on 50 i nters ections i n Al a meda County, CA, more res ea rch i s needed to refi ne the model equa tion a nd determi ne the a ppl i ca bi l i ty of the
res ul ts for other communi ties . The model equa tion i s : Es tima ted pedes tri a n i nters ection cros s i ngs per week = 0.928 * Total popul a tion wi thi n 0.5-mi l es of the i nters ection + 2.19 * Total
empl oyment wi thi n 0.25-mi l es of the i nters ection + 98.4 * Number of commerci a l retai l properties wi thi n 0.25-mi l es of the i nters ection + 54,600 * Number of regi ona l tra ns i t s tations wi thi n
0.10-mi l es of the i nters ection - 4910. Detai l s of the s tudy a re provi ded i n two pa pers : 1) Schnei der, R.J., L.S. Arnol d, a nd D.R. Ra gl a nd. "Extra pol a ting Weekl y Pedes tri a n Inters ection
Cros s i ng Vol umes from 2-Hour Ma nua l Counts ," UC-Berkel ey Tra ffi c Sa fety Center, Tra ns portation Res ea rch Record (Forthcomi ng), 2010, a nd 2) Schnei der R.J., L.S. Arnol d, a nd D.R. Ra gl a nd.
“A Pi l ot Model for Es tima ting Pedes tri a n Inters ection Cros s i ng Vol umes ,” UC‐Berkel ey Tra ffi c Sa fety Center, Tra ns portation Res ea rch Record (Forthcomi ng), 2010.
2. The pedes tri a n vol ume es tima tes produced by the model a re i ntended for pl a nni ng, pri ori tiza tion, a nd s a fety a na l ys i s a t the communi ty, nei ghborhood, a nd corri dor l evel s . Si nce the
model provi des rough es tima tes of pedes tri a n a ctivi ty, a ctua l pedes tri a n counts s houl d be us ed for s i te-l evel s a fety, des i gn, a nd engi neeri ng a na l ys es .
3. The i nters ections s el ected for the s tudy di d not i ncl ude i nters ections i n a rea s wi th very l ow popul a tion dens i ties (<50 peopl e per s qua re mi l e). Therefore, the model i s not a ppropri a te
for i nters ections bel ow thi s dens i ty thres hol d (i .e., the model does not a ppl y i f there a re fewer tha n 64 peopl e wi thi n a 1/2-mi l e ra di us ).
4. The s tudy of Al a meda County, CA found tha t l a nd us e cha ra cteri s tics a re the mos t i mportant fa ctors for predi cting pedes tri a n a ctivi ty. Roa dwa y des i gn fa ctors , s uch a s the pres ence of
s i dewa l ks , medi a n cros s i ng i s l a nds , curb ra di i , or pedes tri a n cros s i ng s i gna l s ma y ha ve mi nor effects on pedes tri a n vol umes , but they a re not a s s i gni fi ca nt for predi cting pedes tri a n
a ctivi ty. However, roa dwa y des i gn fa ctors a re cri tica l for pedes tri a n s a fety a nd comfort. Roa dwa ys mus t be des i gned to a ccommoda te pedes tri a ns of a l l a bi l i ties , rega rdl es s of vol ume.
5. The model output i s a n es tima te of the number of pedes tri a n cros s i ngs duri ng a typi ca l 168-hour week (wi th a n a vera ge s ea s ona l vol ume). Pedes tri a n cros s i ngs a re counted ea ch time
a pedes tri a n cros s es a ny l eg of the i nters ection (e.g., one pers on i s counted twi ce i f they cros s the ea s t l eg a nd then the s outh l eg of a n i nters ection). Pedes tri a ns do not need to cros s
compl etel y i ns i de the cros s wa l k; they a re counted i f i f they cros s wi thi n 50 feet of the i nters ection.
6. The model ma y not perform wel l i n l oca tions cl os e to s peci a l a ttra ctors , s uch a s a mus ement pa rks , wa terfronts , s ports a rena s , regi ona l recrea tion a rea s , a nd ma jor mul ti-us e tra i l s .
Pedes tri a n vol umes i n thes e a rea s tend to be hi ghl y va ri a bl e, wi th hi gh vol umes duri ng certai n s ea s ons or duri ng ni ce wea ther. Bri dges a nd underpa s s es ma y a l s o cha nnel pedes tri a n
a ctivi ty, s o more res ea rch ma y be neces s a ry to a djus t vol ume es tima tes nea r thes e fea tures .
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
                                   Daily Pedestrian Volume on Broadway in Downtown Oakland
                          7,000
                                      Daily Pedestrian Volume (Feb. 16-22, 2009)              Typical Daily Pedestrian Volume
                                      Rain totals are for Feb. 16-22                                       April 2008 to April 2009
                                                                                               No Rain
                          6,000
                                                                        No Rain    No Rain
                                                  Rained 0.96"
                                  Rained 0.59"

                          5,000
Daily Pedestrian Volume




                                                                                                            No Rain
                          4,000




                          3,000
                                                                                                                            Rained 1.65"



                          2,000




                          1,000




                             0
                                   Monday           Tuesday            Wednesday   Thursday     Friday      Saturday           Sunday
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 35
and 57 percent lower than the average volume for the same hour of the week
over the entire year. The effect of rain may be greater on weekends because
more trips may be discretionary.

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

Warm temperatures (slight reduction)
Pedestrian counts taken between 12 p.m. and 1 p.m. on Saturdays showed
that each additional degree Fahrenheit was associated with one percent lower
pedestrian volume. Two weekday models showed that pedestrian volumes may
be 5 to 8 percent lower than average when the temperature is above 80
degrees 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 degrees
Fahrenheit (10 Celsius) were associated with lower pedestrian volumes.

High winds (slight reduction)
The weekday mid-day model showed that higher winds were associated with
lower 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
              2008 Weekly       2009 Weekly
         Pedestrian Volume Pedestrian Volume Absolute Difference
    ID #   based on Counts   based on Counts       (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
                                                            25,000

                                                                                                                                   Line Representing
                                                                                                                                   Perfect Prediction
                                                                                                                                   (Observ. = Pred.)
Weekly Pedestrian Volume Predicted by Pilot Model in 2008




                                                            20,000




                                                            15,000                                                                                  Trendline for
                                                                                                                                                    Observed vs.
                                                                                                                                                   Predicted Data

                                                            10,000




                                                             5,000




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




                                                            -5,000
                                                                                          Weekly Pedestrian Volume "Observed" in 2009
       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=0.900
Independent variables significant at 90% confidence level
            Which Intersection Features are
            Associated with Pedestrian Risk?

Pedestrian Crossings (+)
While intersections with more pedestrian
crossings have more pedestrian crashes,
there may be a “safety in numbers” effect
(i.e., lower crash risk per crossing).
    (Expected Effect*: 100% more pedestrian crossings, 49% more crashes)


Motor Vehicle Volume (+)
There may be a “danger in numbers” effect
with mainline motor vehicle volume, but need
to explore the influence of congestion and
speed.
     (Expected Effect*: 100% more mainline AADT, >100% more crashes)
             Which Intersection Features are
             Associated with Pedestrian Risk?
Number of Right-Turn-Only Lanes (+)
Intersections with more right-turn-only lanes
may have longer crossing distances and more
complex interactions between drivers and
pedestrians.
    (Expected Effect*: 1 more right-turn-only lane, 53% more crashes)

Number of Driveway Crossings (+)
Intersections with more non-residential
driveway crossings within 50 ft. may have
more conflict points; drivers may focus on
entering or exiting motor vehicle lanes.
    (Expected Effect*: 1 more driveway crossing, 33% more crashes)


Medians (-)
Mainline and cross-street legs with medians
have a refuge that allows pedestrians to
cross one direction of traffic at a time, which
may make crossing safer.
    (Expected Effect*: Medians on mainline roadway crossings, 75% fewer crashes)
            Which Intersection Features are
            Associated with Pedestrian Risk?

Number of Commercial Properties (+)
Intersections with more commercial
properties within 0.1 miles may have more
drivers looking at signs and for parking; more
pedestrians may cross between cars.
    (Expected Effect*: 100% more pedestrian crossings, 49% more crashes)


Percentage of Residents Under 18 (+)
A greater percentage of young pedestrians
within 0.25 miles may indicate that more of
the people crossing are less experienced and
have higher risk crossing busy streets.
     (Expected Effect*: 100% more mainline AADT, >100% more crashes)

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:25
posted:9/25/2012
language:English
pages:60