Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
29 CHAPTER 3. PEDESTRIAN SAFETY DATABASES This chapter describes the pedestrian safety databases assembled for use in this research from the City of Toronto, the City of Charlotte, and two metropolitan areas in Minnesota. SIGNALIZED INTERSECTIONS IN TORONTO The City of Toronto, Ontario, Canada, has over 1,500 signalized intersections in its jurisdiction. The City has collected both vehicle and pedestrian volume data at most of these intersections which creates a rich database for pedestrian safety research. Intersections meeting the following criteria were included in this study: ⢠signalized control ⢠not the terminal of a freeway ramp ⢠no one-way intersection legs ⢠no turn restrictions For each site, the research team obtained data elements on intersection characteristics, signal data, vehicle and pedestrian volumes, and vehicle-pedestrian crashes. Intersection Characteristics Table 10 summarizes the data elements that were available for signalized intersections in Toronto. Most of these variables were available in an existing City database. The research team supplemented the existing data through a review of aerial photographs to obtain data on the presence of median refuge islands and to verify data on the number of lanes on each intersection leg. Vehicle and Pedestrian Volumes Vehicle and pedestrian volume data were available for each intersection leg, but no vehicle turning movement counts were available. The expansion factor for pedestrian volumes discussed above were based initially on a 1985 FHWA study by Zegeer et al. (10). As part of that study, there was a need to develop pedestrian volume expansion factors to adjust short-term pedestrian counts to pedestrian ADTs. Since cities typically do not collect pedestrian volumes on a 24-hr basis, the only available counts were a large sample of pedestrian volume counts taken in Seattle, Washington, which were summarized by hour of the day and also by area type (e.g., CBD, fringe, and residential). The 1985 study used those data to compute pedestrian volume adjustment factors to allow for expanding shorter counts to approximate pedestrian volumes on a 24-hr basis. This pedestrian
30 ADT expansion methodology was further refined, checked, and used in a 2005 FHWA study by Zegeer et al. (3). These adjustment factors are given and explained in detail on page 67 of the 2005 FHWA report. Vehicle-Pedestrian Collision Data Data for vehicle-pedestrian collisions at the signalized intersections in Toronto were available from City records for seven years from 1999 through 2005. Hard-copy police reports were reviewed for many of these collisions to verify their location relative to the intersection in question. Vehicle-pedestrian collisions were attributed to the intersection if they occurred within 76 m (250 ft) of the intersection and were related to the operation of the intersection. TABLE 10. Summary of Available Data Elements for Signalized Intersections in Toronto and Charlotte. Data element Available in Toronto Database Available in Charlotte Database INTERSECTION-LEVEL DATA Number of intersection legs X X Intersection/skew angle X X Presence of lighting X INTERSECTION-LEG-LEVEL DATA Number of through lanes X X Number of left-turn lanes X X Number of right-turn lanes X X Right-turn treatment X Presence of marked pedestrian crosswalks X X Presence of median X X Presence of curb parking X Presence of sidewalks (left and right) X Presence and type of pedestrian signal X X Type of left-turn phasing X Curb return radius X One-way vs. two-way traffic operation X X Posted speed limit X Presence and type of turn restrictions X X VOLUME DATA Average daily traffic volume by leg X X Vehicle turning movement counts by leg X Pedestrian crossing volume by leg X X SIGNALIZED INTERSECTIONS IN CHARLOTTE The City of Charlotte has approximately 600 signalized intersections in its jurisdiction. Given that Charlotte Department of Transportation (CDOT) collects vehicle and pedestrian counts at these intersections on a regular basis, many of these intersections were suitable for this study. Intersections meeting the following criteria were included in this study: ⢠signalized control
31 ⢠not the terminal of a freeway ramp ⢠no one-way intersection legs ⢠no turn restrictions ⢠not equipped with automated enforcement For each site, the research team collected data on the intersection characteristics, signal data, vehicle and pedestrian volumes, and vehicle-pedestrian crashes. Intersection Characteristics Geometric and physical characteristics of an intersection can affect pedestrian safety. Some characteristics are particular to the whole intersection, whereas some are leg specific and can be different for different legs of an intersection. Table 10 summarizes the data that were available for signalized intersections in Charlotte. Most of these data were collected by examining high-resolution aerial photography and signal plans provided by CDOT. The streetlight data were obtained in a GIS format from the local power company and spatially joined to the intersections to determine how many were within 30 m (100 ft). The presence and type of turn restrictions were determined by brief field visits to each intersection. Vehicle and Pedestrian Volume The main source for vehicle and pedestrian volume data was the set of turning movement data supplied by CDOT. The city conducts 12-hour turning movement counts (7:00am â 7:00pm) of most intersections on a regular basis (approximately once every two or three years). The turning movement count broke down the movements on each leg to through, right, left, and pedestrian. Vehicle Annual Average Daily Traffic (AADT) was calculated from an expansion of the 12-hour count. CDOT supplied a specific expansion factor for each intersection count. Years for which there was no count were filled by extrapolation. Percentage of right and left turns were calculated by dividing the number of turning vehicles by the total vehicles on the leg. For intersections with no turning movement counts available, traffic volume data were obtained from the Metropolitan Planning Organization (MPO). The MPO regional network contained volumes on most major streets. The volumes were estimations based on vehicle trip generation and distribution. Approximately 3 percent of intersections had no turning movement count and relied on MPO traffic volume data. Pedestrian daily volumes were also calculated from an expansion of the 12-hour count. There was no expansion factor for pedestrian volumes available from CDOT. Based on past research, the team used an expansion factor of 1.16 to convert a 12-hour pedestrian volume to a daily volume. Pedestrian volumes were not available for sites relying on MPO traffic volume data, and these sites were excluded from subsequent analyses. The expansion factor for pedestrian volumes discussed above were based initially on a 1985 FHWA study by Zegeer et al. (10). As part of that study, there was a need to develop pedestrian volume expansion factors to adjust short-term pedestrian counts to pedestrian ADTs.
32 Since cities typically do not collect pedestrian volumes on a 24-hr basis, the only available counts were a large sample of pedestrian volume counts taken in Seattle, Washington, which were summarized by hour of the day and also by area type (e.g., CBD, fringe, and residential). The 1985 study used those data to compute pedestrian volume adjustment factors to allow for expanding shorter counts to approximate pedestrian volumes in a 24-hr basis. This pedestrian ADT expansion methodology was further refined, checked, and used in a 2005 FHWA study by Zegeer et al. (3). These adjustment factors are given and explained in detail on page 67 of the 2005 FHWA report. Vehicle-Pedestrian Collision Data Pedestrian crash data were obtained from CDOT in a GIS format. Crash data were available for a period of nine years from 1997 to 2005. Crashes had been manually geocoded by city staff. For most years, if a crash occurred within 30 m (100 ft) of an intersection, the crash was placed at the intersection (in GIS terms, the crash was said to be âsnappedâ to the intersection point). If a crash occurred farther than 30 m (100 ft) from an intersection, it was placed on the appropriate road segment. Given this rough method of determining whether a crash was intersection-related, the team desired more accuracy. To this goal, crashes were matched to the records in the state pedestrian crash database. North Carolina has a detailed pedestrian and bicycle crash database that includes information such as crash typing, which gives further description of the pre-crash actions of each party. The state database was used to determine with more accuracy whether or not the crash was intersection-related. Unfortunately, crashes could not be matched by a unique identifier, so they had to be matched based on date and location (street names). Some crashes could not be matched to the state record in this manner. In these cases, the city specification of whether the crash was intersection-related was kept as the final rule. Land Use and Demographic Characteristics Data on land use and demographic characteristics of the area surrounding each intersection in Charlotte was assembled through analysis of planning data available in GIS format. The available data elements included: ⢠presence of bus stops within 300 m (1,000 ft) of the intersection ⢠presence of schools (either public or private) with 300 m (1,000 ft) of the intersection ⢠presence of parks within 300 m (1,000 ft) of the intersection ⢠number of alcohol sales establishments within 300 m (1,000 ft) of the intersection ⢠average per capita income of all census block groups within 300 m (1,000 ft) of the intersection ⢠number of square feet of buildings on commercial land parcels partially or entirely within 0.8 km (0.5 mi) of the intersection ⢠number of commercial structures on commercial land parcels within 0.8 km (0.5 mi) of the intersection
33 ⢠number of commercial land parcels within 300 m (1,000 ft) of the intersection ROADWAY SEGMENTS IN MINNESOTA A database of vehicle-pedestrian collisions for roadway segments in Minnesota was available from the research performed in Phases I and II of this project (25). The database assembled for Phases I and II consisted of five years of crash data (1998-2002), including data for vehicle-pedestrian collisions. The roadway segment crash database includes all vehicle- pedestrian collisions on the roadway segments of interest that did not occur at an intersection and were not related to the operation of an intersection. Table 11 presents the list of roadway segment characteristics included in the database developed for Phases I and II of the research. For the roadway segments located in the Minneapolis-St. Paul and St. Cloud metropolitan areas, the database from Phases I and II was expanded to include eight years (1998-2005) of vehicle-pedestrian collision data. The following land use and demographic variables were added to these data through a combination of on-line database and field reviews: ⢠Type of development (commercial/residential/mixed) ⢠Presence and location of schools within three blocks of the roadway segment ⢠Presence of school crossings within the roadway segment or at adjacent intersections ⢠Location and traffic control for school crossings ⢠Presence and location of parks within three blocks of the roadway segment ⢠Type of park facilities ⢠Number of alcohol sales establishments within roadway segment ⢠Number of bus stops within roadway segment or at adjacent intersections TABLE 11. Site Characteristics Data Obtained for Roadway Segments on Urban and Suburban Arterials in Minnesota (25). Data element Description of data element Primary source Beginning landmark Name of intersecting street or other landmark at beginning of block Highway agency data or field review Beginning milepost Milepost or log mileage at beginning of block to tie field locations to accident data Highway agency data Bicycle route (marked) Presence of bicycle route marked by signs Field or videolog review Bicycle facilities Presence of bicycle facilities including sidepath, marked bicycle lane, or wide curb lane Field or videolog review Driveway locations Location of driveway (side of road and distance from beginning of block) Aerial photograph and field review Driveway types Each driveway was classified into one of seven categories (see accompanying text) Aerial photograph and field review Ending landmark Name of intersecting street or other landmark at end of block Highway agency data or field review
TABLE 11. Site Characteristics Data Obtained for Roadway Segments on Urban and Suburban Arterials in Minnesota (25) (Continued) 34 Data element Description of data element Primary source Ending milepost Milepost or log mileage at end of block to tie field locations to accident data Highway agency data Grade Roadway grade within block (level, moderate, or steep) Field or videolog review Horizontal curve length Length of horizontal curve (mi) computed from beginning and ending locations Computed Horizontal curve location Distance of beginning and end of horizontal curve from beginning of block (mi) Aerial photographs Horizontal curve radius Radius of horizontal curve (ft) measured on aerial photograph Aerial photographs Intersections A basic data set was collected for the intersections at each end of each block including number of legs, side of road (for three-leg intersections), traffic control device, and type of pedestrian facilities (if any). This data set is less extensive than the data collected for the full intersection study sites. Field or videolog review Lane width Width of through lanes (ft) not including gutters. Measured at first block in a series of consecutive blocks and points of change. Lane width is averaged over multiple lanes where present. Field measurement Length of site Length of site (mi) from beginning landmark to end landmark. Measured from center of intersection where intersections are site boundaries. Highway agency data Lighting Presence of street lighting (none, intersection only, or continuous lighting along street) and presence of other ambient lighting Field review Median opening location Distance of median opening from beginning of block (mi). Applicable to divided streets only. Aerial photograph or field review Median opening type Type of median opening (conventional/directional) Aerial photograph or field review Median type Presence and type of median (none, raised, depressed, flush) Field or videolog review On-street parking Presence and type of on-street parking (none, parallel parking, angle parking) Field or videolog review Pedestrian crosswalk (midblock) Location and type of midblock crosswalk (if any) Field or videolog review Roadside hazard rating Rating on 1 to 7 scale (see accompanying text) Field or videolog review Roadway type Number of through lanes and divided undivided (2U, 3T, 4U, 4D, or 5T) as defined in Chapter 2 of this report. Field or videolog review Route number or street name Route number or street name for arterial used to tie field locations to accident data Highway agency data
TABLE 11. Site Characteristics Data Obtained for Roadway Segments on Urban and Suburban Arterials in Minnesota (25) (Continued) 35 Data element Description of data element Primary source Shoulder type Type of shoulder (paved, gravel, turf, composite) and presence/absence of curb. Determined separately for each side of the road. Field or videolog review Shoulder width Width of shoulder (ft). Determined separately for each side of the road. Entered as zero for curbed sections. Field measurement Sidewalks Presence/absence of sidewalk. Determined separately for each site of the road. Field or videolog review Speed limit Posted speed limit (mph) or speed limit applicable under state law Field or videolog review Traffic volume ADTs for each year of study period. Interpolated between count years when not available for every year. Highway agency data The major limitation of the Minnesota roadway segment data is that it includes traffic volume, but not pedestrian volume data. DESCRIPTIVE STATISTICS FOR AVAILABLE DATABASES Signalized Intersections Table 12 summarizes the number of intersections, number of vehicle-pedestrian collisions, and exposure measures in the available databases for signalized intersections in Toronto and Charlotte. Data are tabulated separately for three-leg signalized (3SG) intersections and four-leg signalized (4SG) intersections. Table 13 summarizes descriptive statistics for four key variables in the databases for intersections in Toronto and Charlotte. The key variables summarized in the table are: ⢠major-road ADT (veh/day) ⢠minor-road ADT (veh/day) ⢠total pedestrian volume crossing all intersection legs (pedestrians/day) ⢠maximum number of lanes crossed by pedestrians at intersection considering presence of refuge islands Roadway Segments Table 14 summarizes the number and length of roadway segments, number of vehicle- pedestrian collisions, and exposure measures in the available database for roadway segments in Minnesota. Data are tabulated separately for five types of roadway segments: ⢠two-lane undivided arterials (2U) ⢠three-lane arterials including a center two-way left-turn lane (TWLTL) (3T) ⢠four-lane undivided arterials (4U) ⢠four-lane divided arterials (i.e., including a raised or depressed median) (4D) ⢠five-lane arterials including a center TWLTL (5T)
36 TABLE 12. Number of Intersections, Number of Vehicle-Pedestrian Collisions, and Exposure Measures for Signalized Intersections in Toronto and Charlotte. Vehicle-pedestrian collisions Intersection type Number of intersections Number of vehicle- pedestrian collisions Average study period duration (years) Total entering vehicles (100 millions) Total crossing pedestrians (100 millions) per intersection per year per 100 million entering vehicles per 100 million crossing pedestrians TORONTO 3SG 366 681 6.72 209.5 7.8 0.28 3.25 87.5 4SG 1,166 4,530 6.89 1,223.0 53.5 0.56 3.70 84.7 CHARLOTTE 3SG 84 47 8.02 65.1 0.1 0.07 0.72 597.3 4SG 267 294 8.28 262.0 1.4 0.13 1.12 204.7
37 TABLE 13. Descriptive Statistics for Key Data Elements at Signalized Intersections in Toronto and Charlotte. 3SG intersections 4SG intersections Variable Mean Median Standard deviation Minimum value Maximum value Mean Median Standard deviation Minimum value Maximum value TORONTO Major-road ADT (veh/day) 30,055 29,032 11,902 811 74,226 30,138 28,820 12,024 6,576 80,187 Minor-road ADT (veh/day) 6,562 4,709 6,599 152 51,412 11,570 8,726 9,299 164 49,100 Total pedestrian volume crossing all intersection legs (pedestrians/day) 867 309 1,291 0 12,548 1,823 898 3,210 1 34,129 Maximum number of lanes crossed by pedestrians at intersection considering presence of refuge islands â 5 â 2 9 â 5 â 2 9 CHARLOTTE Major-road ADT (veh/day) 22,959 23,363 8,160 6,173 44,157 24,369 22,932 11,944 5,764 73,235 Minor-road ADT (veh/day) 7,025 5,219 5,215 79 21,695 8,096 6,162 6,866 371 43,481 Total pedestrian volume crossing all intersection legs (pedestrians/day) 32 17 55 0 398 178 51 556 0 6,002 Maximum number of lanes crossed by pedestrians at intersection considering presence of refuge islands â 4 â 2 6 â 4 â 2 9
38 TABLE 14. Number and Length of Roadway Segments, Number of Vehicle-Pedestrian Collisions, and Exposure Measures for Minnesota Database. Roadway segment type Number of segments Total length of segments (mi) Number of vehicle- pedestrian collisions Average study period duration (years) Average vehicle ADT (veh/day) Total veh-mi of travel (100 millions) Vehicle- pedestrian collisions per mi per year Vehicle- pedestrian collisions per 100 million veh-mi 2U 486 63.1 34 8 10,725 19.76 0.067 1.72 3T 169 18.7 15 8 13,462 7.35 0.100 2.04 4U 548 53.7 95 8 15,900 24.93 0.221 3.81 4D 273 41.8 45 8 28,932 35.31 0.135 1.27 5T 39 6.2 12 8 16,608 3.01 0.242 3.99
