Guidebook on Identification of High Pedestrian Crash Locations

CHAPTER 7. SUPPLEMENTAL MATERIALS

This chapter presents several supplemental material examples and uses them to illustrate the method undertaken to identify locations of interest. The chapter also includes a discussion of the advantages and disadvantages for the more common assessment methods.

Supplemental Material A—Example of Safety Index

Los Angeles uses an index that considers the following:

• Number injured.
• Number of fatalities (with a 1.5 factor).
• Age of the pedestrian (an additional point is assigned for youth and older pedestrians).
• Health and equity index. In the Health Atlas for the City of Los Angeles document (table 2, page 179), the components considered in the city’s community health and equity index are provided.⁽²⁸⁾ For example, a weight of 35 points is provided for the variable of hardship index.

Supplemental Material B—Screening Method Examples

Site Identification Analysis

Identification of points or segments is readily available in certain GIS packages, such as ArcGIS. The optimized version of this package analyzes a layer of data (pedestrian crashes in this case), creates a grid to group the data in the context of their extent, and computes the Getis-Ord Gi statistic, which is useful to find areas in the grid with high (and low) representation of a variable of interest (pedestrian crashes in this case). Figure 16 shows the output of running this tool with its default parameters. The area near the center of Austin clearly has more pedestrian crashes relative to the whole county. However, the size of the squares of the grid is relatively large for further, more nuanced decisions.

Screen capture ©Texas A&M Transportation Institute using ArcGIS software by ESRI. ArcGIS Desktop: Release 10.4.1. Service Layer Credits: Esri, HERE, DeLorme, USGS, Intermap, INCREMENT P, NRCan, METI, NGCC, ©OpenStreetMap. Crash data provided by the Texas Department of Transportation.
Figure 16. Graphic. Pedestrian preliminary assessment of Austin-area pedestrian crashes.

A grid of higher resolution can be set for increased detail by limiting the analysis to the area previously identified as being the primary location with a higher number of crashes. In this case, the reduced grid is set to have cells of 328 by 328 ft. Figure 17 shows the distribution of the cells with at least one crash inside. There is a clear concentration of pedestrian crashes in the Austin downtown area and, more specifically, along a west/northwest–east/southeast corridor that includes 6th Street. This visualization presents smaller areas with a high number of pedestrian crashes. To further focus the attention to more problematic subregions, the Getis-Ord Gi analysis can be run again on a different variable: the crash frequency per square. Figure 18 shows the result of this action.

Figure 18 displays two cells whose counts are exactly eight pedestrian crashes in 5 yr, yet one of those cells is flagged as a hot spot (Neches Street and East 7th Street), while the other one is not (South Lamar Boulevard and Lamar Square Drive). The difference between these two sites is their Getis-Ord local statistic. For Neches Street and East 7th Street, that statistic amounts to 3.0598; in contrast, the Getis-Ord statistic is slightly smaller (2.854) for South Lamar Boulevard and Lamar Square Drive.

Screen capture ©Texas A&M Transportation Institute using ArcGIS software by ESRI. ArcGIS Desktop: Release 10.4.1.4.1 Service Layer Credits: Esri, HERE, DeLorme, USGS, Intermap, INCREMENT P, NRCan, METI, NGCC, ©OpenStreetMap. Crash data provided by the Texas Department of Transportation.
Figure 17. Graphic. Pedestrian preliminary assessment of area near downtown Austin, TX.

A higher Getis-Ord statistic indicates a higher count of events (crashes) within a context of cells with high counts. Thus, given a high count of crashes, the cell with a lower statistic may be an isolated case of high crashes, but the cell flagged as a hot spot by the higher Getis-Ord statistic indicates overrepresentation of pedestrian crashes at that cell and its neighbor cells. Only a handful of sites were labeled as hot spots after this step (16 cells, as shown in figure 18). Interestingly, 4 out of 16 hot-spot cells are along the axis of 6th Street. The following example utilizes only the crashes on this arterial to demonstrate sliding window crash screening.

Screen capture ©Texas A&M Transportation Institute using ArcGIS software by ESRI. ArcGIS Desktop: Release 10.4.1. Service Layer Credits: Esri, HERE, DeLorme, USGS, Intermap, INCREMENT P, NRCan, METI, NGCC, ©OpenStreetMap. Crash data provided by the Texas Department of Transportation.
Figure 18. Graphic. Identification of locations of concern using small grids.

Sliding Window Screening for 6th Street

The sliding window approach uses overlapping windows for smoothing any errors in crash location reporting. For this screening method, a window length of 984 ft is selected, performance measures are calculated for that window, the window is then moved for a given smaller offset (e.g., 328 ft), and the analysis is repeated for the shifted window. This procedure is repeated multiple times for as many windows as it takes to cover the complete corridor. An example application of this technique using the 6th Street corridor in Austin, TX, is discussed next.

As shown in figure 19, the corridor of 6th Street between Mopac Expressway and Interstate (I)-35 is roughly 2.2 mi long and had 104 pedestrian crashes between 2009 and 2014. Thirty-seven sliding windows of 984 ft can be defined for this corridor. Figure 20 shows the resulting windows. The crash frequency was computed per sliding window, resulting in the map shown in figure 21.

There are 6 windows with 16 or more pedestrian crashes, as shown in figure 22.

Screen capture ©Texas A&M Transportation Institute using ArcGIS software by ESRI. ArcGIS Desktop: Release 10.4.1. Service Layer Credits: Esri, HERE, DeLorme, USGS, Intermap, INCREMENT P, NRCan, METI, NGCC, ©OpenStreetMap. Crash data provided by the Texas Department of Transportation.
Figure 19. Graphic. 6th Street corridor area of analysis.

Screen capture ©Texas A&M Transportation Institute using ArcGIS software by ESRI. ArcGIS Desktop: Release 10.4.1. Service Layer Credits: Esri, HERE, DeLorme, USGS, Intermap, INCREMENT P, NRCan, METI, NGCC, ©OpenStreetMap. Crash data provided by the Texas Department of Transportation.
Figure 20. Graphic. Sliding windows shown as transparent, overlapping oblongs for 6th Street corridor analysis.

Screen capture ©Texas A&M Transportation Institute using ArcGIS software by ESRI. ArcGIS Desktop: Release 10.4.1. Service Layer Credits: Esri, HERE, DeLorme, USGS, Intermap, INCREMENT P, NRCan, METI, NGCC, ©OpenStreetMap. Crash data provided by the Texas Department of Transportation.
Figure 21. Graphic. Sliding window analysis results.

Screen capture ©Texas A&M Transportation Institute using ArcGIS software by ESRI. ArcGIS Desktop: Release 10.4.1. Service Layer Credits: Esri, HERE, DeLorme, USGS, Intermap, INCREMENT P, NRCan, METI, NGCC, ©OpenStreetMap. Crash data provided by the Texas Department of Transportation.
Figure 22. Graphic. Segments of concern identified from the sliding window analysis.

Since the six critical windows identified overlap, these windows define three areas suited for more granular assessments. Figure 23 shows the first area of concern. This area of interest ranges through five intersections, from Nueces Street to Colorado Street. It is clear from figure 23 that the communality between these three windows is that they all contain the intersection of Guadalupe Street. Incidentally, the previous analysis identified this intersection as one of the top four sites of concern along 6th Street. One can conclude from these two analyses that the intersection of 6th Street and Guadalupe Street has an overrepresentation of pedestrian crashes compared to the rest of the 6th Street corridor (from the sliding window analysis) and that its surroundings also tend to have overrepresentation of pedestrian crashes (from the site identification analysis).

Screen capture ©Texas A&M Transportation Institute using ArcGIS software by ESRI. ArcGIS Desktop: Release 10.4.1. Service Layer Credits: Esri, HERE, DeLorme, USGS, Intermap, INCREMENT P, NRCan, METI, NGCC, ©OpenStreetMap. Crash data provided by the Texas Department of Transportation.
Figure 23. Graphic. One of the areas of concern—includes critical intersections on 6th Street at Nueces Street, San Antonio Street, Guadalupe Street, Lavaca Street, and Colorado Street.

Figure 24 shows the other two segments identified from the sliding window analysis: the single window from Trinity to Red River Street and the two overlapping windows from Sabine Street to the I-35 Service Road. The site identification analysis identified the latter as a location of concern as well.

Screen capture ©Texas A&M Transportation Institute using ArcGIS software by ESRI. ArcGIS Desktop: Release 10.4.1. Service Layer Credits: Esri, HERE, DeLorme, USGS, Intermap, INCREMENT P, NRCan, METI, NGCC, ©OpenStreetMap. Crash data provided by the Texas Department of Transportation.
Figure 24. Graphic. Two segments of interest along 6th Street: Trinity through Red River Street and Sabine Street to I-35 Service Road.

Supplemental Material C—Online Maps

Numerous city and State transportation departments along with FHWA are making pedestrian crash data available online through interactive maps. These online maps allow anyone with access to a Web browser to perform a variety of functions with regard to pedestrian crashes, including the following:

• Zoom out to get a citywide view, or zoom in to get a street- or intersection-specific view.
• Display the crash density by grouping crashes that occur close to each other.
• View pedestrian crashes in the context of nearby land use, socioeconomic attributes, or planned street or highway improvements.
• Filter/subset the crashes by year, crash severity, crash cause, and other crash attributes (e.g., gender, day of week, road agency ownership).
• Click on an individual crash or high crash location to get additional information.

The following are examples of such maps:

• The Federal Highway Administration shows pedestrian fatal crashes for 2009 to 2013 on its HEPGIS site: http://hepgis.fhwa.dot.gov/fhwagis/ViewMap.aspx?map=Annual+Fatal+Crashes|Pedestrian+Fatal+Crashes+2009-2013.
  
  
• California has funded the development of the Transportation Injury Mapping System to provide data and mapping analysis tools and information for traffic safety–related research, policy, and planning. The mapping system is available at https://tims.berkeley.edu/. Several tools are available, including interactive maps and the ability to query collision data, such as pedestrian-involved collisions.
  
  
• As part of the Austin Vision Zero initiative, Vision Zero ATX, which is a local nonprofit organization, created a color-coded map that indicates each fatality’s mode of travel (see http://www.visionzeroatx.org/austin-fatality-map/). This map does not group nearby crashes to show crash density but does allow the user to filter by several attributes, including gender, control of road, time of crash, and day of week.
  
  
• As part of its Vision Zero initiative, the City of Portland, OR, displays all traffic fatalities and serious injuries on a map that illustrates fatality and serious injury crash density. The map shows 10 yr of crash data (2005 through 2014), displays density with different-sized symbols, allows the user to click on a particular crash for additional information, and allows the user to view pedestrian crashes separately from other crashes. Additionally, under the pedestrian-only crash tab, this map highlights (with a yellow line) the 20 pedestrian high crash corridors as determined by the city’s crash analysis. It is available at http://pdx.maps.arcgis.com/apps/MapSeries/index.html?appid=cf122cd3b4ef46f0ac496b2d61d554e9.
  
  
• As part of its Vision Zero initiative, the City of Los Angeles, CA, displays its high injury network for people walking and biking who are killed or seriously injured in the context of other variables, such as a community health and equity index.⁽²⁸⁾ The Los Angeles DOT also used this high injury network to prioritize Safe Routes to School (SRTS) projects. While SRTS prioritization preceded Vision Zero, the city noted that the top 50 schools were all within 0.25 mi of the high injury network. The map is available at http://visionzero.lacity.org/high-injury-network/.
  
  
• As part of a pedestrian safety initiative called WalkFirst, the City of San Francisco, CA, displays high pedestrian injury corridors and intersections in a map-based interface. The WalkFirst site has high injury corridors and intersections categorized by crash profiles (i.e., causal crash factors such as low nighttime visibility, left-turning vehicle at signalized intersection, alcohol use, etc.) such that users can click on a certain crash profile to see where that causal factor contributes to a high crash density. The interactive map also uses color-coded lines to indicate crash density on these high injury streets and can be seen at http://walkfirst.sfplanning.org/index.php/home/streets.
  
  
• In a separate analysis of pedestrian safety in Los Angeles, CA, the Los Angeles Times identified the 817 most dangerous intersections in Los Angeles County using the California Highway Patrol’s Statewide Integrated Traffic Records System (http://graphics.latimes.com/la-pedestrians-how-we-did-it/). The LA Times analysis considered three factors in designating dangerous intersections:
  
  
  • Total number of pedestrian collisions.
  • Proportion of collisions that involved a pedestrian.
  • Proportion of pedestrian crashes that were fatal.
    
    
• It is important to note that the resulting map shows the density of dangerous intersections, not the density of actual pedestrian crashes. The intent of the graphical map was to show, from a regional perspective, the location of problem areas (zones with multiple intersections) with respect to pedestrian safety (see http://graphics.latimes.com/la-pedestrians/).
  
  
• The North Central Texas Council of Governments, the metropolitan planning organization for the Dallas–Ft. Worth region, has developed static pedestrian crash maps that are available online. The maps show individual crash locations overlaid on a color-coded density map and include pedestrian and bicyclist crashes from 2010 through 2014. An example is available online at http://www.nctcog.org/trans/sustdev/bikeped/BikePedCrashInfo.asp.
  
  
• Massachusetts DOT (MassDOT) displays an interactive map of top crash locations for pedestrians and bicyclists. This map groups nearby pedestrian crashes into intersection clusters based on the crash frequency occurring, with the top 200 ranking intersections displayed with a different-colored symbol. There are several different types of cluster displays that can be viewed, depending on the desired year range and type of crash (pedestrian or bicyclist). The map is located at http://gis.massdot.state.ma.us/maptemplate/topcrashlocations.
  
  
• The Houston Chronicle has compiled and displays pedestrian crash data in an interactive map. The map can be used to show individual crashes and associated attributes, like road, weather conditions, and contributing factors. A heat map that groups nearby pedestrian crashes is also viewable and uses color coding to indicate relative crash density. There are also filters that allow users to focus on a selected crash severity or contributing crash factor. An example of the Houston, TX, map is available at http://www.houstonchronicle.com/default/media/Pedestrian-accident-map-252958.php.
  
  
• FDOT has a website crash visualization tool that permits the user to modify year, intersection type, crash harmful event (which permits filtering on pedestrian crashes), and others. An example of a FDOT map is available at https://fdotewp1.dot.state.fl.us/TrafficSafetyWebPortal/FivePercent/LocationList.aspx.
  
  

Supplemental Material D—Advice from Previous Studies

Several previous studies have documented analyses to identify high crash locations, and these studies have used different approaches to identify and rank locations. Examples of approaches used and lessons learned include the following:

• A study explored different ranking methods to identify high pedestrian crash zones for the Las Vegas Metropolitan area.⁽²⁹⁾ The ranking methods included crash frequency, crash density, and crash rate, as well as composite methods such as the sum-of-the-ranks and the crash score methods. In addition, two density methods (simple method and kernel method) were used within a GIS. Results showed a significant variation in ranking when individual methods were considered, while the rankings of the high pedestrian crash zones were relatively consistent when the sum-of-the-ranks method and the crash score method were used. The authors recommended that composite methods should be used in ranking high pedestrian crash zones.
  
  
• A study using data from San Francisco, CA, found that although the pedestrian crash frequency was higher in the vicinity of the central business district, the pedestrian crash rate (crash frequency normalized by pedestrian exposure) was higher in the periphery of the city.⁽³⁰⁾ The authors noted that disregarding pedestrian exposure could significantly affect the results.
  
  
• Crash data can be combined with other data to develop a ranked list of locations for safety improvements. For example, Ottawa, Canada, combined crash data with public input and findings from the Intersection Safety Indices (ISI) tool.^(31,32) Ped ISI allows the ranking of intersections based on their traffic control, number of lanes, 85th-percentile speeds, average daily traffic, and the predominant land use in the area. The combination of crash data with other datasets recognizes that pedestrian crashes are rare compared to vehicle crashes, and the combination of crashes with surrogates permits the consideration of other variables associated with pedestrian activity or situations known to be challenging for pedestrians.
  
  
• In a 2013 study in Florida, the authors observed that for small clusters (i.e., roadway segments less than 0.2 mi), normalized crash frequencies were inflated, and these segments were ranked among the highest.⁽³³⁾ Clusters were then categorized into intersections and corridors using a 500-ft threshold. Segments shorter than the threshold were identified as intersections, while those longer than the threshold were identified as corridors.
  
  
• MassDOT identified top locations where reported collisions occurred between pedestrians and motorists.⁽³⁴⁾ The crash cluster analysis methodology used a fixed search distance (for both pedestrian and bicycle crashes of 328 ft compared to 82 ft for intersection locations) to merge crash clusters together. Because of the relatively small number of reported pedestrian crashes in the crash data file, the clustering analysis used crashes from the 10-yr period from 2005 through 2014, instead of the 3-yr analysis for intersection locations. The actual crash clusters can be viewed on the interactive maps at http://gis.massdot.state.ma.us/maptemplate/topcrashlocations/.