Using GIS to identify pedestrian high crash locations

Using GIS to identify pedestrian high crash locations

SHARE

Abstract

Shashi S. Nambisan
Professor of Civil Engineering & Director

Srinivas S. Pulugurtha
Assistant Research Professor & Assistant Director

Natachai Wongchavalidkul
Graduate Student

Vinod Vasudevan
Graduate Student

Transportation Research Center
University of Nevada, Las Vegas, 4505 S. Maryland Parkway, Box 454007, Las Vegas, NV 89154-4007
Tel.: (702) 895-1338; Fax: (702) 895-4401
Email: [email protected], [email protected], [email protected], [email protected]

Abstract
Moderate daily physical activities such as walking and bicycling are essential ingredients of a healthy lifestyle. However, in United States, studies have shown that the average number of walk trips and bike trips decreased by more than 40 percent during the last two decades. The decrease in physical activity is governed by several aspects, which include congestion and air quality, job-housing location imbalance, longer commuting times, diminishing open space and agricultural lands, absence of community life, urbanization, and, growth and development of the environment. Aside from this, residents of suburban developments often depend on their cars for trips to destinations within the neighborhood because of circuitous street layouts, long travel distances, lack of pedestrian safety, and lack of pedestrian facilities such as sidewalks, pedestrian crosswalks, and pedestrian signals. It is imperative that communities develop neighborhoods, which provide safe pedestrian and bicycling environments, more green space, network connectivity, and access to transit systems in order to control the growing pedestrian safety and physical activity related concerns.

Pedestrian high crash locations have to be identified to implement problem specific pedestrian safety countermeasures (such as mid-block pedestrian crossings, pedestrian activated flashing signals, smart lighting, in-pavement lighting, enlarged pedestrian heads, pedestrian countdown timers, etc.) so as to improve pedestrian safety. The focus of this paper is to develop Geographic Information Systems (GIS) based methodology to identify pedestrian high crash locations. The GIS-based methodology includes the following steps:

  1. Geo-code reported and recorded pedestrian crash data. Geo-coding may be done using street name and reference street name, milepost or address of a pedestrian crash in the database as a reference system.
  2. Overlay pedestrian crash data on the street network and identify corridors (a series of street segments) with a significant number of pedestrian involved crashes.
  3. Select pedestrian high crash locations along these corridors to conduct detailed analysis and implement countermeasure programs targeting specific locations on these corridors with high pedestrian crash problems.
  4. Estimate the number of pedestrian crashes in the vicinity of each pedestrian high crash locations.
  5. Rank the locations based on computed pedestrian crash indices (highest to the lowest).

Data obtained for the Las Vegas metropolitan area is used to demonstrate the working of the methodology. Since the study area primarily comprises of the Las Vegas metropolitan area, the pedestrian crash data were address matched using the street name and reference street name of crashes in the database and street centerline (SCL) coverage. A total of 3,710 pedestrian involved crashes were reported and recorded in the Las Vegas metropolitan area during the years 1996 to 2000. Approximately 95 percent of these pedestrian crashes were address matched. The pedestrian crash coverage is then overlaid over the SCL coverage to identify corridors with significant number of pedestrian crashes.

34 pedestrian high crash locations were then identified in the selected corridors. These 34 locations were selected such that they represent about one-third of total number of pedestrian crashes in the Las Vegas metropolitan area. Pedestrian crashes within 100 feet of each location and the potential population at risk in the vicinity of each location were identified using spatial analysis capabilities of ARC/INFO, a GIS program. Two different types of crash indices were computed based on the number of crashes in the vicinity of the location and population residing in the vicinity of the location.

Crash Index 1 is computed by dividing the percent of pedestrian crashes at a location by the percent of population residing in the vicinity of the location, and then multiplied by 100. The percent of pedestrian crashes at a location is estimated by dividing the number of pedestrian crashes in the vicinity of a site by total number of pedestrian crashes in the study area multiplied by 100. The percent of population in the vicinity of a location is estimated by dividing the population residing in the vicinity of a location by the total population of the study area. Crash Index 2 is computed by multiplying the percent of pedestrian crashes at a location with 100. It was felt that Crash Index 2 is more appropriate for ranking purposes based on the land-use characteristics of the study area. Hence, the selected 34 locations were ranked based on Crash Index 2.

The results obtained from the GIS-based methodology were promising and are as expected. The next step is to identify appropriate Intelligent Transportation Systems (ITS) and engineering related technologies to allocate safety funds for pedestrian safety improvements. These technologies should be based on potential underlying explanatory factors in order to enhance pedestrian safety in the region.