Date of Award

6-2013

Degree Name

Master of Science in Engineering

Department

Civil and Construction Engineering

First Advisor

Dr. Jun-Seok Oh

Second Advisor

Dr. Valerian Kwigizile

Third Advisor

Dr. Joseph W. McKean

Fourth Advisor

Dr. Ron Van Houten

Access Setting

Masters Thesis-Open Access

Abstract

This research investigates spatial factors, that impact non-motorized exposures and crashes, and examines the relationship between exposure at signalized intersections and crash frequencies involving pedestrians and bicycles at a census tract level. The four Michigan cities of Ann Arbor, East Lansing, Flint and Grand Rapids are examined. After completing an extensive literature review to provide a list of required data, 92 intersections are selected from the four cities to collect pedestrian and bicyclist counts. Five-year crash records, special attractions, non-motorized facilities and transportation infrastructure data for the four cities are gathered for the four cities. This data are later mapped and processed by ArcGIS for further analyses. Buffer zones are produced at signalized intersections (quarter miles for pedestrians and half miles for bicycles) to create log-linear models to estimate and verify the number of pedestrians and bicycles. Next, negative binomial models are developed to predict non-motorized crashes at a census tract level to identify influential and correlation factors of crashes with exposure. The results show a correlation between exposure and crashes. They also indicate that land use, certain age groups, income, home-based work mode choice, access points, public transportation and special attractions have an influence on non-motorized exposure and crashes. Based on these findings, the importance of selection of appropriate crash risk measures is discussed.

Share

COinS