An empirical Bayes model to assess deer-vehicle crash safety in urban areas in Iowa

Publication Year:
2010
Usage 379
Downloads 356
Abstract Views 23
Repository URL:
https://lib.dr.iastate.edu/etd/11398
DOI:
10.31274/etd-180810-451
Author(s):
Baird, Michael James
Publisher(s):
Iowa State University; Digital Repository @ Iowa State University
Tags:
Deer-Vehicle Crash; Empirical Bayes; Traffic Safety
thesis / dissertation description
Deer-vehicle crashes are a growing problem in Iowa. In 2008, deer-vehicle crashes represented 12% of all the crashes reported, which include 9 fatalities and 442 injuries. This is especially true in urban areas of Iowa, where the problem has been increasing. There has been quite a bit of research conducted on countermeasure action that could help solve this problem. However, there has been little previous work that attempted to model deer-vehicle crashes in urban areas using the two data sources available: deer carcass salvage reports and deer-vehicle crash reports. The objective of this thesis is to assess the safety of roadway segments using both deer-vehicle crash and deer carcass salvage data in an empirical Bayes model to predict crashes in select urban areas of Iowa.In this thesis, three cities were selected with long-running deer management programs for evaluation. Data were collected from both the deer-vehicle crash and carcass salvage data bases. Records were reconciled to help eliminate double counting. Count data models were estimated that examined crash frequency as a function of roadway and environmental factors. The count model estimates were used to develop safety performance functions as part of an empirical Bayes analysis to assess the safety of sections of state-maintained roadway. Results were discussed, limitations were examined, and recommendations were made for future work.