Bowling Green Traffic Accident Study

Bowling Green Traffic Accident Study

Summary:

In year 2004, there are 42,636 fatalities due to traffic accident in U.S. and 964 of them are from Kentucky. The fatality rate in Kentucky (23.25 per 100,000 person) is significantly higher than the national average (14.52 per 100,000). As a matter of fact, there are a total of 157,232 accidents reported alone in Kentucky and the direct economic cost mounted to $3.114 billions. In order to reduce traffic accident and improve road safety, it is very important to know how, where and when traffic accidents happened. Contrary to the common sense that the occurrences of traffic accidents are rare and random in space and time, and are mainly subject to human behavior, traffic accidents are indeed anticipated to form clusters (commonly known as “hot spots”) in the geographic space and over time. This is because their occurrence is usually tied to traffic volumes, which themselves exhibit distinct spatial and temporal patterns. For instance, some places with high volume traffic are more likely to have high traffic accidents. Another group of factors behind traffic accident include natural environmental characteristics such as weather (snow and fog), the configuration of highway networks such as locations of access and egress points, and the deficient design and maintenance of highways. Therefore, a good understanding of the spatial and temporal distribution of accidents makes a considerable contribution to developing appropriate accident reduction programs. For instance, people who are aware of being observed tend to modify their behavior. This phenomenon could then be used to encourage individuals to behave more safely when driving if we can provide feedback to the drivers about their behavior on the road. Therefore knowing where and when traffic accidents tend to occur, law enforcement can conduct more efficient patrols while highway departments can disseminate to drivers more effectively the critical information about roadway conditions.

In this study, we plan to study the distribution of traffic accidents in Bowling Green, Kentucky. We will first geo-code all traffic accidents from 2000 to 2005 in Geographic Information Systems (GIS). Then we will analyze accident patterns in space and time, including the identification of “hot-spots” of high accident occurrence. This will be done using the combination of geographic maps and advanced spatial statistical methods. Lastly, we will study the changes of traffic accidents patterns to find out the factors that cause the shift. The findings of this study are expected to be used for better decision making in Bowling Green and State Police departments and assist these law-enforcement agencies to develop more effective accident reduction programs.

Steps:

Acquire current and historical traffic accident data
Geo-coding traffic accidents in GIS
Analyze traffic accident patterns in space and time, e.g. identify hot-spots for different types of traffic accidents. This will be done using geographic maps and some advanced spatial statistical methods.
Analyze the factors behind and performed nonlinear fit and analysis of traffic accident data based on traffic flow, climate, road geometric characters
Study the changes of traffic accident pattern in the past 5 years

Preliminary results:

Summary Table

Year	Total	Injury	Fatality
2002	3604	Accident #: 732 (20.3%) Person #: 1081	Accident #: 7 Person #: 7
2003	3392	Accident #: 622 (18.3%) Person #: 957	Accident #: 1 Person #: 1
2004	3373	Accident #: 633 (18.8%) Person #: 898	Accident #: 9 Person #: 9
2005	3288	Accident #: 543 (16.5%) Person #: 788	Accident #: 4 Person #: 5