Determination of traffic hotspots using spatial statistical methods: Case study Afyonkarahisar-Konya

Traffic accidents became a national problem in our country cause economic and social losses more than natural disasters. The results of the traffic accidents have alarming dimension both economical and human. Beside the results such as death and economic losses, they cause the mental effects as reminding the accident during the life by occurring disability. It is necessary to find practical and economic solution methods for resolving these problems and reduce losses. The most important of these methods is identify and improvement of dangerous segments which are called as crash hotspots in traffic accident. According to the development of computer technology, Geographical Information Systems (GIS) also has developed and these systems have been started to use in the analyses of crash hotspot as currently being used in many areas today. Because of the accident data and road information are based on spatial data they can be analyzed with GIS in various types. In this study, the accidents happened between 2005-2011 years in Afyonkarahisar-Konya roadway was investigated using the spatial statistic methods of Kernel Density Estimation, Getis Ord Gi* and Moran?s I. Accidents were investigated according to their numbers, road geometry, numbers of death and injuries and their happening time in the day. It was tried to determine advantages and disadvantages of these three spatial statistic methods based on the related data. Key Words: Kernel Density Estimation, Getis-Ord Gi*, Moran?s I, Hotspot