Time Series Analysis of Accidents in Chandigarh City

The rapid increase in the loss of life through traffic accidents has created a major health and development concern. The United Nation has declared 2011-2020, the Decade of Action for Road Safety. In this paper, identification of high frequency accident locations in Chandigarh city is carried out using Geographical Information System (GIS) technology and the dependency of the said accidents is analysed in terms of time as a variable. Four years of traffic data is collected from the Chandigarh traffic police in the city. All the accident points are plotted in the geographical co-ordinate digitized map of Chandigarh integrated to a tolerance of 100m. Subsequently, collective event converts the accident point data to weighted point data and gives the icount of coincident accident locations. The distance band required for the analysis is calculated using Incremental Spatial Autocorrelation. The weighted point data and distance band is then used in the analysis of hot spot (Getis-Ord Gi*), which identifies statistically significant hot spots and cold spots. The time dependency of the identified hotspot is further checked using one-way ANOVA. The analysis carried out, (F