Accident Analysis and Generating Prediction Maps of Crash Spots using GIS

Recognizing critical crash hot spots is a key role in decreasing accident probability patterns and has an important place in using acceptable methods on street networks and junctions. For this purpose, determining efficient criteria, individually and socially, based on related mathematical operators can identify and calculate different influence values for each criterion. As a consequence, various theoretical and practical models can be guarantee safety and improve transporting conditions of roads networks. From this, the main accident factors can biologically be identified at cross roads. In this paper, GIS capabilities are used to estimate and predict the probability of crash spots at cross roads and consider safety guarantees in the transportation system of Mashhad. Statistical data obtained from traffic observations of the transportation network of Mashhad are collected in question forms and prepared using mathematical and statistical processes such as density estimation and sample interpolation methods, based on the predefined standards and rules. In addition, probability values of accidents are evaluated and assessed using an integrated pattern of accident estimation based on expert opinions of each junction. Finally, Simple overlay, IDW and Kriging (with different neighbor points) are compared to recode values of crash numbers, and after producing the probability map of accidents, modeling validation can be determined in the control points. Presenting the Kriging method using 7 neighbors can improve the transportation system and will present an efficient solution for managing and controlling accidents.