@article { author = {Kalantari, Ali and Alian, Sahar}, title = {Analysis of road accidents with emphasis on environmental and road characteristics in geospatial information system Case study of Karaj-Kandovan axis}, journal = {Human Geography Research}, volume = {54}, number = {2}, pages = {563-582}, year = {2022}, publisher = {University of Tehran}, issn = {2008-6296}, eissn = {2423-7736}, doi = {10.22059/jhgr.2021.314926.1008216}, abstract = {Road traffic accidents are considered as one of the most important causes of human death in all over the world and in our country Iran. The increase of daily travels and traffic volume are the key factors of high number of traffic accidens in Iran. In a road transportation system, human behaviour, road and vehicle are the main contributing factors in traffic accidents. While human errors relating to driver performance are the cause of almost 50 to 95% of crashes, road and environmnal factors in different locations and times of driving may also affect driver performance and increase the risk of a crash. Identification of road hotspots is a geographical decision. GIS offer both spatial and temporal dimensions required for data preparation, spatial analysis, decision making and management of road crashes. The main objective of this research is the spatial analysis of casualty crashes and prioritization the available risk factors that contribute in crash incidence. The stady ares, is a part of Karaj-Chalous road, a main mountainous road connecting Karaj to North of Iran. In doing so, in a GIS contex, a decision tree method is used for crash analysis. The decision tree is a hierarchical knowledge structure that represents a sequence of decision rules. The purpose of the method is to determine which attribute or criterion of attributes provides the best distribution of the actual data set regardless of the value of the given attribute. According to the objective of the research, this research method was conducted in several stages. Firstly, available sources and previouse similar studies were reviewd, and the accident data points were collected. Secondly, the points were corrected based on the collected data source and judtified and implemented in kilometers. Thirdly, information about the slope and climate of the area, the intersections and used points around the study area were collected and the required layers were prepared and standardised. Next, prepared and standardized layers were processed and drawn by the decision tree model and weighted decision tree model. Then the prepared layers were combined and the resulting output probability map of the accidents along the study area was prepared. The results of studies showed that, according to the analysis, the most important effective factor in the occurrence of accidents is: curvature with a coefficient of 0.51. After that, the intersection layer with the coefficient of 0.21, and climate with the coefficient of 0.144 are the second and third factors respectively. Other variables such as density and slope are the next available probabibility factors of crash incidence in the study area. Based on the coefficients, a zoning map of the probability of accidents was prepared. The prepared map shows the high intensity of the probability of a crash in selected points. According to the obtained map, the highest risk area is at the entrance to the road and at points of intersection where the curvature of the road increases. The results of this research and prioritization of the crash risk factors in decreasing the costs of road safety improvement.}, keywords = {Road Accidents,Desition Tree,GIS,Kandavan}, title_fa = {تحلیل تصادفات جاده‏ ای با تأکید بر خصوصیات محیط و جاده در سیستم اطلاعات مکانی مطالعة موردی: محور کرج– کندوان}, abstract_fa = {در این مقاله به تحلیل توزیع مکانی و احتمال وقوع تصادفات جاده ‏ای و عوامل مؤثر بر آن‏ها به ‏ویژه دو عامل محیط و جاده در سیستم اطلاعات جغرافیایی پرداخته شده است. منطقة مورد مطالعه محور کرج- کندوان از محورهای کوهستانی ارتباطی تهران- شمال است. با استفاده از روش سلسله ‏مراتبی درخت تصمیم ‏گیری این پژوهش در چند مرحله انجام گرفت. در مرحلة اول منابع موجود مطالعه و نقاط تصادف جمع ‏آوری شد. در مرحلة دوم اصلاح نقاط صورت پذیرفت و نقاط نسبت به مبدأ جمع ‏آوری اطلاعات توجیه و به‏ صورت کیلومتر اجرا شد. در مرحلة سوم اطلاعات مربوط به شیب جاده، درجة انحنا، نقاط تقاطع، درصد تراکم، اقلیم، و کاربری اطراف منطقة مورد مطالعه جمع‏ آوری و سپس لایه‏ های اطلاعاتی ایجاد، آماده‏ سازی، و استاندارد شد. در مرحلة بعد، لایه‏ های استانداردشده توسط مدل درخت تصمیم و درخت تصمیم وزن‏ دار پردازش و ترسیم شد. سپس، با تلفیق لایه ‏ها نقشة احتمال وقوع تصادف در محدودة مطالعاتی تهیه شد. نتایج مطالعات نشان داد که مهم‏ترین عامل مؤثر در وقوع تصادف با توجه به تحلیل انجام ‏شده متغیر انحناست و متغیرهای تقاطع، اقلیم، تراکم، و شیب در اولویت ‏های بعدی قرار دارند.}, keywords_fa = {تصادفات جاده ‏ای,درخت تصمیم,سیستم اطلاعات جغرافیایی,کندوان}, url = {https://jhgr.ut.ac.ir/article_79953.html}, eprint = {https://jhgr.ut.ac.ir/article_79953_538a569e29e1cbcd7300f0a78f2cbeaf.pdf} }