نوع مقاله : مقاله پژوهشی

نویسندگان

• جلال سمیعا ¹
• منوچهر رنجبر شوبی ²
• عامر نیک پور ³

¹ استادیار گروه جغرافیا، دانشکده علوم انسانی و اجتماعی، دانشگاه مازندران، بابلسر، ایران

² کارشناس پایگاه امداد و نجات جمعیت هلال احمر بابلسر

³ دانشیار گروه جغرافیا و برنامه‌ریزی شهری، دانشکده علوم انسانی و اجتماعی، دانشگاه مازندران، ایران

چکیده

احتمال وقوع تصادف در جاده کندوان به ­عنوان یکی از زیباترین و شگفت ­انگیز ترین جاده های توریستی کشور از دغدغه های مهم مسافران در بازدید از استان مازندران محسوب می شود. در این پژوهش، با استفاده از قابلیت های GIS، الگوی توزیع مکانی– زمانی تراکم تصادفات جاده­ای، الگوی توزیع مکانی– زمانی خوشه های تصادفات و سطح معنی­ داری آماری آنها با استفاده از روش های تخمین تراکم کرنل و تحلیل نقاط حاد مورد بررسی قرار گرفته است. به این­ منظور، داده های تصادفات جاده ­ای به ­وقوع پیوسته دربازه زمانی سال های 1401-1395 که توسط سازمان امداد و نجات جمعیت هلال احمر استان مازندران جمع آوری شده، مورد استفاده قرار گرفته است. نتایج حاصل­ از تحلیل توصیفی و آماری تصادفات جاده ­ای نشان می ­دهد که در این بازه زمانی، 2084 تصادف درامتداد این جاده به­ وقوع پیوسته که در آن 9076 نفر مصدوم و 52 نفر فوت شده ­اند. بیشترین تعداد وقوع تصادفات در سال های 1400 و 1401 بوده که نسبت به تعداد پایین وقوع تصادفات در سال های 1398 و 1399 به دلیل مشکلات مرتبط با بیماری کرونا و منع مسافرت، رشد قابل ملاحظه ­ای داشته است. علاوه بر این، نتایج حاصل از تابع تخمین تراکم کرنل، نشان دهنده تراکم بالای تصادفات در منطقه های پل زنگوله، تونل کندوان، سیاه بیشه، مجلار، پل اوشن و شاه چشمه در بازه زمانی 1401- 1395 است. همچنین نتایج بدست آمده از تحلیل نقاط حاد نشان می­ دهد که در منطقه های پل زنگوله، تونل کندوان، مجلار، سیاه بیشه و پل اوشن خوشه های مکانی تصادفات شناسایی با مقدار میانگین Z-score = 3.12 و سطح اطمینان 90-95 % شناسایی شده اند. نتایج این تحقیق می­ تواند در جهت شناسایی عوامل مؤثر بر شکل­ گیری خوشه های مکانی تصادفات و افزایش ایمنی حمل و نقل جاده­ای در محور کندوان مورد استفاده قرار بگیرد.

کلیدواژه‌ها

• GIS
• تراکم کرنل
• تحلیل نقاط حاد
• خوشه های تصادفات
• مازندران

موضوعات

• سیستم اطلاعات جغرافیایی

عنوان مقاله [English]

Spatio-temporal analysis of accidents along Kandovan road using Geostatistical methods

نویسندگان [English]

• Jalal Samia ¹
• Manouchehr Ranjbar Shoobi ²
• Amer Nikpour ³

¹ Faculty member, Department of Geography, Faculty of Humanities and Social Sciences, University of Mazandaran, Babolsar, Iran

² Red Crescent relief and rescue organization of Babolsar

³ Associate professor, Department of geography and urban planning, University of Mazandaran

چکیده [English]

Extended abstract
Introduction
Visiting Mazandaran province could be a fascinating and memorable trip due to its amazing natural touristic attractions such as Caspian Sea and mount Damavand. The three main roads naming Kandovan, Haraz and Firoozkooh can be used to access Mazandaran province. Among them, passing through Kandovan road is fascinating with its beautiful natural landscapes. At the same time, this road is also known as one of the most dangerous roads of Iran due to its mountainous location and the potential occurrence of different types of climatic and geomorphologic hazards. Apart from these dangers, the occurrence of accidents in Kandovan road is one of the main concerns of tourists visiting west parts of Mazandaran province and also the local governments providing relief and rescue services and facilities to injured people. Therefore, it is crucial to identifying the dangerous sections of this road in order to minimize fatalities and socio-economic losses. The purpose of this research is to investigate the spatio-temporal density pattern of road accidents and also to identify accidents clusters along Kandovan road. 
Material and methods
To this end, we used road accidents information along Kandovan road, collected by the relief and rescue bases of Red Crescent organization of Mazandaran province in the period of 2016 to 2022. Information like location, date, and the number of death and injuries in the road accidents along this road were used in this research. First, we used GIS, spatial and statistical analyses in order to get insight from road accidents distribution and statistics. In the next step, Kernel Density Estimation – a Geostatitical measure – was used to investigate the general spatial density pattern of road accidents in the period of 2016-2022 and also the spatio-temporal density pattern of road accidents in every year from 2016 to 2022. Furthermore, the hot spot analysis was implemented to the distribution of road accidents in this period in order to find out whether accidents are clustered, dispersed or randomly distributed. Both general spatial pattern and annual spatio-temporal patterns of accidents were investigated using hot spot analysis. With this, accidents clusters reflected as hot spots were identified based on the Getis-Ord Gi^*index and the associated Z-score, P-value and Gi-bin statistics. In this context, the number of accident clusters, the length of road in the accident clusters and the percentage of observed accidents in the clusters were computed from 2016 to 2022.  
Results and discussion
Results show that 2084 accidents were occurred in the period of 2016-2022 with 9076 injuries and 52 deaths. The most number of accidents was occurred in 2022 following the end of Corona lockdown in 2021. Also, several parts of Kandovan road indicated to contain the highest number of accidents density. Besides, the accident density pattern changes spatially and temporarily with an increasing trend in the number of accidents density from the end year of Corona disease epidemic in 2020. Results from hot spot analysis also identified several accidents clusters along this road in the period of 2016-2022. In this context, road accidents clusters were identified in Zangouleh Bridge, Majlar, Siah bisheh, Knadovan tunnel and Ushen Bridge with average Z-score value of 3.12, average P-value smaller than 0.05 and confidence interval of 90 to 99%. The total length of road in these clusters was more than 14 kilometer which contains around 60 % of the total accidents. The spatio-temporal distribution pattern of accidents clusters and also road lengths in the identified clusters change decreasingly in the period of 2016-2022. The results of this research can be used to investigate the reasons behind the occurrence of road accidents in the high accidents density sections and also in accidents clusters identified along the road. Taking proper preparation and mitigation strategies can be beneficial in proper crisis management of road accidents in order to avoid human causalities and socio-economic losses.
Conclusion
We conclude that kernel density estimation and hot spot analysis are effective geostatistical approaches to investigate the density pattern of road accidents and also to identify accidents clusters. In order to increase the safety of Kandovan road, the factors contributing to accidents occurrence in highly dense accidents sections of road and also in accidents clusters need to be identified, and with implementing proper measures, their effects can be minimized.

کلیدواژه‌ها [English]

• GIS
• Kernel density
• Hot spot analysis
• accidents clusters
• Mazandaran