Key points are not available for this paper at this time.
Background Fractal theory, developed by Benoit B. Mandelbrot in 1983, serves to describe intricate patterns in nature that traditional geometry finds challenging to measure. This theory employs the fractal dimension as an index to gauge the complexity and connectivity of patterns in response to changes in geographic scales. Recent research has applied this theory to analyze the road network structure, with compelling evidence suggesting that analyzing the road network using fractal theory offers a more insightful understanding of it. Analyzing road networks through fractal theory reflects their morphological complexity and spatial organization. Urban road networks with pronounced fractal features often exhibit well-developed branching structures and enhanced accessibility. While previous studies have utilized fractal theory to understand road networks, none have explored its connection with road safety. This study aims to fill this gap by employing fractal theory to quantify road network structure and find its association with road safety. Objectives This study aims to: a) introduce a methodology for evaluating the connectivity of road network structures using fractal theory and b) create a safety model that explores the association between road traffic crashes and the fractal dimension. Method This study utilizes the geometrical fractal dimension (Dg) to assess the uniformity of road network distribution, which is an indicator of connectivity. The box-counting method is employed for Dg calculation, facilitated by a developed geoprocessing Python script. Applying this methodology, the study estimates Dg for Delhi's road network at the ward level, utilizing OpenStreetMap data. The analysis covers 289 municipal wards, with crash data extracted from police records for 2017–19, encompassing 3765 fatal crashes. Geographically weighted Poisson regression is then employed to model ward-level crash rates, considering variables such as Dg, nightlight intensity (an economic status indicator), and population density. Results and Conclusion The results indicate a positive association between road network connectivity (higher Dg) and crash rates, implying that wards with a more interconnected structure witness more crashes. Increased connectivity tends to result in more vehicles taking shortcuts through neighbourhoods, impacting road safety. However, a negative association was observed with nightlight intensity, suggesting that economically prosperous wards experience lower crash rates.
Nazir et al. (Fri,) studied this question.