Road network structure of Varanasi city based on connectivity and spatial pattern

Better road network connection and coverage result in better transport, which is the fundamental foundation of urban growth. This paper is focussed on the assessment of connectivity, coverage and spatial pattern of the road network of Varanasi city. Connectivity has been analysed with the help of nine indices—alpha index, beta index, gamma index, cyclomatic number, aggregate transportation score, eta index, grid-tree pattern, connected node ratio and completeness of network. Coverage has been measured in terms of network density and intersection density. Fractal dimension has been utilised to reveal the spatial pattern of road network. The road network shows tree or spinal pattern, indicating very poor connectivity in Varanasi city. The wards near the core of the city show better connectivity and coverage while the peripheral wards have poorer connectivity. Fractal dimension values show opposite pattern as the wards at the edge of the city have more regular network compared to the wards near central business district (CBD). The correlation analysis shows statistically significant correlation between the connectivity, coverage indices and fractal dimension. Two regression models- ordinary least square (OLS) and geographically weighted regression (GWR) have been applied to show interaction between connectivity and spatial pattern. GWR with adaptive kernel type is the most suitable model with R² value 0.50. Cluster analysis including all the indices shows four types of clusters in the city. The highest network development is seen in the cluster acquiring the city core and it gradually lowers down towards the edge.