Modeling spatial determinants of urban expansion of Siliguri a metropolitan city of India using logistic regression

Abstract

Siliguri, a metropolitan city of West Bengal, has been experiencing extensive and rapid urban outgrowth from the last 2 decades. This tremendous urban expansion leads to the loss of natural landscape, agriculture lands, forest cover, and creating problems to run urban utility services effectively in expanded areas. Spatiotemporal assessment and modeling of urban expansion are very crucial as well as helpful for better management of sprawl areas. Therefore, in the present investigation, we have studied the responsible driving factors for urban expansion of the Siliguri metropolitan area form the period 1991–2017 with the help of binary logistic regression using random and stratified sampling. Sixteen independent variables have been included in this model, and these are elevation, slope, distance to the forest distance to the river, distance to agriculture, land value, proximity of road, distance to rail, proximity of old city, proximity of education, proximity of medical, proximity of utility services, built-up density, distance to the canal, population density. This research shows that over the past 2 decades, the built-up area has been expanded rapidly in the town. Results obtained from the model explain that elevation, the proximity of the major road, land value, the proximity of education center, medical center are the most important factors of urban expansion from 1991 to 2017. Interpolated probability map obtained from the model shows that most urban expansions will take place nearby the old urban areas and along the major roads in the southwest direction. Edge expansion is a dominant process rather than infill development in the area. The area under curve of receiver operating characteristics is 0.88 that specifies the predicted probability surface of the urban growth is correct and the model is valid.