Investigating the influence of spatial characteristics on cycling volume: A multi-scale geographic weighted regression approach

Abstract

Cycling has seen a remarkable rise, signifying a paradigmatic move towards sustainable, eco-friendly, and efficient commuting alternatives in the contemporary urban setting. Cities also encourage this trend by establishing cycle lanes, bike-sharing programs, and incentives for frequent riders. To enhance these motivations from an urbanistic perspective, it is essential to comprehend the influence of urban characteristics on cycling volume and to incorporate this understanding into data-driven decision-making processes.

This research examines the Bicification project data from Istanbul with a spatial perspective. Utilising a comprehensive array of spatial big data, the study explores the impact of urban land use, transport services, land morphology, and sociodemographic factors on cycling volume through a Multi-scale Geographically Weighted Regression (MGWR). With an Adj R² value of 0.68, the model demonstrates a strong relation between cycling volume and several factors, including biking park stations, park and ride points, pier stops, rail stops, transfer points, main roads, elevation, population, industrial facilities, health facilities, sports areas, and residential areas. The findings will serve to develop a data-driven strategic approach to promote cycling in Istanbul.