Identification of Urban Scaling Behavior for Transportation Mode Share

Abstract

Urban scaling effects provide a means to formulate a general understanding of the processes that govern transportation system outcomes. Urban scaling analysis assumes that the quantity of interest follows a power-law distribution with respect to an urban area’s population. This distribution gives rise to a non-linear scaling exponent which then characterizes the scaling regime, either sublinear or super-linear where super-linear scaling reflects increasing returns with an increase in population while sublinear scaling represents economies of scale. This analysis approach has previously not been applied to the transportation system to study general features of urban transportation mode share, accessibility, and congestion. This work explores the effect of the selected geographic urban scale, population, and population density on the non-linear scaling exponent for these different transportation variables of interest. Notably, US transportation mode share exhibits remarkably general features which are sensitive to the urban scale, population, and population density. Single occupancy vehicle mode share follows a negative, sublinear scaling regime with respect to both population and population density. Transit mode share exhibits positive, sublinear scaling with respect to population and super-linear scaling with respect to population density. Non-motorized transportation modes experience positive, sublinear scaling for both population and population density. These results are strengthened by relating the findings to measures of transportation accessibility and congestion which are also expected to influence the observed transportation mode share. Finally, locations with unique mode share characteristics are identified to corroborate these findings with expectations and explore the effects of regional geography on transportation mode share.