LAUSR

Abstract Taxicab is one of the critical components of the urban mobility in most metropolitan areas on the world, especially in the emerging countries. Last year in Shanghai, the total 50,000 taxicabs carry 2.8 million passengers and travel 17.1 million kilometers every day. The tremendous amount of the stochastic and independent trips across the city makes the taxi problem more complex than transit such as bus and metro. Multi-source data over space and time provide an efficient way to explore the taxi ridership and travel pattern. In this study, we first summarize the historical operation records of Shanghai taxicabs, and then compare the temporal and spatial variation patterns of the taxi trips in Shanghai and New York City using GPS data. A regression model is developed to investigate the relationship between ridership and variables including urban land use, resident population, employment and car ownership. The results of this study may first improve the transportation model, and also benefit the efficiency of taxi operation and policy making.