Measuring the Destination Accessibility of Cycling Transfer Trips in Metro Station Areas: A Big Data Approach.

Cycling is a green, sustainable, and healthy choice for transportation that has been widely advocated worldwide in recent years. It can also encourage the use of public transit by solving the "last-mile" issue, because transit passengers can cycle to and from transit stations to achieve a combination of speed and flexibility. Cycling as a transfer mode has been shown to be affected by various built environment characteristics, such as the urban density, land-use mix, and destination accessibility, that is, the ease with which cyclists can reach their destinations. However, cycling destination accessibility is loosely defined in the literature and the methods of assessing cycling accessibility is often assumed to be equivalent to walking accessibility using the same decay curves, such as the negative exponential function, which ignores the competitive relationship between cycling and walking within a short distance range around transit stations. In this study, we aim to fill the above gap by measuring the cycling destination accessibility of metro station areas using data from more than three million bicycle-metro transfer trips from a dockless bicycle-sharing program in Shenzhen, China. We found that the frequency of bicycle-metro trips has a positive association with a trip distance of 500 m or less and a negative association with a trip distance beyond 500 m. A new cycling accessibility metric with a lognormal distribution decay curve was developed by considering the distance decay characteristics and cycling's competition with walking. The new accessibility model outperformed the traditional model with an exponential decay function, or that without a distance decay function, in predicting the frequency of bicycle-metro trips. Hence, to promote bicycle-metro integration, urban planners and government agencies should carefully consider the destination accessibility of metro station areas.