Bicycle-metro integration is an efficient method of solving the “last mile” issue around metro stations. Built environment is believed to have a significant effect on cycling behavior. However, transfer cycling around metro stations, as a specific type of cycling behavior, has often been overlooked in transport research. In addition, static contextual units such as circular or street-network buffers are typically used to delineate metro catchment areas of transfer cycling trips. These methods are inaccurate to represent the actual geographic contexts of cycling trips, according to the uncertain geographic context problem (UGCoP). Thus, in this study, bicycle-metro catchment areas are delineated based on aggregating the end points of over three million transfer cycling trips. The impact of the built environment on transfer cycling behavior is also explored.

First, we find that the aggregate-points buffer outperforms traditional static buffers in predicting transfer cycling trips. Second, we also identify a high level of spatial heterogeneity in catchment area and transfer cycling density between urban and suburban areas. Third, residential and working population density, bus stop density, and metro stations accessibility have a significant effect on bicycle-metro transfer cycling.

自行车-地铁一体化是解决地铁站周边“最后一公里”问题的有效方法。建筑环境被认为对骑行行为有显着影响。然而，绕地铁站换乘单车作为一种特殊的骑行行为，在交通研究中经常被忽视。此外，静态上下文单元（例如圆形或街道网络缓冲区）通常用于描绘换乘自行车旅行的地铁集水区。根据不确定的地理环境问题（UGCoP），这些方法不能准确表示自行车旅行的实际地理环境。因此，在本研究中，自行车-地铁集水区是根据超过 300 万次换乘骑行的终点的聚合来划分的。

首先，我们发现聚合点缓冲区在预测转移循环行程方面优于传统的静态缓冲区。其次，我们还确定了集水区的高度空间异质性以及城市和郊区之间的转移循环密度。第三，居住和工作人口密度、公交车站密度和地铁站可达性对自行车-地铁换乘循环有显着影响。