Prediction method development for metro station and exit ridership based on the surrounding urban built environment evaluation

The integration of metro stations into urban spaces is the key to achieving transit-oriented development. Previous studies have focused on the driving factors of ridership at metro stations, but have lacked consideration of exit design based on the surrounding urban built environment. In this research, 23 typical stations of different types in Xi’an were selected to investigate the influence of the built environment on station and exit ridership. The vector model was established with an 800 m circular buffer around the station. Based on the model, 21 built environment indicators within the 800 m buffer area were calculated for the station. Similarly, 21 built environment indicators were also calculated using the sector where the exit is located as the buffer area for the exit dimension. Afterwards, six typical time periods were selected for passenger flow monitoring, and 19 indicators were calculated for both dimensions. Finally, partial least squares regression models were constructed and the contribution of different built environment factors to ridership was quantified. The models could explain 38–83% of the ridership variation in the station dimension and 16–51% in the exit dimension. Commercial and official functions had the most significant impact on ridership. The method for predicting ridership at stations and exits during different typical periods could guide the planning and design of metro stations and their surroundings.