Understanding the Spatial Alignment Between Public Rental Housing and Metro: Evidence from Zhengzhou, China

The population for each PRH community was estimated by multiplying the number of households by the citywide average household size of 3.46 persons per household, obtained from Zhengzhou's Seventh National Population Census. While we acknowledge the limitation of not having household-level census data for PRH, our field survey provided ancillary support for the population estimation framework. Through informal conversations with property management staff at 22 of the surveyed communities, we confirmed that the reported number of households for each community was consistent with management records. This at least verifies the reliability of the 'household count' component of our population estimation.Metro network and passenger flow data. Data on metro lines and station locations were sourced from the AMap application programming interface for October 2024, ensuring alignment with the operational network during the study period. Passenger flow data was provided by the Zhengzhou Metro Corporation and consists of anonymized smart card transaction records from September 1 to 30, 2024. These records contain card IDs, entry and exit station names, and timestamps. From this, we extracted the inbound passenger flow for each station during the weekday morning peak period (7:00–9:00 AM) and calculated the average daily inbound flow for this period for each station. This passenger flow data is critically used in our model to quantify station-level congestion and its impact on waiting time.Transportation network data. The road network data for Zhengzhou was sourced from the 2023 version of the electronic map database. It includes vector data for primary roads, secondary roads, branch roads, and inter-urban highways. This dataset provides the fundamental network structure for conceptualizing travel paths.

各PRH社区的人口规模通过将家庭户数与全市平均家庭规模（每户3.46人）相乘估算得到，该平均家庭规模取自郑州市第七次全国人口普查数据。尽管我们意识到本次研究未获取PRH社区的家庭级普查数据这一局限性，但我们开展的实地调研为本次人口估算框架提供了辅助支撑。通过与22个受访社区的物业管理人员进行非正式交谈，我们确认各社区上报的家庭户数与物业记录相符，这至少验证了本次人口估算中“家庭户数统计”部分的可靠性。

地铁网络与客流数据。地铁线路及站点位置数据取自2024年10月的高德地图（AMap）应用程序编程接口，确保与研究时段内的运营网络保持一致。客流数据由郑州地铁集团（Zhengzhou Metro Corporation）提供，为2024年9月1日至30日的匿名智能卡交易记录，包含卡号、进出站名称及时间戳。我们从中提取了各站点工作日早高峰时段（7:00–9:00）的进站客流，并计算了该时段各站点的日均进站客流量。该客流数据是本模型量化站点级拥堵及其对等待时间影响的核心依据。

交通网络数据。郑州市道路网络数据取自2023版电子地图数据库，包含主干道、次干道、支路及城际高速公路的矢量数据。该数据集为出行路径的概念化建模提供了基础网络结构。