Diagnosing 15-minute community life circle service levels and simulating policy scenarios in a high-density urban core: Shanghai’s Huangpu District

The 15-minute community life circle is a neighborhood-scale framework for organizing daily service provision and assessing access to services in high-density urban cores. Yet existing studies have not sufficiently integrated service-level measurement at the residential community scale, identification of key predictors, and policy scenario simulation, limiting the translation of baseline assessments into planning practice. This study examines 1396 residential communities in Huangpu District, Shanghai, integrating public service facility locations, community boundaries, OSM road-network data, and census data. A network-based shortest-path method measures life-circle service levels, and operational network-distance thresholds of 500 and 800 m classify communities as A, B, and C, from highest to lowest service level. A C5.0 decision tree identifies key predictors and derives classification rules. Two predefined scenarios—transportation improvement and development intensity adjustment—then compare predicted changes in community categories and population coverage. Results reveal spatial variation in service levels. Category A communities cover 54.60% of the district’s population and tend to be concentrated in core commercial and waterfront areas; Category C communities cover 11.32% and tend to be concentrated in older residential areas in the northwest. The C5.0 model achieves an overall test-set accuracy of 81.74%. Road network density has the highest predictor importance (0.42), and transportation-related indicators have greater combined importance (0.53) than development-intensity indicators (0.19). Under the two scenarios, the Category A population share increases by 7.98 and 5.24 percentage points, respectively, whereas the Category C share decreases by 4.88 and 4.16 percentage points. Given the specified model, parameters, and spatial conditions, the transportation improvement scenario produces larger model-predicted changes. Decision-tree rules and scenario predictions may provide supplementary model-based information for identifying communities requiring priority review and comparing predefined policy scenarios in Huangpu District, while offering a methodological reference for life-circle diagnosis and planning review in high-density urban districts with similar built environments.