Analysis of the interactive effect of built environment on online ride-hailing travel distribution considering temporal and spatial heterogeneity

Understanding how the built environment influences the spatial distribution of ride-hailing trips is essential. However, the interactive effects among built environment factors remain insufficiently explored. Using ride-hailing trip data and multi-source geospatial information from Haikou City, this study applies the Optimal Parameter Geodetector (OPGD) to examine the impact of the Modifiable Areal Unit Problem (MAUP) on ride-hailing travel modeling and to quantify interactions among factors. The results indicate that the proposed mixed land use calculation enhances model interpretability across different spatial scales compared with conventional approaches. Socioeconomic attributes and road density exhibit relatively weak independent effects on ride-hailing travel, but their explanatory power increases substantially when interacting with public management and commercial services. These findings provide empirical evidence to support urban planning and the optimization of online ride-hailing services.