Accessibility modeling and multiobjective trade-off optimization for sports facility site selection

Traditional population census data often fail to capture the temporal dynamics and spatial heterogeneity crucial for accurate sports facility site selection. To overcome this limitation, this study integrates optical satellite remote sensing (OSRS) imagery and nighttime light (NTL) data to enhance the precision of accessibility modeling and demand assessment. High-resolution OSRS imagery enables detailed extraction of urban morphology and transportation networks, while NTL intensity serves as a reliable proxy for socioeconomic activity and population mobility. By fusing these heterogeneous datasets, a dynamic spatial demand surface is generated, allowing the recalibration of demand point weights in accordance with real-world spatiotemporal variations. The optimized framework introduces a multi-objective trade-off (MOT) model that balances accessibility, equity, and cost-effectiveness through a Pareto-based evolutionary algorithm (PEA). Experimental validation across multiple urban regions confirms that the proposed OSRS–NTL-enhanced method significantly improves spatial coverage and reduces inequity in service allocation compared with conventional census-based approaches. The findings demonstrate that integrating optical and nightlight remote sensing provides a more accurate representation of human–environment interactions, offering a robust data foundation for dynamic, adaptive, and sustainable urban facility planning.