Quantifying blue-green space cooling thresholds in a subtropical water-network city

This study examines Nanchang, a typical subtropical water-network city, to quantify the nonlinear impacts of blue‑green space (BGS) patterns on the urban heat island (UHI) effect and identify critical thresholds. Using multi‑source geographic data and a random forest model, it reveals: BGS proportion (BGP) exhibits cooling thresholds at 27.05%, 39.49%, 60.74%, and 71.46%, with optimal cooling at 73.32% and benefits saturating beyond 74.90%; the largest patch index of BGS (LPI_BG) must exceed 13.18 to be effective; building density (PBPA) above 19.34% sharply intensifies UHI; while increasing building height variation (AUCHR > 2.25) and BGS aggregation (AI_BG > 81.44) enhance cooling. Recommendations include setting neighborhood BGP targets within 60%–74% and strictly controlling building density below 20%. Planning should shift from “area expansion” to “structural optimization,” focusing on continuous, aggregated BGS patches and coordinating three‑dimensional building morphology to improve ventilation. The framework is climate‑adaptive: in high‑humidity environments, higher BGS aggregation and connectivity compensate for constrained transpiration; in regions with pronounced prevailing winds, blue‑green corridors aligned with wind direction provide stronger dynamic cooling, allowing threshold adjustments. The study uncovers “ineffective green space” in subtropical cities, emphasizes multi‑factor coupled regulation beyond single area metrics, and provides empirical evidence for climate‑adaptive planning in hot‑humid regions.