Threshold responses of land surface temperature to local moisture recycling in a mountain-lake-forest composite system

Quantifying water-heat coupling in mountain-lake-forest composite systems is important for alleviating surface heat stress in low-latitude mountain tourist cities, yet existing indicators fail to resolve nonlinear threshold transitions between local recycled water vapor contribution ratio( f local ) and land surface temperature(LST). This study proposes a threshold-based indicator approach by integrating stable isotope end-member analysis with multi-scale land surface temperature remote sensing, and applies it to three subsystems (mountain forest-grassland, basin composite, and lake-water) in the Erhai Basin. The Erhai Basin is dominated by advected water vapor in the rainy season(57.90%) and local recycled moisture in the dry season(61.10%). Distinct thermal response regimes emerge across the three subsystems. The mountain forest-grassland system shows significant negative correlations when f local < 30% in the rainy season and < 45% in the dry season; the basin composite system switches from positive to negative correlation at approximately 50% in the rainy season and exhibits strong negative correlation in the dry season; and in the lake water system, advection water vapor masks the local evaporation cooling signal during the rainy season, while sustained lake-surface evaporation maintains stable heat-sink functionality during the dry season. These divergent thresholds reflect differential synergies between surface evaporation and sub-cloud secondary evaporation across ecosystems. This demonstrates that peripheral lake-forest composite ecosystems can effectively alleviate regional surface heat through local moisture recycling and latent heat exchange, offering a nature-based pathway for thermal management in such composite systems, in which subsystem thresholds provide quantitative evidence for heat mitigation strategies.