Environmental contingency of forest-cloud interactions: Contrasting atmospheric enhancement and suppression mechanisms across Chinese climate gradients

• Environmental moisture availability fundamentally determines forest-cloud interactions across climate gradients. • Forests enhance cloud formation by 2.24 % in temperate Northeast China but suppress by 1.78 % in subtropical Central China. • Causal inference through propensity score matching and structural equation modeling confirms forest effects rather than correlations. • Machine learning reveals forest cover as dominant predictor in moisture-limited regions (importance=0.39) versus precipitation in humid areas (0.32). • Moisture recycling and atmospheric saturation represent contrasting mechanisms governing regional forest-cloud relationships. Forest-cloud interactions exhibit striking regional variations that challenge universal assumptions about forest atmospheric services. Using satellite data (2000–2022) and multiple analytical approaches including correlation analysis, propensity score matching, structural equation modeling, and machine learning, we investigated forest-cloud relationships across moisture-limited Northeast China and moisture-abundant Central China. Northeast China demonstrated strong positive forest-cloud correlations ( r = 0.626, p < 0.001), while Central China showed weak negative correlations ( r =-0.040, p < 0.05). Causal analysis revealed forests enhance cloud formation by 2.24 percentage points in moisture-limited environments but suppress it by 1.78 percentage points in humid conditions. Machine learning confirmed forest cover as the dominant predictor in Northeast China (feature importance=0.39) versus precipitation in Central China (0.32). These findings demonstrate that environmental context fundamentally determines forest-cloud interactions, with implications for climate mitigation strategies requiring region-specific approaches.