Tree shade as a nature-based strategy for mitigating heat exposure but effectiveness varies

Tree shade is widely recognized as an effective nature-based cooling solution to mitigate thermal exposure under a warming climate. However, factors modulating the intensity of shade-related cooling remain poorly understood, particularly beyond urban settings where most prior studies have focused on individual tree traits and local land use. This study examines whether shading effect varies across different landscapes and identifies key temporal and spatial drivers of shade-induced cooling across three survey sites: urban, post-mining, and lakeside environments in Lusatia, Germany. More than 100 trees were assessed for their shading effects. Surface temperature of shaded and adjacent non-shaded surfaces were measured using a handheld thermal camera during heat events in August 2023 and 2024, when daily maximum temperatures exceeded 30°C. Land-use information was derived from field-collected RGB imagery. Additional variables including distance to water and forest, vegetation index and canopy height were extracted from remote-sensing datasets. Shading effects were quantified using paired statistical tests, and an XGBoost regression model combined with post-hoc interpretability analyses was applied to identify key predictors and their influence on cooling intensity. Across all survey sites, shaded surfaces were significantly cooler than non-shaded surfaces, with non-urban areas exhibiting larger cooling effect. The predictive model achieved moderate performance (R² = 0.34). Temporal factors, particularly year and time of day, emerged as the most influential predictors, indicating substantial temporal variability in shade-induced cooling. Spatial configuration also played a critical role: shade-induced cooling increased with distance from forested areas and decreased with distance to water bodies. The relative importance of spatial variables varied by landscape type. Canopy height showed a negative relationship with cooling magnitude, suggesting that areas dominated by shorter trees may enhance shading effectiveness. Vegetation greenness and land-use categories had comparatively minor effects, while landscape type itself exerted no substantial influence. These findings demonstrate that shade-related cooling is governed not only by local tree or land-use characteristics but also by broader environmental context, including surrounding vegetation and landscape configuration. Incorporating multiscale geospatial predictors into microclimate assessments can therefore improve the design of climate-resilient landscapes and heat-mitigation strategies.