Elevation Correction of Forest Biogeophysical Cooling Effect in China

Forests exert significant biogeophysical cooling effects (CE) through processes such as increased evapotranspiration, reduced albedo, and enhanced surface roughness. However, little is known about the extent to which elevation-induced temperature differences bias the observed CE and how this bias interacts with the underlying biogeophysical mechanisms. In this study, we integrated multi-sensor remote sensing products and Shuttle Radar Topography Mission (SRTM) elevation data on the Google Earth Engine (GEE) platform, and applied a spatial-temporal window regression approach to quantify and correct the sensitivity of land surface temperature (LST) to elevation for forest pixels across China from 2001 to 2022. First, we found that forest LST exhibited a significant negative relationship with elevation, leading to systematic CE overestimation by 0.61 K during the day and 0.60 K at night compared with altitude-corrected CE values. Second, after correction, the CE showed clear spatial heterogeneity, with stronger daytime cooling in tropical (−0.54 K) and temperate forests (−0.24 K), and warming in cold (+0.11 K) and arid regions (+0.53 K), while most regions experienced nighttime warming. Among forest types, evergreen needleleaf forests (ENF) exhibited the strongest daytime cooling (−0.36 K), whereas deciduous broadleaf (DBF) and open shrublands (OS) tended to warm. Third, mechanism analysis revealed that elevation correction strengthened the correlations of CE with leaf area index (LAI) and evapotranspiration, while maintaining a significant negative correlation with albedo, indicating that both radiative and non-radiative processes jointly shape the unbiased CE. These findings provide a more accurate quantification of forest CE by eliminating elevation-induced bias, which providing a more accurate assessment of the climate mitigation potential of forests, which is crucial for developing more effective forest management and ecological restoration strategies.