Tree restoration can cool or warm the local climate through biophysical processes. However, the magnitude of these effects remains unconstrained at large scales, as most previous observational studies rely on land surface temperature (Ts) rather than the more policy-relevant air temperature (Ta). Using satellite observations, we show that Ta responds to tree cover change at only 15-30% of the magnitude observed in Ts. This difference is supported by independent evidence from site observations, and can be attributed to the reduced aerodynamic resistance and the resultant flatter near-surface temperature profiles in forests compared to non-forests. At mid- or high-latitudes, the maximum seasonal biophysical Ta warming or cooling only accounts for approximately 10% of the equivalent climate effect of carbon sequestration in terms of magnitude, whereas the biophysical Ts effect can reach 40%. These findings highlight the importance of selecting the appropriate temperature metric in different applications to avoid exaggerating or underestimating the biophysical impacts of forestation.
© 2025. The Author(s).