Contrasting long-term trends, dynamics and driving mechanisms in trees' transpiration and their associated cooling across global biomes over 1995–2018

Recent evidence suggests that record-breaking temperature rise is a major indicator of continuing global climate change. However, the lack of decades-long continuous field data on trees' responses to background climate constrains our comprehension of how climate change influences global patterns and the determinants of trees' capacity to mitigate local climate through cooling effects. Using SAPFLUXNET datasets from 1995 to 2018, we assessed long-term trends, temporal dynamics, and key driving parameters influencing trees' transpiration (T r ) and their associated cooling (i.e., heat energy absorption (Q) and air temperature reduction (ΔT)) across multiple global biomes. Annual T r , Q, and ΔT exhibited an increasing trend in tropical rain forests (TRF) and temperate forests (TF), with ΔT slopes of 0.27 and 0.12, respectively, while decreasing trend in tropical forest savannas (TFS), subtropical deserts (STD), woodlands/shrublands (W/SL), and temperate grassland deserts (TGD), with ΔT slopes of -0.23, -0.06, -0.05, and -0.07, respectively. This can be attributed to the corresponding increase in both air temperature (T a ) and soil water content (SWC) in the biomes of TRF and TF, while T a rises but SWC declines in the biomes of TFS, W/SL, STD, and TGD. Moreover, seasonal trends in T r , Q, and ΔT are shaped by spring T a , summer SWC, and autumn vapor pressure deficit. The ΔT variations were mainly controlled by T a in the biomes of TRF and TF, while precipitation and SWC drove it in other biomes. Variations in ΔT sensitivity to hydrometeorological factors through biomes are associated with stomatal regulation, exhibiting reduced ΔT sensitivity to T a and heightened sensitivity to SWC in the biomes of TFS, W/SL, and desert, which could point to plant adoption of a water-saving strategy. These findings indicated that long-term transpiration patterns in global biomes have considerable biophysical effects on local climate mitigation and improving the universal model's representation of hydrometeorological constraints.