Abstract: Woody plant encroachment (WPE) into grasslands can cool or warm the local climate by altering surface energy through biophysical processes. However, its global effects on land surface temperature (LST) remain elusive since the pervasive interactions between energy pathways, vegetation coverage and structure, and climate context have never been fully addressed. Here, using two decades of satellite observations and a temporally matched effect estimation approach, we demonstrate that WPE induced a net global cooling of daily LST by -0.016 ± 0.002℃ during 2003–2020. The effect is mainly mediated by three biophysical pathways: sensible and ground heat fluxes (H+G), surface albedo (SA), and latent heat flux (LE). Among all analyzed grid cells, 53% of LST changes were dominated by H+G, 40% by SA, and 7% by LE. In detail, H+G and LE reduced LST by -0.513 ± 0.015℃ and -0.068 ± 0.002℃, respectively, whereas SA increased LST by 0.565 ± 0.016℃. Changes of LST and biophysical pathways induced by WPE are closely regulated by vegetation coverage (fractional vegetation cover, fV) and structure (leaf area index, LAI), and climate, where air temperature determines LST change direction and precipitation modulates its magnitude. Specifically, in the warmer regions, WPE generally led to net cooling because of increases in fV and LAI that enhanced non-radiative cooling via increased LE and H+G, although radiative warming via reduced SA was amplified simultaneously. The opposite effects were observed in the cooler regions. Furthermore, the magnitudes of changes in fV, LAI, LST, and pathway-specific contributions tend to be greater in the wetter regions than in the drier regions. Our results underscore the critical role of vegetation coverage and structure in modulating the thermal effects of WPE and highlight the need for climate-dependent land management strategies to mitigate its heterogeneous impacts on surface temperature.
This R code for the research is primarily used to generate the figures for the article. The main code and data corresponding to each figure are clearly marked within the script.
Note: Figure 7 is a conceptual schematic diagram that was created using Microsoft PowerPoint, rather than being directly generated from the R code.