Ma, Bingxin ， Li, Longhui ， Zhang, Yue ， Dou, Yueyue
2026-05-01 GLOBAL AND PLANETARY CHANGE 2026 260(卷), null(期), (null页)
Both land and atmosphere drought can lead to changes in vegetation structure, but their relative contributions are still poorly understood. In this study, eight machine learning algorithms were used to investigate the relative contributions of soil moisture (SM) and vapor pressure deficit (VPD) to the leaf area index (LAI) from 1982 to 2020 in the Northern Hemisphere. The results revealed that the relative contribution of SM to LAI decreased with aridity index (AI) (from 0.63 +/- 0.09 in the arid region to 0.24 +/- 0.09 in the humid region). However, the relative contribution of VPD to LAI increased with AI (from 0.37 +/- 0.09 in the arid to 0.76 +/- 0.09 in the humid region). The relative contribution of SM to LAI was further enhanced when the combined effects of time lag (TL) and time accumulation (TA) were considered, but the relative contribution of VPD to LAI was significantly attenuated in the Northern Hemisphere except the arid region (AI <0.2). In the current context of continuously decreasing SM and increasing VPD, the inferred seesaw effect of land and atmosphere drought on structural changes would help predict the vegetation response to future climate changes more accurately.