Vegetation–Atmosphere–Land Interactions Driven by Precipitation Extremes in Northeast China

Climate change is increasing the frequency and intensity of extreme rainfall events, profoundly affecting vegetation–atmosphere–soil interactions and ecosystem stability. Northeast China (NEC), a major ecological region, is highly sensitive to precipitation variability. However, the annual mechanisms underlying vegetation responses to rainfall extremes, the mediating roles of soil moisture (SM) and vapor pressure deficit (VPD), and the ecosystem-specific differences remain insufficiently understood. This study investigates these processes during 2000–2022 by integrating precipitation extremes, normalized difference vegetation index (NDVI), SM, VPD, and land cover data. Ten rainfall extreme indices were evaluated using the Mann–Kendall (MK) test and Sen’s slope estimator, while NDVI responses were examined through correlation analysis, mixed-effects models, and structural equation modeling (SEM). Results show strong spatial heterogeneity in precipitation extremes, with intensified heavy rainfall in southern NEC and prolonged drought conditions in northern areas. Vegetation exhibited significant greening trends (NDVI slope = 0.0026 yr−1, R2 = 0.718, p < 0.001), accompanied by increasing SM (slope = 0.0478 yr−1, p = 0.003) and mild warming (slope = 0.0005 yr−1, p = 0.045). NDVI showed a strong correlation with SM (ρ = 0.65, p < 0.01) but a weak relationship with temperature (ρ = 0.04, p > 0.05), highlighting SM as the dominant driver of regional greening. Grasslands and cultivated lands were more sensitive to rainfall fluctuations, whereas forests showed greater resilience. SEM results indicate that extreme rainfall affects NDVI mainly through indirect pathways mediated by SM and VPD, with mediation effects exceeding 97%. These findings improve understanding of nonlinear vegetation–atmosphere–land interactions and provide scientific insights for climate adaptation, ecosystem management, and ecological restoration under future climate change.