Driving mechanisms of the decoupling between vegetation greenness and ecosystem resilience in the Yellow River Basin of Inner Mongolia, China (2000−2020)

The asynchronous changes in vegetation greenness and resilience (the capacity of ecosystems to absorb disturbances and maintain their functions) may threaten the long-term sustainability of ecological restoration. However, their coupling mechanisms and responses to climate change and human activities in desertification areas remain poorly understood. Here, we used the kernelized Normalized Difference Vegetation Index (kNDVI) and its temporal autocorrelation to characterize vegetation greenness and resilience from 2000 to 2020 in Yellow River Basin in Inner Mongolia, China. Random forest (RF) models were employed to quantify the dominant drivers of vegetation resilience. Results indicate a persistent greening trend across the study region, 96.62% of pixels experienced kNDVI increases. Meanwhile, resilience exhibits a pronounced turning point around 2011, transitioning from an increasing to a declining trajectory, while the proportion of resilience-declining areas increased from 18.09% in 2000–2010 to 45.85% in 2011–2020. The greenness–resilience relationship transitioned from synchronous increases to systematic decoupling, resulting in a high-greenness–low-resilience regime. The dominant drivers of vegetation resilience shifted from climate factors during 2000–2010 (contributing 51.57%) to grazing pressure during 2011–2020 (contributing 32.12%). Threshold analysis further revealed that vegetation resilience decline in the later decade resulted from the combined effects of intensified hydrological constraints and grazing pressure. Therefore, we recommend incorporating ecosystem resilience into restoration frameworks and coordinating the regulation of water availability and grazing pressure to promote the synergistic enhancement of greenness and resilience, thereby ensuring long-term ecosystem stability.