Spatiotemporal patterns of abrupt vegetation changes and reversal trends in the Yellow River Basin: Climate and human drivers

Abstract

Vegetation changes involve both gradual trends and abrupt changes, with the latter serving as critical indicators for detecting vegetation state transitions and identifying potential environmental stresses. In recent decades, the structure and function of vegetation ecosystems in the Yellow River Basin have undergone significant alterations. However, there remains limited understanding of whether abrupt changes occurred during these alterations and their driving mechanisms. So here, this study utilized the Normalized Difference Vegetation Index (NDVI) as a proxy for vegetation changes, characterized the spatiotemporal distribution of abrupt vegetation changes characteristics based on the Breaks for Additive Seasonal and Trend (BFAST) algorithm, and employed a Random Forest model to elucidate the driving factors behind these changes. The results showed that 73.1% of the vegetation in the basin experienced at least one abrupt change, with 2011 marking the peak year (9.8% affected area). Notably, 21.35% of pixels exhibited trend reversals, a phenomenon masked in previous linear trend studies. Regarding driving factors, positive abrupt changes were primarily driven by moderate warming and extreme precipitation anomalies, while the concentrated positive abrupt changes in 2011 were significantly influenced by the transition between dry and wet years and ecological restoration projects. In contrast, negative abrupt changes predominantly resulted from extreme temperatures and intensified human activities. These findings reveal the causes and response mechanisms of abrupt vegetation changes in the Yellow River Basin, providing new insights and perspectives for exploring the patterns of vegetation changes in this region.