Impact of hydropower construction on vegetation NPP in the Yarlung Zangbo River, Qinghai-Tibet Plateau

Vegetation net primary production (NPP) is integral to the global carbon cycle, especially in the alpine ecosystems of the Qinghai-Tibet Plateau. However, hydropower construction can profoundly disrupt ecosystem health and carbon sequestration capacity, particularly along the Yarlung Zangbo River. Therefore, there is an urgent need to understand the impacts of hydropower construction on riparian vegetation NPP through precise, engineering-scale assessments of long-term changes. Changes in NPP may alter soil carbon inputs and turnover, with potential feedbacks on ecosystem carbon balance. Such assessments are essential for comprehensive ecological evaluations of hydropower projects. In this study, we downscaled MODIS NPP data to 30 m resolution using a convolutional neural network-based model to examine the spatiotemporal dynamics of NPP changes at the engineering scale across disturbance zones (DZs) and influence zones (IZs). We further applied a structural equation model (SEM) and a generalized linear model (GLM) to enable process-based attribution of dam- and climate-driven changes in NPP from the pre-construction, during-construction, to post-construction stages. The results showed that hydropower construction caused persistent NPP reductions in DZs due to engineered surface expansion and vegetation loss, whereas IZs experienced a sharp NPP decline during construction (88.6%), followed by widespread recovery in the post-construction stage (93.9%). The SEM–GLM framework identified dam construction as the predominant driver of NPP variations (standardized path coefficient: −0.35 in DZs; mean relative importance: 38.6% in IZs), exceeding the combined effects of precipitation and temperature. These findings provide critical insights for ecological risk assessment, sustainable ecosystem management, and the coordinated development of engineering infrastructure and the natural environment in future hydropower projects in environmentally sensitive alpine regions.