Abstract
Compared with traditional fossil fuels, wind power, as a typical clean energy source, is capable of effectively reducing carbon emissions. However, recent studies have shown that wind farms (WFs) in some regions may suppress vegetation carbon sequestration (VCS), thereby reducing the regional carbon sink capacity and diminishing the associated life-cycle emission reduction benefits. Therefore, WF siting requires a trade-off between development potential and ecological impact to avoid negative consequences while promoting synergistic gains. This study presents a new WF siting framework that uses a random forest model to predict the impact of WF deployment on VCS, and integrates this analysis with a development potential map obtained from multi-criteria evaluation to generate a comprehensive siting map. Using the Inner Mongolia grassland region as a case study, we demonstrate the implementation of this framework. The results reveal that 9169.09 km2 of grassland has been identified as priority development areas. These areas, which are located mainly in northeastern Xilingol and western Chifeng, not only possess high development potential but also significantly enhance VCS. The siting framework developed in this study achieves synergistic optimization of wind power development and VCS, which addresses the lack of ecological considerations in previous studies and provides new insights for further emission reductions through renewable energy, demonstrating broad application potential.