Assessment of the potential for carbon sink enhancement in the overlapping ecological project areas of China.

Ecological engineering can significantly improve ecosystem carbon sequestration. However, few studies have projected the carbon sink trends in regions where ecological engineering projects overlap and have not considered the different climate change conditions and land use scenarios. Using the ensemble empirical mode decomposition method and machine learning algorithms (enhanced boosted regression trees), the aims of this study to elucidate the stability of carbon sinks and their driving mechanisms in areas where ecological projects overlap and to predict the potential enhancement in carbon sinks under varying climate and human activity scenarios. The findings revealed that: (1) The carbon sinks clearly and steadily increased in regions where five ecological projects were implemented from 1982 to 2019. In contrast, the carbon sinks did not significantly increase in regions with two or three ecological projects. (2) As the number of ecological projects increased, the impact of human activities on the carbon sinks gradually decreased. In eastern China, rapid economic development and significant interference from human activities hindered the growth of carbon sinks. In contrast, in western China, the warming and humidification trend of the climate, large-scale afforestation, and other ecological projects have significantly improved carbon sinks. (3) The regions with five overlapping ecological projects exhibited the greatest enhancement and stability of carbon sinks under different scenarios. Compared with the SSP585 scenario, under the SSP126 scenario, the carbon sinks increased, and their stability was greater. Achieving carbon neutrality requires major ecological projects to account for the limitations imposed by climatic conditions. Instead of isolated projects or the implementation of single restoration measures, a comprehensive approach that uses the synergistic effects of combined ecological strategies is recommended.