Spatiotemporal Persistence and Directed Influence Pathways of Regional CO2 Concentrations in China

Although atmospheric CO2 concentration plays a crucial role in the climate system, the persistence of structural linkages among regions and the directional pathways through which they influence one another have not been systematically quantified. In this study, an interregional CO2 concentration network across China is constructed using daily observations within a climate network framework. The results show that the network exhibits pronounced structural stability on multi-year timescales, with consistently high Jaccard similarity values between networks from different years. Network connectivity is dominated by zero and short time delays (within one day), indicating that regional CO2 variations mainly exhibit coherent fluctuations on short temporal scales. Key source nodes and sink nodes are identified within the network. Strong source nodes are primarily located in Northeast China and parts of inland Northwest China, while strong sink nodes are mainly concentrated in the Xinjiang region, particularly around the Junggar Basin and the Tarim Basin. Pathway analysis using the Dijkstra shortest-path algorithm further reveals a dominant influence pathway extending from Northeast China toward Northwest China, with numerous pathways converging in the Xinjiang region. These findings offer a network-based perspective for examining the spatial influence structure of interregional CO2 concentrations in China and emphasize the importance of cross-regional interactions in shaping regional atmospheric CO2 variability.