Centennial spatiotemporal evolution of AGC in subtropical forests and key driving factors − based on process modeling and geography detector

• Centennial AGC changes of subtropical forest were predicted using BIOME-BGC model. • The total forest AGC in Zhejiang Province will range from 269.55 Tg C to 301.67 Tg C in 2100. • We used the OPGD model to analyze the drivers of future forest AGC changes. • The forest AGC changes were mainly affected by the N content and CO 2 concentration. Forests are the primary carbon reservoirs in terrestrial ecosystems and play a key role in the global carbon cycle. However, the future of the forest carbon cycle and its storage is uncertain due to climate change and socio-economic development. Above-ground carbon (AGC) storage is a crucial indicator for assessing carbon stocks in forests, making it essential to understand AGC trends and influencing factors for effective management. Therefore, this study utilized the BIOME-BGC model to simulate the spatial and temporal distribution of AGC in subtropical forests over a 100-year period (2000–2100) under four SSP-RCP scenarios, focusing on Zhejiang Province, China. The trend of AGC spatiotemporal evolution was analyzed. The optimal parameters-based geographical detector (OPGD) model was also used to explore the key drivers of AGC changes. The results of the study showed that: (1) the correlation coefficients (R) between the simulated and observed AGC values of the BIOME-BGC model are all above 0.7, indicating that the model can effectively simulate the forest AGC dynamic; (2) In the SSP1-2.6 scenario, AGC increased first and then tended to decrease after reaching the peak in 2080; in the SSP2-4.5 scenario, AGC increased year by year; and in the SSP3-7.0 and SSP5-8.5 scenarios, AGC decreased first and then increased year by year; (3) based on the analysis using coefficient of variation (CV), the Theil-Sen Median (TSM) and the Mann-Kendall (MK) tests, the spatial heterogeneity of forest AGC in Zhejiang Province under the four different scenarios was significantly enhanced, and the density of forest AGC was significantly increased; (4) The OPGD factor detection analysis indicated that the primary independent factors influencing the spatiotemporal dynamics of AGC were nitrogen content under the SSP1-2.6 scenario, and atmospheric CO 2 concentration under the SSP2-4.5, SSP3-7.0, and SSP5-8.5 scenarios. The study results provided a scientific foundation for managing carbon sinks in subtropical forests amid global climate change and effective climate adaptations.