Spatial quantification of blue carbon ecosystem stocks is crucial for developing policies to mitigate climate change, especially in regions experiencing ongoing wetland disturbance from biological invasions. We integrated multiple machine learning models with the space-for-time substitution method to quantify the spatiotemporal impact of Spartina alterniflora invasion on tidal marsh sediment blue carbon (soil organic carbon – 'SOC') stocks at 100 cm depth in the Yangtze Estuary. Our results show that the invasive S. alterniflora contributed more than half of the total SOC stocks (2,056 ± 379 Gg C, 1 Gg = 106 kg) in the 27,600 ha tidal marshes of the Yangtze Estuary, which were estimated to be 1,107 ± 176 Gg C. S. alterniflora increased the SOC stocks in the Yangtze Estuary within the first 15 years, but this gain was not sustained in the long term, with a gradual decline (by 13.14 Mg C/ha) observed after 15 years of S. alterniflora growth. We found that sediment salinity, tidal range, and human accessibility were strong indicators for modeling and predicting SOC stocks, with Random Forest providing the best simulation of tidal marsh SOC stocks (R2 = 0.894, RMSE=7.646 Mg C/ha, and MAPE=9.469 %). Our study provides much needed information on blue carbon stocks in the Yangtze Estuary under biological invasion stress, and offers guidance for targeted S. alterniflora management actions in the future.