Abstract
Dams trap organic carbon (OC) and create carbon dioxide (CO 2 ) and methane (CH 4 ) emission hotspots within reservoirs, underlying the ongoing debate over their role in carbon emissions. However, even without dams, this labile OC would still decompose and emit CO 2 or CH 4 during downstream transport and after deposition in estuaries. Thus, relying solely on dammed reservoir observations to assess these emissions may introduce considerable bias. Here, we simulated both dammed and non-dammed scenarios for 5346 global reservoir dams to assess their impact on carbon emissions from retained OC. The results show that reservoir-only approaches overestimate dam-induced carbon emissions by neglecting baseline emissions under no-dam conditions. After baseline correction, dams increase carbon emissions by an average of 21.5% in CO 2 -equivalent (CO 2 -eq), which represents only 18.6% of the emissions estimated using the conventional reservoir-based accounting approach. XGBoost identified reservoir depth and hydraulic residence time as the primary controls on dam-induced carbon emissions.
Funding Information
  • National Natural Science Foundation of China (42422605)
  • National Natural Science Foundation of China (42277060)
  • National Natural Science Foundation of China (42477073)
  • National Key Research and Development Program of China (2024YFC3214500)
  • Natural Science Foundation of Jiangsu Province (BK20240142)
  • US National Science Foundation (2108917)
  • US National Institutes of Health (2P01ES028939-06)
  • Yunnan Provincial Council of Academicians and Experts Workstations (202405AF140006)
  • US National Science Foundation (2418066)

This publication has 43 references indexed in Scilit: