Degassing methane (CH4) through reservoir water compromises hydroelectricity's presumed low-carbon status, which has emerged as a critical hotspot for global carbon dynamics. However, a comprehensive understanding of the involved pathways remains elusive, hindering the accurate estimation of global reservoirs' carbon budget (emission-to-burial ratio). This study presents a holistic upscaling approach to assess methane degassing in global river reservoirs and its impacts on carbon budgets. Firstly, a machine learning model is utilized to characterize the contributions of climate and human factors to annual water residence time. Secondly, the stepwise multiple linear regression method is used to calculate CH4 degassing emissions for each reservoir. Finally, to systematically tackle all sources of uncertainty, separate uncertainty analyses are implemented for the estimates of degassing emissions, areal emissions, and organic carbon burial. Analyzing 30-year data from 6695 reservoirs worldwide, our assessment considers water residence time, temperature, catchment area, and reservoir size. Findings indicate that water releases contribute significantly to global CO2 emissions from reservoirs, elevating the carbon budget by 20 % from 2.02 to 2.18 TgC/year, underscoring the previously underestimated significance of CH4 degassing in shaping the carbon cycle impact of river reservoirs. We propose a redefined threshold for low carbon credits, suggesting that reservoirs with power densities exceeding 6.1 MW/km2, instead of the conventional 4 MW/km2, should qualify. This study underscores the need for sustainable water management and reshaping the carbon dynamics associated with hydroelectricity. Future research can advocate Artificial Intelligence (AI) techniques to enhance water management and mitigate carbon emissions by multi-objectively optimizing reservoir operations.
Keywords: Carbon emission; Hydroelectricity; Methane degassing; River reservoir.
Copyright © 2024. Published by Elsevier B.V.