Spatiotemporal Variations and Emission Characteristics of CO ₂ Fluxes in Deep Large Reservoirs

Climate change driven by greenhouse gas emissions is a global priority. Deep large reservoirs represent significant potential emission sources, yet their greenhouse gas source-sink dynamics and driving mechanisms remain inadequately understood. This study targeted three representative deep reservoirs of the upper Yellow River (Longyangxia (LYX), Lijiaxia (LJX), and Liujiaxia (LJXX)). Comparative analysis of water column conditions during the stratification period (summer) and mixing period (winter) was conducted. We quantified the spatial distribution of dissolved partial pressure of CO 2 ( p CO 2 ) and the water-air interface exchange flux ( F CO₂ ), elucidating their spatiotemporal patterns and carbon source-sink attributes. Key findings include: (1) Concentration and Flux Characteristics: Mean p CO 2 was 765.00 μatm in summer and 739.00 μatm in winter. Corresponding mean F CO₂ was 12.273 mmol/(m2·d) and 9.111 mmol/(m 2 ·d). (2) Regulatory Mechanisms: Thermal stratification, climatic fluctuations, and reservoir-specific properties influenced p CO 2 by modulating carbon cycling pathways and carbonate equilibria. (3) Source-Sink Nature: All three reservoirs exhibited significant spatial gradients and seasonal variations in p CO 2 and F CO₂ , collectively functioning as sources of the CO 2 . This study provides critical theoretical support for quantifying the carbon footprint of hydropower projects.