Spatiotemporal variability and controls of CH4 and CO2 emissions in an alpine reservoir: insights from the Liujiaxia system

• Methane emissions primarily from ebullition in alpine reservoirs. • Seasonal and spatial variations significantly affect emission rates. • Hydrogenotrophic methanogenesis in central zones, acetoclastic in nearshore and upstream areas. • Hydrological management amplifies upstream GHG emissions in alpine reservoirs. Alpine reservoirs represent critical but poorly quantified sources of greenhouse gas (GHG) emissions from inland waters. This study examines spatiotemporal variations and drivers of methane (CH 4 ) and carbon dioxide (CO 2 ) fluxes in the Liujiaxia (LJX) Reservoir (1735 m a.s.l.), a high-altitude hydropower system in the upper Yellow River, through multi-season field campaigns (2021–2023). High-frequency flux measurements, stable isotope analyses (δ 13 C-CH 4 and δ 13 C-CO 2 ), and statistical modeling identified contrasting emission patterns between central and nearshore zones. Ebullition dominated CH 4 emissions (54–86 % of total fluxes), with summer total fluxes surpassing spring and winter levels by factors of 2.8–5.7. Isotopic evidence revealed hydrogenotrophic methanogenesis as the principal pathway in the central zone (α c = 1.053 ± 0.01; δ 13 C-CH 4 = −50.19 ± 4.09 ‰), contrasting with acetoclastic dominance in nearshore regions (α c = 1.028 ± 0.01; δ 13 C-CH 4 = −65.04 ± 6.38 ‰), implicating sediment-driven methane production. CO 2 fluxes exhibited pronounced seasonality, linked to organic carbon mineralization in nearshore sediments. Hydrological regulation amplified upstream GHG emissions, where CH 4 and CO 2 fluxes exceeded downstream values by 3.7- and 1.8- fold, respectively. A random forest model, informed by high-resolution spatiotemporal flux data, estimated annual emissions of 0.23 Gg CH 4 and 20.18 Gg CO 2 . These findings underscore the significant influence of hydrologic management and biogeochemical processes on GHG budgets in alpine reservoirs, providing essential data to improve global models of inland water emissions.