摘要: The Yellow River (YR), China's second-longest river, remains understudied regarding its greenhouse gases (GHGs) emissions, particularly the impacts of urban drainage ditches and wastewater treatment facilities on regional GHGs dynamics. This study investigated methane (CH₄) and carbon dioxide (CO₂) concentrations, fluxes and stable carbon isotopes (δ¹³C-CH₄ and δ¹³C-CO₂) across six main stream, three ditches, and one wastewater treatment site along the upper Lanzhou section of the YR, spanning from the urban entrance (36.176°N, 103.449°E) to the exit of Lanzhou city (36.056°N, 104.020°E). Measured CH₄ diffusion fluxes in mainstem sites ranged from 0.01 to 2.58 mmol·m^-2·d^-1 (mean:0.36 mmol·m^-2·d^-1), while ebullitive fluxes (gas bubbles) ranged from 0.01 to 18.89 mmol·m^-2·d^-1 (mean:0.90 mmol·m^-2·d^-1). CO₂ diffusion fluxes varied between 9.16-92.80 mmol·m^-2·d^-1 (averaged:39.11 mmol·m^-2·d^-1) at these locations. Ebullition (bubble) fluxes accounted for 53.1% ±22.4% (range:9.0% to 98.4%) to total CH₄emissions (diffusion plus ebullition), with peak fluxes occurring during summer, indicating its significance as a CH₄ transport mechanism. Notably, both diffusion CH₄ and CO₂ fluxes and ebullitive CH₄ rates at ditch sites substantially exceeded those in mainstream reaches. The lowest CH₄ and highest CO₂ concentrations were observed at a wastewater treatment site, likely resulting from the removal of high organic loads. Acetoclastic methanogenesis-the process converting acetate-derived methyl groups to CH₄-was identified as the dominant production pathway in both mainstream and ditch environments. CH₄ and CO₂ flux magnitudes in the upper YR (Lanzhou section) were comparable to those observed in subtropical Yangtze River tributaries. These results demonstrate that anthropogenic influences significantly enhance CO₂/CH₄ emissions, and the lateral exports of dissolved carbon (DIC and DOC) in the main stream site was quantified., which cannot be overlooked. The findings emphasize the critical need to account for pronounced spatiotemporal variations in arid-region GHG fluxes to improve basin-scale estimates for the YR.