Unveiling the temporal variability of gas transfer coefficients of streams based on high-frequency dissolved oxygen measurements

Greenhouse gas (GHG) emissions from streams and rivers are important sources of global GHG emissions. As a crucial parameter for estimating GHG emissions, the gas transfer coefficient (expressed as K₆₀₀ at water temperature of 20 °C) has uncertainties. This study proposed a new approach for estimating K₆₀₀ based on high-frequency dissolved oxygen (DO) data and an ecosystem metabolism model. This approach combines the numerical solution method with the Markov Chain Monte Carlo analysis. This study was conducted in the Chaohu Lake watershed in Southeastern China, using high-frequency data collected from six streams from 2021 to 2023. This study found: (1) The numerical solution of K₆₀₀ demonstrated distinct dynamic variability for all streams, ranging from 0 to 111.39 cm h⁻¹ (2) Streams with higher discharge (>10 m³ s⁻¹) exhibited significant seasonal differences in K₆₀₀ values. The monthly average discharge and water temperature were the two factors that determined the variation in K₆₀₀ values. (3) K₆₀₀ was a major source of uncertainty in CO₂ emission fluxes, with a relative contribution of 53.72%. An integrated K₆₀₀ model of riverine gas exchange was developed at the watershed scale and validated using the observed DO change. Our study stressed that K₆₀₀ dynamics can better represent areal change to reduce uncertainty in estimating GHG emissions.