Methane emissions from a small wind shielded lake determined by eddy covariance, flux chambers, anchored funnels, and boundary model calculations: a comparison.

Lakes are large sources of methane, held to be responsible for 18% of the radiative forcing, to the atmosphere. Periods of lake overturn (during fall/winter) are short and therefore difficult to capture with field campaigns but potentially one of the most important periods for methane emissions. We studied methane emissions using four different methods, including eddy covariance measurements, floating chambers, anchored funnels, and boundary model calculations. Whereas the first three methods agreed rather well, boundary model estimates were 5-30 times lower leading to a strong underestimation of methane fluxes from aquatic systems. These results show the importance of ebullition as the most important flux pathway and the need for continuous measurements with a large footprint covering also shallow parts of lakes. Although fluxes were high, on average 4 mmol m(-2) d(-1) during the overturn period, water column microbial methane oxidation removed 75% of the methane and only 25% of potential emissions were released to the atmosphere. Hence, this study illustrates second the importance of considering methane oxidation when estimating the flux of methane from lakes during overturn periods.