Lakes are hotspots for CH4 and CO2 effluxes, but their magnitude and underlying drivers are still uncertain due to high spatiotemporal variation within and between lakes. We measured CH4 and CO2 fluxes at high temporal (hourly) and spatial resolution (approx. 13 m) using 24 automatic floating chambers equipped with continuously recording sensors that enabled the determination of diffusive and ebullitive gas fluxes. Additionally, we measured potential drivers such as weather patterns, water temperature, and O2 above the sediment. During five days in autumn 2021, we conducted measurements at 88 sites in a small, shallow eutrophic Danish Lake. CH4 ebullition was intense (mean 54.8 μmol m−2 h−1) and showed pronounced spatiotemporal variation. Ebullition rates were highest in deeper, hypoxic water (5–7 m). Diffusive CH4 fluxes were 4-fold lower (mean 15.0 μmol m−2 h−1) and spatially less variable than ebullitive fluxes, and significantly lower above hard sediments and submerged macrophyte stands. CO2 concentration in surface waters was permanently supersaturated at the mid-lake station, and diffusive fluxes (mean 919 μmol m−2 h−1) tended to be higher from deeper waters and increased with wind speed. To obtain mean whole-lake fluxes within an uncertainty of 20 %, we estimated that 72 sites for CH4 ebullition, 39 sites for diffusive CH4 fluxes and 27 sites for diffusive CO2 fluxes would be required. Thus, accurate whole-lake quantification of the dominant ebullitive CH4 flux requires simultaneous operation of many automated floating chambers. High spatiotemporal variability challenges the identification of essential drivers and current methods for upscaling lake CH4 and CO2 fluxes. We successfully overcame this challenge by using automatic floating chambers, which offer continuous CH4 and CO2 flux measurements at high temporal resolution and, thus, are an improvement over existing approaches.
Bibliografisk noteFunding Information:
The work is part of the project “Supporting climate and biodiversity by rewetting low-lying areas” from the Independent Research Fund Denmark ( 0217-00112B ) to KSJ. We thank Aage V. Jensen's foundation for grants supporting the PhD research of JSS. We are thankful for the CH 4 sensors funded by the COWI foundation ( A-155.03 ) and the Carlsberg grant for funding the Ultraportable Greenhouse Gas Analyzer ( CF21-0166 ). We thank David Bastviken for discussions of our approach and results and David Stuligross for proof-reading the manuscript.
© 2023 The Authors