Temperature Proxies as a Solution to Biased Sampling of Lake Methane Emissions. Issue 14 (20th July 2020)

• Record Type: Journal Article
• Authors: Jansen, Joachim; Thornton, Brett F.; Wik, Martin; MacIntyre, Sally; Crill, Patrick M.
• Is Part Of: Geophysical research letters. Volume 47:Issue 14(2020)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-07-20
• Subjects: methane -- northern lakes -- ebullition -- diffusion -- sampling bias -- temperature proxies; Geophysics -- Periodicals; Planets -- Periodicals; Lunar geology -- Periodicals; 550
• DOI: 10.1029/2020GL088647
• Languages: English
• ISSNs: 0094-8276

Abstract

Abstract: Lake emissions of the climate forcing trace gas methane (CH4 ) are spatiotemporally variable, but biases in flux measurements arising from undersampling are poorly quantified. We use a multiyear data set (2009–2017) of ice‐free CH4 emissions from three subarctic lakes obtained with bubble traps ( n = 14, 677), floating chambers ( n = 1, 306), and surface concentrations plus a gas transfer model ( n = 535) to quantify these biases and evaluate corrections. Sampling primarily in warmer summer months, as is common, overestimates the ice‐free season flux by a factor 1.4–1.8. Temperature proxies based on Arrhenius functions that closely fit measured fluxes ( R 2 ≥ 0.93) enable gap filling the colder months of the ice‐free season and reduce sampling bias. Ebullition (activation energy 1.36 eV) expressed greater temperature sensitivity than diffusion (1.00 eV). Resolving seasonal and interannual variability in fluxes with proxies requires ∼135 sampling days for ebullition, and 22 and 14 days for diffusion via models and chambers, respectively. Key Points: Multiyear lake methane flux monitoring reveals sampling only in summer (here: July, August) can overestimate ice‐free emissions by 43–76% Fluxes increase predictably with temperature and solar irradiance, enabling sampling bias correction via proxies based on energy input Temperature proxies can be more cost‐effective than measurements at estimating ice‐free fluxes that are not greatly affected by hysteresis