Effects of Boreal Lake Wetlands on Atmospheric 13CH3D and 12CH2D2

Recently, we developed a theoretical model to investigate the potential use of 13CH3D and 12CH2D2 as tools for tracking atmospheric methane budget. We used electronic structure methods to estimate kinetic isotope fractionations associated with the major sink reactions of CH4 in air (reactions with •OH and Cl•), and literature data with reconnaissance measurements of the relative abundances of 13CH3D and 12CH2D2 to estimate the compositions of the largest atmospheric sources. Here we present new methane rare isotopologue data from boreal wetlands, comprising one of the most important sources, in order to evaluate the robustness of the model. Boreal wetlands (>55° N) account for more than half of the wetland area in the Northern hemisphere. We analyzed methane samples from high latitude lakes representing different geographical regions, geological and ecological contexts, methane fluxes, and isotopic signatures. Using clumped isotopes of CH4 we are able to determine the likely production mechanism for natural CH4 samples. So far, all of our analyzed samples except one plot in the microbial pure-culture methanogenesis field (Young et al. 2017) with ranges of -0.2‰ to +1.2‰ for Δ13CH3D, and -29.6‰ to -18.2‰ for Δ12CH2D2. These compositions are far from equilibrium. The one exception, from Lake Doughnut, Alaska, exhibits Δ13CH3D and Δ12CH2D2 values of +5.2‰ and +18.7‰, respectively, which fall near ambient thermodynamic equilibrium values. This may be an effect of methanotrophy. Mean Δ13CH3D and Δ12CH2D2 for all lake samples are +1.7‰ and -15.4‰ respectively, compared to our original estimate of +6.1‰ and +21.2‰ for the wetland methane source based on an assumption of equilibrium. If we assume that these samples are representative of the overall wetland source, Δ13CH3D decreases by 0.8‰ and Δ12CH2D2 decreases by 0.6‰ in our model of bulk atmospheric methane. Δ13CH3D and Δ12CH2D2 values of air (including •OH and Cl• sink reactions) are estimated to be +3.6‰ and +112.9‰, respectively. Even if we exclude the Lake Doughnut sample, the total effect on Δ12CH2D2 in air is no more than 2‰. Our model predicts that sink reactions generate a distinctly elevated Δ12CH2D2 (by 104‰) relative to the source composition. In contrast, differences in source compositions have a comparatively small effect, on the order of a few per mil at most.