Biased sampling of methane release from lakes and ponds: A problem for extrapolation

Monday, 15 December 2014
Martin Wik1, Brett F Thornton1, Patrick M Crill2 and David Bastviken3, (1)Stockholm University, Dept. of Geological Sciences, Stockholm, Sweden, (2)Stockholm University, Stockholm, Sweden, (3)Linköping University, Linköping, Sweden
Lakes and ponds at high latitudes are important sources of atmospheric methane (CH4). However, their overall contribution is largely unconstrained, partly due to sampling campaigns with short measurement periods and few sampling locations against a backdrop of high spatiotemporal variations in the emissions. Using CH4 emission data from a total of 49 studies, we point out that high-end fluxes that average above 200 mg CHm-2 d-1 were measured during sampling intervals of 1 week or less. The range of these short-term fluxes is also twice that of those measured during sampling intervals of 3 months or longer. Further, measurements of small shallow thaw ponds are overrepresented in the literature, which adds error and complexity when scaling CH4 emissions regionally and globally. We use these and other examples to suggest improved sampling strategies that will reduce biases of CH4 emission estimates from high latitude inland waters.