Landfill methane emissions were measured continuously using the eddy covariance method from June to December 2010. The study site was located at the Bluff Road Landfill in Lincoln, Nebraska, USA. Our results show that landfill methane emissions strongly depended on changes in barometric pressure; rising barometric pressure suppressed the emission, while falling barometric pressure enhanced the emission, a phenomenon called barometric pumping. There was up to a 35-fold variation in day-to-day methane emissions due to changes in barometric pressure. Wavelet coherence analysis revealed a strong spectral coherency between variations of barometric pressure and methane emission at periodicities ranging from 1 day to 8 days. Power spectrum and ogive analysis showed that at least 10 days of continuous measurements was needed in order to capture 90% of the total variance in the methane emission time series at our landfill site. From our results, it is clear that point-in-time measurements taken at monthly or longer time intervals using techniques such as the trace plume method, the mass balance method, or the closed-chamber method will be subject to large variations in measured emission rates because of the barometric pumping phenomenon. Estimates of long-term integrated methane emissions from landfills based on such measurements could yield uncertainties, ranging from 28.8% underestimation to 32.3% overestimation. Our results demonstrate a need for continuous measurements to quantify annual total landfill emissions. This conclusion may apply to the study of methane emissions from wetlands, peatlands, lakes, and other environmental contexts where emissions are from porous media or ebullition. Other implications from the present study for hazard gas monitoring programs are also discussed.

No abstract

No abstract