Testing the near-field Gaussian plume inversion flux quantification technique using unmanned aerial vehicle sampling

Abstract

Methane emission fluxes from many facility-scale sources may be poorly quantified, potentially leading to uncertainties in the global methane budget. Accurate atmospheric measurement-based flux quantification is urgently required to address this. This paper describes the first test (using unbiased sampling) of a near-field Gaussian plume inversion (NGI) technique, suitable for facility-scale flux quantification, using a controlled release of methane gas. Two unmanned-aerial-vehicle (UAV) platforms were used to perform 22 flight surveys downwind of a point-source methane gas release from a regulated cylinder with a flowmeter. One UAV was tethered to an instrument on the ground, while the other UAV carried an on-board prototype instrument (both of which used the same near-infrared laser technology). Both instruments were calibrated using certified standards to account for variability in the instrumental gain factor, assuming fixed temperature and pressure. Furthermore, a water vapour correction factor, specifically calculated for the instrument, was applied and is described here in detail. We also provide guidance on potential systematic uncertainties associated with temperature and pressure, which may require further characterisation for improved measurement accuracy. The NGI technique was then used to derive emission fluxes for each UAV flight survey. We found good agreement of most NGI fluxes with the known controlled emission flux, within uncertainty, verifying the flux quantification methodology. The lower and upper NGI flux uncertainty bounds were, on average, 17 %±10(1σ) % and 227 %±98(1σ) % of the controlled emission flux, respectively. This range of conservative uncertainty bounds incorporate factors including the variability in the position of the time-invariant plume and potential for under-sampling. While these average uncertainties are large compared to methods such as tracer dispersion, we suggest that UAV sampling can be highly complementary to a toolkit of flux quantification approaches and may be a valuable alternative in situations where site access for tracer release is problematic. We see tracer release combined with UAV sampling as an effective approach in future flux quantification studies. Successful flux quantification using the UAV sampling methodology described here demonstrates its future utility in identifying and quantifying emissions from methane sources such as oil and gas extraction infrastructure facilities, livestock agriculture, and landfill sites, where site access may be difficult.