Many methods have been applied to monitor fugitive methane gas from landfills. Recently, there have been suggestions to use a framework utilizing an unmanned aerial vehicle (UAV) for landfill gas monitoring, and several field campaigns have proved that a rotary UAV-based measurement has advantages of ease of control and high-resolution concentration mapping on the target planes. Research on the application of rotary UAVs for quantifying the whole-site methane emissions from waste landfills is limited so far. This study aimed to establish the methodology as an efficient and reliable method by developing field measurements and data processing procedure, evaluating errors by potential factors, and field applications.

A measurement system composed of a lightweight methane detector and a rotary UAV, and a procedure from field measurements to data processing were ii prepared. Potential error-inducing factors associated with the measurement system and the procedure were experimentally or theoretically assessed. In the detector reliability test, the methane detector had sufficient resolution for field application, and the critical UAV velocity required was obtained to ensure the credibility of the proposed measurement system. When spatial interpolators were applied to field data from the measurement system, the empirical Bayesian kriging demonstrated the best prediction of methane concentrations at unmeasured points. A wind estimation method using GPS/IMU data of UAV was also evaluated. Near-field experiments showed that the method produced wind vector estimates comparable to the wind parameters measured by a mechanical anemometer.

Field campaigns and following analysis demonstrated that it was able to successfully estimate wind vectors at multiple heights in contrast to a fixed anemometer. Estimated parameters in reasonable ranges, and explicable correlations between parameters also supported the validity of the wind estimator. Some of the evaluation results provided representative errors that were used in the comprehensive uncertainty analysis, as well as the validation of components. Multiple field campaigns were conducted at the Dangjin-si Resource Circulation Center, Dangjin, Korea. The estimated methane emission rates from seven campaigns ranged from 406.4 to 3,640 kg/ha/day, which was comparable to the emission rates modeled based on the IPCC guidelines. The total uncertainties combining effects of detection errors, interpolation errors, and wind variations were below 6 % for five cases, and below 23 % for three cases. Although the largest contributor turned out to be interpolation errors in most cases, it would be detection errors to lead to a significant reduction in the uncertainties of this methodology in the near future. There were failures in field iii campaigns due to misplacement of measurement planes or weak wind, which presented practical problems in the actual applications.

This work contains a complete description of the methodology, its evaluation, and its showcases both of success and failure. Complementation by near-field, denser measurements allowed to utilize the gas analyzer of compromised performance without a great aggravation of uncertainties. We sought the improvement to the emission estimation accuracy by introducing UAV-based wind estimation method. Messages for the practical applications of the methodology could be drawn from multiple field campaigns. This study could be also useful for other UAV-based studies in the near future, especially in the current situation where UAVs are widely employed for airborne measurement or remote sensing regardless of field.