Characterizing methane emissions from high-organic-content landfills via integrated UAV and ground-based monitoring

Landfill methane emissions are spatially heterogeneous and difficult to link to individual surface features. We combined unmanned aerial vehicle (UAV) remote sensing and ground measurements to characterize methane patterns at a large high-organic-content landfill in Northwest China. UAV hotspots were compared with ground hotspot maps at 281 locations, using ground measurements at 3-4 cm and 1.5 m above the cover surface as separate references. At a 30 m matching distance, recall was 51.5-51.9 % and precision was 70.6 %. Areas with overlapping ground and UAV hotspots were selected for closer examination using orthophotos and field observations. Nearby features included damaged high-density polyethylene (HDPE) geomembranes, cracks in loess cover, cover-material interfaces, and gas collection infrastructure. Universal kriging was used to interpolate discrete UAV path-integrated methane observations onto a continuous grid and support local mass-balance screening at selected locations. Screening estimates were higher and more variable at selected hotspots in the 3-year loess-covered zone than at those in the 5-8-year zones. The ten selected hotspots in the 3-year zone had a mean kriging-assisted mass-flow estimate of 28.73 g s⁻¹. These comparisons describe selected hotspots and do not isolate waste-age or cover-material effects. A UAV methane column enhancement near the landfill boundary had no corresponding near-surface hotspot, leaving its source unresolved. The field comparison supports using UAV hotspots to prioritize ground inspection, while the unmatched observations show the need to examine locations beyond UAV hotspots and investigate nearby features before attributing methane enhancements to specific sources.