USGS Science⌕ Search
Home/Archive/USGS

USGS · 70278993

Quantifying methane emissions from a rich fen with uncrewed aircraft systems in boreal Alaska

J.M. Tomlin·C.W. Edgar·E. Seroki·E.S. Euskirchen·E.S. Kane·M.Y. McPartland·L.T. Iraci·Kristen L. Manies
View original source ↗Export metadata ↓

Abstract

Thawing of permafrost in northern latitudes is accelerating, potentially releasing substantial amounts of methane (CH 4 ) as forested permafrost plateaus transition into wetlands. This ecosystem shift alters the carbon exchange between the soil and atmosphere, influencing the permafrost-carbon feedback. Monitoring these changes may require measurement platforms operating across varied spatial and temporal scales. Recent advancements in small uncrewed aircraft systems (sUAS) enable high resolution CH 4 flux quantification in remote, complex terrains; however, comparisons with established methods such as eddy covariance flux towers remain limited. We used a hexacopter sUAS to quantify CH 4 emissions from the Alaska Peatland Experiment, a wetland within the Bonanza Creek Experimental Forest. Using an ensemble of methods to define the background CH 4 concentration, along with near surface emissions from soil chambers, helped constrain our flux estimates. The sUAS method yielded an average flux of 0.0077 ± 0.0019 mol s −1 CH 4 , within a factor of two concurrent tower-derived total source flux estimates (0.0036 ± 0.00042 mol s −1 CH 4 ). To assess spatial drivers of observed fluxes, we conducted a 2D footprint analysis and overlaid the results with high-resolution hyperspectral land cover classification, quantifying vegetative contributions within each footprint. This revealed higher fen representation in sUAS measurements (73.8%) than in tower footprints (58.8%), and lower tussock meadow representation (15.6% and 30.3%, respectively). These differences were consistent with known variation in vegetation-specific CH 4 emissions. Our results highlight that combining footprint modeling with land cover characterization can enhance interpretations of CH 4 fluxes and guide cross-platform comparisons.

Explore related subjects

JGR AtmospheresArticle

Keep this discovery

BibTeXRIS

J.M. Tomlin, C.W. Edgar, E. Seroki, E.S. Euskirchen, E.S. Kane, M.Y. McPartland, L.T. Iraci, Kristen L. Manies. 2026-07-30. Quantifying methane emissions from a rich fen with uncrewed aircraft systems in boreal Alaska. https://doi.org/10.1029/2026jd046424

Cite the original work for its findings. Save a collection to share your selection of sources.

Discover connections

Alaska placeArticle subjectJGR Atmospheres subject

Connections use source metadata and explicit phrase matches, not verified experimental comparisons.

KEEP EXPLORING

Related discoveries

USGS2026

Assessment of extreme subsurface hydrologic conditions captured during atmospheric river storms in the San Francisco Bay area (California, USA) with applications to shallow landslide early warning

An increase in soil pore water pressure is the typical trigger for the majority of landslides caused by rainfall. Atmospheric river storms, common to the west coast of North America during the winter season, can deliver landslide triggering rainfall resulting in severe impacts to coastal communities. Using a network of hydrologic monitoring stations situated within landslide-prone terrain in the San Francisco Bay area of California (USA), we assess the meteorologic conditions and resulting hydrologic and landslide response resulting from eight consecutive storm events that caused thousands of shallow landslides during the winter of 2022–2023. We find disparate hydrological responses and resultant degrees of observed landsliding ranging from < 1 landslide/km2 to 18 landslides/km2 at the monitoring sites that reflect the interplay and differences between rainfall delivery, subsurface hydrological characteristics, and geotechnical properties at each site. Antecedent soil moisture from both early season rainfall and the first storm in the sequence played a critical role in setting up some hillslopes for failure. Subsequent storms then generated elevated pore water pressures for several hours with associated landsliding. However, we find that the occurrence of widespread landsliding required not only sufficient pore water pressure magnitude in susceptible hillslopes, but also full and prolonged effective soil saturation throughout hillslope profiles. Landslides may still occur at lower values and durations of effective saturation but are likely to be less extensive regionally. We present these findings within the context of research directions and improvements to landslide early warning systems first suggested by researchers 40 years ago.

California↗
USGS2025

A glimpse into the future of tectonic tremor monitoring

Tectonic tremor is a weak, long-duration seismic signal often observed in subduction zones and on some other plate-bounding faults. Because of tremor's characteristically low amplitude (and low signal-to-noise) and lack of clear phase arrivals, detecting and locating tremor usually requires techniques distinct from those applied to typical earthquakes. Major advances in detection and understanding of tremor have derived in the past from a powerful combination of new data and new analysis techniques. In a recent study, Sagae et al. (2025, https://doi.org/10.1029/2025jb031348 ) exploit that combination again, developing a new machine-learning based workflow and applying it to the S-net cabled seismic network in the Japan trench offshore northern Honshu. Their approach, although complex, succeeds in detecting several times more tremor activity than earlier studies, resulting in new insights and providing a blueprint for similar approaches that could be applied elsewhere. As real-time earthquake monitoring adopts similar tools, it may present an opportunity to bring tremor monitoring into operational workflows. In turn, this could solidify tremor monitoring as a component of future operational earthquake forecasting.

JGR Solid Earth↗
USGS2026

SlideDetect: Spatio-temporal landslide detection using a three-dimensional convolutional neural network

Landslides pose a serious and ongoing threat to both human lives and infrastructure worldwide; therefore, it is of interest to predict where and when landslides are likely to occur. Advances in machine learning techniques have spurred numerous studies aimed at estimating relative landslide propensity, but are limited to spatial (as opposed to temporal) prediction due to the sparsity of landslide timing data. We address this data gap by training SlideDetect, a 3-dimensional convolutional neural network (3D CNN), to identify landslides based on their spatial and temporal occurrence within multitemporal image stacks. We use an inventory of landsides triggered by the 2018 Hokkaido earthquake and two years of monthly composite optical imagery spanning this event. The model can identify not only landslide location but also landslide date with an area under the precision-recall curve (PR-AUC) of 0.84. We further present a new standard for presenting PR curve results that explicitly compares model performance at different confidence thresholds, allowing for clearer model evaluation and comparison. Our new approach to constraining landslide timing paired with this more consistent and objective method for evaluating model performance shows considerable promise, and with further application and testing, SlideDetect could enhance the data availability and tools needed to advance landslide hazard and risk assessments.

JGR Machine Learning and Computation↗