Evaluating uncertainty in Sweden’s national methane inventory for landfills: A UAV-based case study at Filborna

Abstract

Methane (CH₄) emissions from landfills are a key uncertainty in Sweden’s national greenhouse gas inventory. This thesis evaluates whether UAV-based CH₄ measurements can complement inventory-based estimates, using Filborna landfill in southern Sweden as a case study. UAV-derived emissions from 2022–2024 were compared with an IPCC Tier 2 First Order Decay model, reproduced from Sweden’s national inventory methodology and applied to the Filborna Waste Facility. For the UAV-derived emissions, uncertainty was assessed for wind speed, wind direction, kriging interpolation, sensor precision, and background estimation. Wind direction was the dominant source of uncertainty in the UAV-based flux estimates. The reproduced First Order Decay model showed good agreement with Sweden’s national inventory estimates at the national scale, while UAV measurements revealed short-term, site-specific variability that the model cannot capture directly. UAV methods cannot replace inventory modelling, but can provide independent observations to evaluate assumptions, identify variability, and improve confidence in landfill CH₄ emission estimates.

Popular Abstract

Methane from Landfills: What Can Drones Tell Us?
Landfills can release methane long after waste has been buried. Methane is an important greenhouse gas, and knowing how much is released from landfills matters for climate reporting and emission reduction. However, landfill methane is difficult to estimate because emissions can change over time, vary across the landfill surface, and depend on weather conditions. This project investigated methane emissions from Filborna landfill in Helsingborg, Sweden, and compared modelled emissions with direct measurements from drones and satellites.
In national greenhouse gas inventories, landfill methane emissions are usually estimated using a model based on how much waste was landfilled in the past and how quickly the organic material breaks down. This project first recreated the Swedish national landfill methane model and then applied a similar model to Filborna using site-specific waste data. The model showed that methane emissions from Filborna likely peaked in the early 1990s and have decreased strongly since then. For 2024, the model estimated an average emission rate of about 86 kg methane per hour.
The modelled emissions were compared with direct measurements from the landfill. These included older site-level measurements and drone-based methane measurements from 2022, 2023, and 2024. The drone measurements showed that methane was not released evenly across the landfill. Instead, emissions appeared in localised plume structures, meaning that some areas released more methane than others during the flights. This shows why direct measurements are useful: they can reveal spatial patterns that a long-term model cannot show.
Overall, the model generally estimated higher methane emissions than the measurements for most of the period from 2001 to 2024. In more recent years, the modelled and measured values became more similar. This suggests that the model captures the broad long-term decrease in methane emissions, but does not fully describe short-term variation or site-specific conditions. Drone measurements are also not perfect annual estimates, because each flight only represents a short moment in time under specific wind and weather conditions.
Satellite observations were also examined, but no quantifiable methane emissions were detected over Filborna. This does not mean that the landfill emitted no methane. Instead, the emissions were likely too small, too diffuse, or too affected by weather and surface conditions to be detected reliably from space. Satellites are useful for finding large methane plumes, but many landfill emissions are spread out over a larger area and may fall below current satellite detection limits.
Drones are better suited for this kind of site because they can fly closer to the methane plume and measure emissions in more detail. They can help identify where emissions occur and can support landfill monitoring. However, drone measurements also have uncertainties. In this project, wind direction was the largest source of uncertainty, because the calculated methane emission depends strongly on how the wind carries the plume through the measurement area. Wind speed and the estimated background methane concentration also affected the results.
The project also tested how processing choices influenced the drone-based emission estimates. The assumed surface roughness of the landfill affected how wind speed was calculated at flight height, which then changed the methane estimate. The method used to estimate background methane also mattered. A rolling background method gave lower emissions than a static background method, but sometimes followed the methane plume itself and may have removed part of the real methane signal.
These results show that models, drones, and satellites each have different strengths. The landfill model is useful for estimating long-term national emissions, but it cannot fully represent the detailed behaviour of one landfill. Satellites can cover large areas, but current technology may miss smaller or diffuse landfill emissions. Drones provide detailed site-level measurements and can help test whether model assumptions are realistic.
In the future, repeated drone measurements at different landfills, seasons, and weather conditions could help improve methane estimates in Sweden. Drone data should not replace national inventory models, but they can support them by showing how emissions behave in the real world. Combining models with direct measurements could reduce uncertainty and improve how methane emissions from the waste sector are reported.

• author: van Kampen, Fleur
• supervisor: Marko Scholze; Tim Arnold
• organization: Department of Earth and Environmental Sciences (MGeo)
• alternative title: A case study using UAV-based methane measurements at Filborna landfill to assess how site-scale emissions compare with Sweden’s national inventory estimates and associated uncertainties
• course: NGEM01 20261
• year: 2026
• type: H2 - Master's Degree (Two Years)
• subject: Earth and Environmental Sciences
• keywords: Physical Geography and Ecosystem Analysis, methane, landfill emissions, UAV measurements, greenhouse gas inventory, waste management
• publication/series: Student thesis series INES
• report number: 791
• language: English
• id: 9236729
• date added to LUP: 2026-06-22 12:36:13
• date last changed: 2026-06-22 12:36:13