Peatlands store vast amounts of carbon but are increasingly at risk of becoming carbon sources due to climate change and human disturbance. This thesis aims to understand the mechanisms driving carbon loss dynamics, particularly soil respiration (i.e., a key ecological process that releases CO2 from the soil into the atmosphere), in these vulnerable systems. Using multi-sensor Unmanned Aerial Vehicles (UAVs) remote sensing, field monitoring, and incubation experiments in a temperate sloping peatland (Hautes Fagnes, Belgium), we show that peat thickness is strongly associated with topographic features at the macro-scale, while SOC storage is more related to micro-topographic variability. Second, UAV-borne data were utilized to map carbon storage spatial distribution and predict the spatiotemporal dynamics of soil temperature and moisture, the main controls on spatiotemporal variability in soil respiration. Next, by modelling hourly soil CO2 fluxes and mapping daily daytime soil CO2 fluxes, we identified soil CO2 emission hot spots and hot moments, which contributed disproportionately to the total soil CO2 emissions across the landscape. Last, soil organic matter functional composition is the primary control of potential soil respiration, while topography-induced thermal-hydrological conditions exert cascading effects by regulating soil biogeochemical properties. Overall, this thesis shows that integrating UAV remote sensing can improve our mechanistic understanding of peatland soil respiration across heterogeneous landscapes. The findings provide insights into carbon dynamics and support peatland conservation and climate change mitigation.
Li, Y. (2025). Soil respiration dynamics in a temperate peatland: insights from Unmanned Aerial Vehicle (UAV) remote sensing and in-situ monitoring. https://hdl.handle.net/2078.5/266387