Quantifying the influence of soil forming factors on carbon stock in the High Fens region and predicting its spatial distribution

Details

Supervisors

Vanacker, Veerle

Faculty

Faculté des sciences

Degree label

Master [120] en sciences géographiques, orientation générale, à finalité approfondie

Abstract

enThis study investigates the environmental factors controlling Soil Organic Carbon (SOC) stocks in the High Fens peatlands and evaluates the potential of spatial modelling approaches to predict their distribution. An integrated approach combining field sampling, laboratory analyses, remote sensing, and statistical modelling was applied across the plateau. A total of 36 soil pits were sampled in contrasting environmental settings. SOC stocks were estimated using carbon concentration, dry bulk density, and soil depth measurements. Environmental predictors including topography, hydrology, vegetation indices, soil type, and land cover were analyzed to assess their influence on SOC variability. Results indicate that land cover and soil type are the main factors controlling SOC distribution. Peaty wetland environments displayed significantly higher SOC stocks than forested silty-stony soils. While topographic and hydrological variables contributed to explaining peatland dynamics, their direct statistical influence remained secondary. A Gamma Generalized Linear Model proved more suitable than classical linear regression due to the skewed distribution of SOC values. In parallel, a Random Forest model successfully predicted the spatial distribution of SOC and highlighted the importance of non-linear environmental interactions. Overall, the study demonstrates that combining field observations with remote sensing and machine-learning approaches provides an effective framework for assessing SOC dynamics in peatland ecosystems. The results also emphasize the importance and vulnerability of the High Fens as a major long-term carbon reservoir under increasing anthropogenic and climatic pressures.