Spatial modelling of polycyclic aromatic hydrocarbon distribution in a Canadian ice wedge polygon tundra landscape

Permafrost thaw due to climate change raises concerns about the potential remobilisation of organic pollutants such as Polycyclic Aromatic Hydrocarbons (PAHs) in Arctic environments. This study aims to identify environmental controls and create maps of PAH storage in two catchment-scale study areas, Komakuk Beach and Ptarmigan Bay, on the Canadian Beaufort coast (Yukon), while dealing with the lack of data concerning organic contaminants in Arctic soils. Soil samples were collected using a stratified random sampling design based on the catchment area and quaternary geology. The samples were analysed for 22 different PAHs, including 16 PAHs classified as priority pollutants by the US EPA, using accelerated solvent extraction and gas chromatography - triple quadrupole mass spectrometry. The dataset included measurements of PAHs, soil organic carbon content (SOC %), depth, and landform types from permafrost sediment and soil samples. Significant differences in PAH concentrations were found across soil depths and between low-centered polygon (LCP) and high-centered polygon (HCP) landforms, reflecting varying biogeochemical and hydrological dynamics in permafrost degradation forms. Random Forest models were used to predict spatial PAHs distributions in the study areas, divided by molecular weight and depth intervals, using SOC% data from digital soil mapping as an environmental support variable. The modelling approach showed promise for intermediate soil layers but faced challenges in the surface and deep layers owing to data limitations. This study addresses a critical knowledge gap and demonstrates the potential of upscaling PAH distribution data in Arctic permafrost regions using limited ground data. These findings highlight the complex relationships between PAHs, soil carbon, and permafrost landforms, emphasising the need to consider these factors when assessing contaminant dynamics in thawing permafrost. This study identifies spatial patterns and landscape-level factors for PAH upscaling, which can support further studies across Arctic permafrost regions.