Integrated Simulation of Polycyclic Aromatic Compounds in the Athabasca River Basin

The environmental risks posed by polycyclic aromatic compounds (PACs) associated with increasing oil and mineral mining activities have become a major concern in Alberta. Due to their diversity and complex behavior, basin-scale surface water PACs simulation based on traditional modeling tools is restricted, thereby hindering decision-making. To address this, we designed an integrated simulation framework that combines predictive relationships among PACs with mechanism-based models and implemented it in the Athabasca River basin (ARB) in Alberta. The predictive relationships were obtained through a preliminary analysis based on principal component analysis, clustering, and regression. For mechanism-based simulation, a Python-based Soil Water Assessment Tool-Load Calculator (SWAT-LC) was developed and coupled with SWAT and the Water Quality Analysis Simulation Program 8 (WASP8) to describe the fate and transport of PACs in both the terrestrial and aquatic systems. Our results show that: 1) Out of 76 PACs studied, two clusters were identified, including one with 66 PACs exhibiting seasonal patterns, and another with 10 PACs marked by significant uncertainties. Chrysene and naphthalene were chosen from the respective cluster as representatives for mechanism-based modeling; 2) The established mechanism-based model demonstrated overall acceptable to satisfactory performance for chrysene at different sites (NSE=-0.28~0.33, d=0.34~0.71, PBIAS=0.09%~36.68%), although was less successful in describing the fluctuations of naphthalene; 3) Evidence indicates that seasonalities in PACs are petrogenic and are mainly driven by soil-water processes, while surface wash-off in the oil sands region and wet depositions lead to concentration spikes in river water; 4) The predictive relationships of the other 74 PACs are robust along the Athabasca River mainstem, showing great potential for facilitating rapid decision-making in the future.