Integrative modeling of POPs output flux from soil at a regional scale: A comprehensive approach

Abstract

Plot-scale natural attenuation models provide valuable insights into localized pollutant behavior but struggle to account for regional-scale hydrological processes. Existing research has predominantly concentrated on single processes, lacking comprehensive models to describe the output flux of persistent organic pollutants (POPs) by transport and transformation from soil at a regional scale. To address this gap, a model was developed by combining natural attenuation processes (e.g., degradation, volatilization, plant uptake) with hydrological transport processes (e.g., leaching, water washout, sediment transport). The model was validated using data from a petrochemical area in China and compared with the previous model (mass balance and neural networks model) to assess pollutant output flux. Results indicated that water washout was the dominant output pathway from soil for both Phenanthrene (Phe) (94.67 %) and Benzo(a)pyrene (BaP) (98.33 %). Phe exhibited a broader output flux range (0–67.8 mg∙m⁻²∙a⁻¹) compared to BaP (0–12.9 mg∙m⁻²∙a⁻¹), due to its higher volatility and solubility. Performance evaluation through 10-fold cross-validation yielded coefficient of determination (R²) values greater than 0.7 and root mean square error (RMSE) below 3 %, outperforming the previous model. Sensitivity analysis revealed that the soil organic carbon mass fraction (f_oc) was the most influential parameter at both a plot and regional scale. This study fills a gap in environmental research by providing a comprehensive model for accurate estimates of POPs output fluxes from soil.