A Process-Based Approach to Model Pesticide Dynamics in Non-Floodplain Wetlands: WetQual-PEST

Abstract

Pesticides can pose significant environmental challenges once they enter water bodies. Wetlands are increasingly recognized as cost-effective, nature-based solutions for attenuating pesticide loads delivered to surface waters, yet pesticide cycling in wetlands remains poorly represented in many water-quality models. This thesis develops and evaluates a physically based modelling framework (WetQUAL-PEST) to capture pesticide fate in treatment and non-floodplain wetlands. Stochastic calibration against 1991 Atrazine data in Des Plaines Project Wetland (Illinois, USA) produced daily concentration Nash–Sutcliffe efficiencies of 0.81–0.92 (median 0.91) and a median PBIAS of 2.35%. Applying the plant-extended framework to lindane data from the shallow, three-stage Hájek Wetland+ system (Czech Republic) highlighted photolysis rate, water-column porosity and Henry’s law constant as most sensitive parameters, whereas macrophytes accounted for approximately 3% of the total lindane mass removal but contributed to a ~5% reduction in peak aqueous concentrations. These findings demonstrate the importance of incorporating detailed vegetation processes, daily hydrological data, and data driven uncertainty analysis to improve predictive skill of wetland models. Overall, this research demonstrates that WetQUAL-PEST is well suited for screening daily pesticide loads and concentrations in wetlands, and quantifying key process controls, across diverse compounds and wetland configurations.