Inverse modelling of transient water flow and reactive herbicide transport in vegetated weighable lysimeters

Intensive agricultural activities that result in the release of agrochemicals such as nitrate or pesticides pose a significant threat to groundwater quality and the safe supply of drinking and irrigation water. In this study, we investigate the transport and fate of four herbicides (terbuthylazine, metolachlor, prosulfuron, nicosulfuron) and some of their metabolites within two lysimeters over a period of 3 years. The studied lysimeters contain undisturbed soil profiles dominated by sandy gravel (Ly1) and clayey sandy silt (Ly2), and planted annually with maize.

Aims of this study include an improvement of flow and transport process understanding by identifying processes that contribute to the observed herbicide concentrations peaks in lysimeter outflow water and by determining dynamics and rates of those. Stable water isotopes at the outflow (δ¹⁸O) and drainage observations were used to calibrate the soil hydraulic and transport parameters of each soil. Within the modelling with HYDRUS-1D, the interaction between mobile and immobile soil water as well as the preferential flow with single and dual porosity approach were considered. In addittion, different assumptions on water uptake by roots and on root growth were compared. In a next step, reactive transport parameters for the transport of herbicides in lysimeter outflow were inversely estimated, which would eventually reach groundwater. Different modelling approaches for sorption kinetics were compared to identify governing processes.

Measured metabolites concentrations and analysed stable carbon isotopes (δ¹³C) of terbuthylazin and metolachlor in the drainage and findings from reactive transport modelling indicate a contribution of biodegradation activity, however, sorption processes seem to dominate herbicide fate in the lysimeters. The implementation of root water uptake could improve the explanation of observed lysimeter outflow rates. The dual porosity approach for flow and transport simulation resulted in a significant improvement of model representation compared to other approaches for soil hydraulic models. Surprisingly, non-equilibrium or non-linear sorption considerations for the charged herbicides nicosulfuron and prosulfuron did not lead to improved model results. In this study we could show that stable carbon isotopes can be utilized to identify biodegradation effects in the unsaturated zone. Furthermore, modelling results highlight that the consideration of various processes influencing unsaturated flow and reactive transport is crucial for decision-support models addressing sustainable groundwater management that aims at minimizing impacts of agrochemicals to soil and groundwater.