On the relevance of spray drift from boom application for catchment-scale pesticide fate assessment
- Pesticides are essential to modern agriculture, yet their movement through the environment poses risks to non-target ecosystems. Landscape-scale fate models are key tools for assessing environmental exposure, but they must adequately represent all relevant transport pathways to produce reliable predictions. Spray drift is a critical yet underrepresented pathway in landscape-scale pesticide fate models. This thesis addresses this knowledge gap by developing and applying a mechanistic modeling approach, the Droplet and Atmospheric Dispersion drift (DAD-drift) model, to simulate spray drift deposition from boom sprayers across complex and heterogeneous agricultural landscapes.
The DAD-drift model integrates a mechanistic droplet model, a micrometeorological module, and a three-dimensional Gaussian puff dispersion model to simulate the motion, evaporation, and atmospheric transport of spray droplets at high spatial resolution. Evaluation against two field-scale drift studies that cover a wide range of environmental conditions and nozzle types demonstrated very good agreement between predictions and observations, confirming accurate and robust model behavior across diverse application scenarios.
The role of spatial representations of application areas and non-target areas was evaluated by employing DAD-drift in virtual experiments, comparing its predictions with simplified algorithms previously applied in eco-hydrological modeling. The findings indicate that spatial simplifications can introduce errors of an order of magnitude in predicted pesticide deposition, with major consequences for subsequent exposure assessment. A complementary analysis identified the droplet size distribution and the distance to non-target areas as key drivers of spray drift deposition at the landscape scale. Crucially, the effects of wind speed, temperature, and humidity were found to be distance-dependent, reflecting the shift from near-field to far-field drift regimes and highlighting the necessity of scale-aware modeling strategies.
Finally, the DAD-drift model was integrated into a high-resolution Soil and Water Assessment Tool (SWAT+) eco-hydrological model of an agricultural catchment in Germany, extending the model with a previously absent drift component. The model setup was constructed entirely from publicly available datasets (e.g., terrain, soil, land use, hydrology, and meteorology), and included realistic multi-year crop rotations derived from field-level records. The catchment hydrology was calibrated and validated against observed discharge, achieving very good model performance. The coupled framework simulated both hydrological and atmospheric pesticide transport pathways across multiple years and active ingredients. Incorporating spray drift led to substantially higher predicted environmental concentrations in adjacent surface waters, particularly for substances that are otherwise less mobile due to strong sorption to soil. Ensemble-based hydrological parameterization revealed considerable uncertainty in predicted concentrations, underscoring the importance of uncertainty assessment in landscape-scale fate evaluations.
Taken together, this work delivers the first mechanistic and transferable integration of spray-drift modelling into catchment-scale eco-hydrological modelling, providing a strong framework for linking landscape hydrology with atmospheric drift processes, thereby strengthening the scientific basis for environmental risk assessment, mitigation planning, regulatory evaluation, and ecological impact assessment.