Related Experiment Video
Updated: Sep 13, 2026

Vegetated Treatment Systems for Removing Contaminants Associated with Surface Water Toxicity in Agriculture and Urban Runoff
Published on: May 15, 2017
VERBER - A regional scale screening tool to identify freshwater at risk of pesticide contamination
Anna Lindahl1, Gustaf Boström2, Mikaela Gönczi2
1Department of Soil and Environment, Swedish University of Agricultural Sciences, P.O. Box 7014, Uppsala, SE-75007, Sweden.
Abstract:
Water authorities in European countries are required to carry out regular impact assessments of pesticides in their water bodies under the EU Water Framework Directive (Directive 2000/60/EC; WFD). This requirement poses a significant challenge for water authorities in Sweden, which has up to 100,000 lakes and more than 500,000 km of water-courses. To support their work under the WFD, we developed VERBER for surface waters, a GIS-based model for pesticide risk assessment and management. VERBER predicts average concentrations of active substances originating from arable fields at the outlet of sub-catchments in inter-connected networks of lakes and water-courses in Sweden. The modelling approach combines a parsimonious steady-state mass balance model with a classification tree trained on Swedish environmental monitoring data, to estimate the fraction of an applied pesticide lost to surface waters based on hydrological characteristics of soils, pesticide properties and application season. According to VERBER model runs for seven active substances (either priority substances in the WFD with particularly hazardous properties or river basin-specific pollutants) at the national scale, 2.2% of sub-catchments in Sweden are at risk of being affected by pesticide concentrations that constitute chronic toxicity to aquatic organisms. A comparison between average surface water concentrations predicted by VERBER and measured concentrations of the seven active substances and two additional substances that often exceed the EU drinking water limit value (0.1 μg L-1) indicated that the model predictions are generally realistic, although it does also seem to be conservative in a few cases.

