Chemical exposure and risk in river basins: Bayesian integration of environmental model simulations and monitoring data

The EU's Chemicals Strategy for Sustainability aims to incorporate environmental risk of combined exposure, stemming from unintentional chemical mixtures in the environment, into regulatory risk assessments. Within the ENCORE (ENvironmental CO-exposure and Risk Estimation) project, a probabilistic modeling framework was developed for assessing chemical risk to aquatic ecosystems at the watershed level across Europe. The developed probabilistic framework can synthesize multiple sources of information at EU level, as well as uncertainty in both exposure and hazard information. The framework builds upon a chemical exposure model originally used in the EU project SOLUTIONS for large-scale European domains, which is being further developed in ENCORE. More specifically, a workflow was developed for a Bayesian network (BN) model that can update prior probabilities of chemical exposure derived from process-based simulation (predicted exposure) data. Key data sources integrated into the BN included the pan-European publicly available dataset Waterbase Water Quality (WISE-6), from the Water Information System for Europe managed by the European Environment Agency (EEA). Bayesian updating is used to integrate this new evidence (chemical monitoring data) improving accuracy of risk calculation. A pilot study, using a subset of pesticides in Belgium, was selected to develop and test the implementation of this probabilistic approach using BNs. This pilot will serve as a proof-of-concept before this risk modeling approach is scaled up to larger European regions. The goal of this approach is to identify chemicals with high contribution to risk in the aquatic environment, by accounting for spatial exposure patterns in watersheds across large regions of Europe and temporal patterns across months.