|
(21-19) Uncertainty analysis in geo-referenced probabilistic ecological risk assessment (PERA). Verdonck, Frederik*,1, Jaworska, Joanna2, Janssen, Colin3, Vanrolleghem, Peter1, 1 Ghent University, Coupure Links 653, Gent, Belgium2 Procter & Gamble, Temselaan 100, Strombeek-Bever, Belgium3 Ghent University, Jozef Plateaustraat 22, Gent, Belgium ABSTRACT- The goal of a risk assessment is to estimate the likelihood and the extent of adverse effects occurring to humans and ecological systems due to possible exposure(s) to substances. Environmental risk assessment is based on the comparison of an environmental concentration (EC) with species sensitivity (SS) derived from chronic toxicity data. In a probabilistic framework, EC and SS are treated as random variables coming from probability distributions (respectively ECD and SSD). As a result, the outcome is a risk distribution. The goal of this poster is to extend the PERA framework with an uncertainty analysis. Currently, uncertainty is not often considered in the ECD, SSD and the resulting risk. The methodology for calculating the risk with its uncertainty interval is based on Monte Carlo analysis. In PERA, the contribution of spatial variability to the ECD can be quite high. By geography referencing the risk assessment, the spatial variability is explicitly accounted for in each local risk assessment. Of course, the geo-referenced risks can then again be combined in a lumped, non-geo-referenced but refined risk making it more realistic. As a result, it improves decision support of individual chemicals. A case study was performed on the pesticide atrazine in the river catchments of Flanders (Belgium). Key words: atrazine, variability, uncertainty |
