|
1L (2) - Exposure Modelling (TH13/3) Separation of true uncertainty and interindividual variability in exposure modelling: a case study concerning the coherence of exposure and intake standards of toxic pollutants. Ragas, Ad1, Büchner, Frederike1, Huijbregts, Mark1, 1 University of Nijmegen, Nijmegen, The Netherlands ABSTRACT- Input parameters of exposure models can be uncertain due to a lack of knowledge (true uncertainty) or due to variability (i.e., interindividual variability). An ordinary Monte Carlo simulation combines both types of uncertainty in one output distribution, which impedes an unambiguous interpretation of the results. Separation of true uncertainty and interindividual variability can be realized by means of a nested Monte Carlo simulation. In this study, nested Monte Carlo simulation was applied to the human exposure model NORMTOX (Ragas & Huijbregts, Regul. Toxicol. Pharmacol. 27(3), 251-264, 1999), which predicts the lifetime-averaged daily intake from soil, air, surface water, drinking water and food products. NORMTOX was used to estimate the intake of 54 different substances under the assumption that all exposure media are polluted up to their environmental quality objective (EQO). The estimated intake distribution of each substance was compared with its acceptable daily intake (ADI). For nine substances, EQOs and ADIs are clearly incoherent, i.e., the ADI is exceeded if all exposure media are polluted up to their respective EQO levels. The results also indicate that the output variance of the estimated intake levels is mainly caused by interindividual variability. True uncertainty has a significant impact only on the intake levels of lead, cadmium, thiram and methylmercury. The main source of this true uncertainty is the intake of soil particles, which is an important, but highly uncertain, exposure parameter for these substances. Key words: nested Monte Carlo simulation, exposure, uncertainty, variability |
