Scientific and engineering improvement of environmental risk management - Environmental Deterioration and Contamination: Problems and their Management - page 412

Figure 11.12 Environmental monitoring, risk assessment and risk reduction, the key tools of

environmental management.

estimation of unknown parameters, quality, impurities of chemical substances and their
use scenario carry the uncertainties during the risk assessment procedure. Exposure,
i.e., the predicted concentration of the potentially contaminating chemical substance
in a certain environment is a highly uncertain value, including uncertainties of the
chemical substance's properties, the environmental characteristics and the interaction
during the transport, fate and the effects of the chemical in the environment. Hazard
assessment resulting in the “no effect'' concentration of a chemical substance is also
highly uncertain due to the applied models—chemical models, single species bioassays
or microcosms for the ecosystem and in vitro or in vivo animal models for humans—as
well as the extrapolation methods both to the ecosystem and to humans. Both expo-
sure and hazard are uncertain, their ratio, the risk characterization ratio will be
exponentially uncertain.
The sources of uncertainties have been identified by WHO IPCS (2006) as scenario
uncertainty, model uncertainty and parameter uncertainty.
Scenario uncertainty is rooted in specifying a consistent scenario with the identified
use of the chemical substance. Scenario uncertainty includes wrong or incomplete infor-
mation on a chemicals use scenario (descriptive errors), aggregation errors (approxi-
mations for volume and time), errors of assessment (using an inadequate model) and
errors of incomplete analysis (overlooking an important exposure pathway). It is an
uncertainty in the concept, in the establishment of the conceptual model of the problem.
Model uncertainty is caused by the inadequacy of the model used for the scope
and purpose of the assessment. Models are simplifications of reality, with a distance
between model and reality. The model's uncertainty is due to this distance from the
realty, which depends on the extrapolation error, mistakes in the creation of the
model (oversimplification, false structure, incorrect parameters) and dependency errors