Human Health Risk Assessment under Uncertainty Via Possibilistic Approaches

Pollutants from different manmade or artificial sources to the environment greatly impact human beings. The natural environment contributes the most significant proportion to human radiation exposure. The establishment of industries is another factor of environmental pollution. The primary pathways of exposure to human beings are Inhalation of Air, Intake of Water, Intake of Food, etc.

Consequently, an assessment is required to carry out in order to determine the intensity of impact on human health by hazardous substances when released into the environment. Accordingly, health risk assessment is carried out to measure the amount of risk to humans from hazardous substances present in the environment. For this purpose, the health risk assessment models of the Environmental Protection Agency, USA (EPA-USA, 2010) are adopted. Health risk assessment models are functions of parameters that are more often uncertain in nature. Various types of fuzzy numbers are considered to deal with the uncertain model parameters, and different arithmetic techniques are adopted to combine the fuzzy input numbers. Henceforth, human health risk assessments are performed.

Then, the Dempster-Shafer theory of evidence (DST) is studied, and for that, first, we propose three strategies of assigning BPA when only three values of the parameter are known, viz. minimum, maximum, and most likely value. Likewise, we characterize the arithmetic combination of interval focal elements. We discuss fuzzy focal elements and define Dempster’s rule of combination for fuzzy focal elements. We extended the idea of the arithmetic combination of interval focal elements for fuzzy focal elements. Risk assessment is performed for interval focal elements using the proposed arithmetic combination of focal elements (PACF), and Yager arithmetic combination of focal elements (YACF), and results are analyzed. Finally, risk assessment is performed using fuzzy focal elements.

In health risk assessment models, the representation of a few parameters is probabilistic in which mean and standard deviation (variance) are available in the form of interval or fuzzy numbers. In contrast, a few other parameters are considered possibilistic. Therefore, risk assessment is also carried out using the hybrid framework.

Furthermore, then risk assessment is performed utilizing Probability-Possibility transformations. Likewise, we have presented a comparative case study of uncertainty propagation by taking three different scenarios using probability-possibility transformation satisfying consistency conditions.