Issues on environmental risk assessment

Risk assessment on the environmental sciences is different from health risk assessment in that there is no experimental repetition. Lack of experimental repetition can be solved using random function models. However, the impact of random function model choice is not always fully understood. Examples of this impact comparing the often used multi-Gaussian model to alternative models with high connectivity of extreme values show the importance of carefully tailoring the selection of the random function model to reproduce those characteristics of the data that are most consequential to the risk being assessed.