Bayesian uncertainty assessment in multicompartment deterministic simulation models for environmental risk assessment

Abstract

SUMMARYWe use a special case of Bayesian melding to make inference from deterministic models while accounting for uncertainty in the inputs to the model. The method uses all available information, based on both data and expert knowledge, and extends current methods of 'uncertainty analysis' by updating models using available data. We extend the methodology for use with sequential multicompartment models. We present an application of these methods to deterministic models for concentration of polychlorinated biphenyl (PCB) in soil and vegetables. The results are posterior distributions of concentration in soil and vegetables which account for all available evidence and uncertainty. Model uncertainty is not considered.

A probabilistic approach to exposure risk assessment

“…We refer to exposure risk as the probability of disease occurrence in response to environmental contamination, whereas the process by which the potential adverse effects of exposure are assessed and characterized is known as exposure risk assessment. Bates et al (2003), amongst others, delineate four major activities involved in the exposure risk assessment process: (1) hazard identification identifies the potential hazards and the alleged health defects, (2) exposure assessment identifies populations, which could be exposed and the possible pathways, (3) dose-response assessment quantifies the relationship between levels of exposure and levels of potential adverse effects and, (4) risk characterization combines the results of the previous three steps to determine some outcome of interest.…”

Estimating Uncertainty in the Revised Universal Soil Loss Equation via Bayesian Melding

“…Bayesian melding has begun to be used for a range of problems, such as urban simulation (Ševčíková, Raftery, and Waddell 2007), whale management (Raftery, Givens, and Zeh 1995;Poole and Raftery 2000), forest ecosystem (Radtke, Burk, and Bolstad 2002;Radtke and Robinson 2006), environmental risk assessment (Bates, Cullen, and Raftery 2003), disease transmission (Spear et al 2002), pollution levels (Fuentes and Raftery 2005), and HIV prevalence modeling (Alkema, Raftery, and Clark 2007).…”

Continental-Scale, Data-Driven Predictive Assessment of Eliminating the Vector-Borne Disease, Lymphatic Filariasis, in Sub-Saharan Africa by 2020

“…Bayesian data-driven modeling frameworks have received considerable attention in this regard given their ability for not only facilitating the induction of a dynamical system from data, but also in the use of multiple data sources for constraining the parameters of a model to capture the local transmission features of a spatial setting [21,22,33,[39][40][41].…”