Monte Carlo analysis and Bayesian decision theory for assessing the effects of waste sites on groundwater, I: Theory

Marín, Carlos M.; Medina, Miguel Á.; Butcher, Jonathan B.

doi:10.1016/0169-7722(89)90002-8

Cited by 35 publications

(14 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…We state that the worth of sampling data can be best evaluated by the use of Bayesian decision analysis, which has been applied for contaminated soil and groundwater by several authors, for example, Massmann and Freeze (1987), Marin et al (1989), Freeze et al (1990Freeze et al ( , 1992, Massmann et al (1991), James and Freeze (1993), James and Gorelick (1994), James et al (1996), Rosén and Wladis (1998), Wladis et al (1999), Rosén (2002), Back (2003), andNorrman (2004). A special application of Bayesian decision analysis is referred to as Bayesian data worth analysis, which is described below, primarily based on the works by Freeze et al (1992) and James and Freeze (1993), although it is a wellestablished methodology in Bayesian decision theory, see for example, Lindley (1997).…”

Section: General Description Of Data Worthmentioning

confidence: 99%

Calculating the optimal number of contaminant samples by means of data worth analysis

Norberg

Rosén

2006

Environmetrics

View full text Add to dashboard Cite

SUMMARYThis is a study of Bayesian data worth analysis in environmental clean-up applications. Its focus is on calculating the worth of simultaneously taking several (soil) samples in a small homogeneous area and on finding the optimal number of samples to take, by relating the reduction in risk cost from sampling, that is, data worth, to the cost of the samples. Even though the cost of one sample may be higher than the risk cost reduction it provides, this study shows that several samples may be cost-efficient. This is mainly due to two factors: one is that the unit sample cost often decreases as the number of samples increase; another, more important, factor is that the data worth of several samples typically is higher than the worth of fewer samples.

show abstract

Section: General Description Of Data Worthmentioning

confidence: 99%

Calculating the optimal number of contaminant samples by means of data worth analysis

Norberg

Rosén

2006

Environmetrics

View full text Add to dashboard Cite

show abstract

“…The derivation of a realistic loss function, however, is not an easy task, as is shown in such works as Rodriquez-Iturbe (1976a, 1976b) and Bogardi and Bardossy (1985). Marin et al (1989) presented a method for making regulatory decisions regarding waste disposal facilities that incorporated uncertainty and allowed for a small set of monitoring alternatives to be evaluated based on the predicted reduction in the concentration variance.…”

Section: Variance-reduction Analysismentioning

confidence: 99%

Long-term Monitoring Plan for the Central Nevada Test Area

Hassan¹

2003

View full text Add to dashboard Cite

EXECUTIVE SUMMARYThe groundwater flow and transport model of the Faultless underground nuclear test conducted at the Central Nevada Test Area (CNTA) was accepted by the state regulator and the environmental remediation efforts at the site have progressed to the stages of model validation and long-term monitoring design. This report discusses the long-term monitoring strategy developed for CNTA. Subsurface monitoring is an expensive and time-consuming process, and the design approach should be based on a solid foundation. As such, a thorough literature review of monitoring network design is first presented. Monitoring well networks can be designed for a number of objectives including aquifer characterization, parameter estimation, compliance monitoring, detection monitoring, ambient monitoring, and research monitoring, to name a few. Design methodologies also range from simple hydrogeologic intuition-based tools to sophisticated statistical-and optimization-based tools.When designing the long-term monitoring well network for CNTA, a number of issues are carefully considered. These are the uncertainty associated with the subsurface environment and its implication for monitoring design, the cost associated with monitoring well installation and operation, the design criteria that should be used to select well locations, and the potential conflict between different objectives such as early detection versus impracticality of placing wells in the vicinity of the test cavity. Given these considerations and the literature review of monitoring design studies, a multi-staged approach for development of the long-term monitoring well network for CNTA is proposed. This multi-staged approach will proceed in parallel with the validation efforts for the groundwater flow and transport model of CNTA. Two main stages are identified as necessary for the development of the final long-term monitoring well network for the site.The first stage is to use hydrogeologic expertise combined with model simulations and probability based approaches to select the first set of monitoring wells that will serve two purposes. The first is to place the wells in areas likely to encounter migration pathways thereby enhancing the probability of detecting radionuclide migration in the long run. The second objective is crucial in the short run and is aimed at using this set of wells to collect validation data for the model. The selection criteria should thus balance these two objectives. Based on the results of the validation process that progresses concurrently with the first monitoring stage, either more wells will be needed in this first stage or the second stage will be initiated. The second monitoring design stage will be based on an optimum design methodology that uses a suitable statistical approach, combined with an optimization approach, to augment the initial set of wells and develop the final long-term monitoring network.The first-stage probabilistic analysis conducted using the CNTA model indicates that the likelihood of migration away from th...

show abstract

“…Finding optimal distribution functions for model transport parameters or solving the inverse problem for the contaminant transport equation has two major problems Bayes' theorem, we have -P(IAI A)P(A)dA [14] where P(AIIA) is the updated or posterior distribution of the random transport variable A that predicts contaminant concentrations at a pumping well that are very close to actual measured concentrations. Figure 5 represents the steps in the development of transport models for a contaminated aquifer and how distribution functions for model transport variables are adjusted by Bayesian updating to provide better estimates of these variables for a particular contaminated aquifer (53)(54)(55)(56)(74)(75)(76)(77) [15] for noncarcinogens R (D) = 0, for D< Dt [16] where D£ is the threshold dose below which there is no observed adverse effect. In actuality, zero exposure doses cannot be achieved, and risk assessment is concerned with the determination of safe exposure doses or the distribution of safe exposure doses that are greater than zero but produce a negligible increase in cancer incidence over background.…”

Section: Contaminant Transport Modelsmentioning

confidence: 99%