Bayesian networks in environmental and resource management

Abstract

This overview article for the special series, "Bayesian Networks in Environmental and Resource Management," reviews 7 case study articles with the aim to compare Bayesian network (BN) applications to different environmental and resource management problems from around the world. The article discusses advances in the last decade in the use of BNs as applied to environmental and resource management. We highlight progress in computational methods, best-practices for model design and model communication. We review several research challenges to the use of BNs in environmental and resource management that we think may find a solution in the near future with further research attention.

Socio-hydrological modelling: a review asking "why, what and how?"

“…The advantages of using Bayesian networks come directly from the modelling approach: uncertainties are directly and explicitly accounted for since all inputs and outputs are stochastic (Kelly et al, 2013), and the use of Bayes' theorem means that probability distributions of output variables may be "updated" as new knowledge and data become available (Barton et al, 2012). Using Bayes' theorem also allows the use of prior knowledge, since distributions of output parameters are required to be specified prior to model start-up (to then be changed and updated), and these prior distributions may be informed by the literature (Barton et al, 2012). The fact that there are relationships (albeit stochastic rather than deterministic) between variables also means that direct causal links between variables may be established (Jellinek et al, 2014).…”

“…The fact that there are relationships (albeit stochastic rather than deterministic) between variables also means that direct causal links between variables may be established (Jellinek et al, 2014). The drawbacks in using BNs are the difficulties present in modelling dynamic systems, since BNs tend to be set up as "acyclic" (Barton et al, 2012) (though object-oriented (Barton et al, 2012) and dynamic Bayesian networks (Nicholson and Flores, 2011), which can model dynamic feedbacks, are being developed and becoming more prevalent), and in the potential statistical complexities present. A Bayesian network may be seen as a stochastic version of a system dynamics model, and so many of the criticisms of SD models may also be applicable to BNs; in particular, the fact that BNs are largely based around datadefined relationships (as opposed to physically determined or process-based relationships) between variables means that BNs can only yield deterministic (albeit stochastically deterministic) results that arise from data.…”

Tales from a thousand and one ways to integrate marine ecosystem components when assessing the environmental status

“…Only quantitative up to a certain level Probabilistic (Barton et al, 2008(Barton et al, , 2012Lehikoinen et al, 2013Lehikoinen et al, , 2014 Bayesian statistics Produces a probability estimate of how likely the area is in GEnS; managers can decide the acceptable undertainty…”

“…Barton et al (2012) demonstrate how to use the probabilistic approach in the DPSIR framework in the case of eutrophication management. There are several other examples in the recent literature about how to evaluate various management measures under uncertainty to optimize one target, such as eutrophication (Barton et al, 2008;Lehikoinen et al, 2014) and oil spill severity (Lehikoinen et al, 2013).…”

Inferring Missing Climate Data for Agricultural Planning Using Bayesian Networks

“…To overcome these hurdles, we propose to use a Bayesian network (BN), which is a mathematical model that graphically represents conditional probabilistic dependencies between variables. BNs can deal with uncertainty, missing data, missing (hidden) variables and small datasets; it is possible to learn the graphical structure and the parameters of the model from data, literature, expert knowledge or a combination of all [11][12][13][14]. Another practical advantage of using BN is the availability of free software [15,16].…”