Bayesian Networks in Environmental Risk Assessment: A Review

Area of Science:

• Environmental Science
• Ecology
• Risk Assessment

Background:

• Environmental risk assessment (ERA) evaluates adverse effects of human activities on ecosystems.
• Bayesian networks (BNs) offer a probabilistic approach to synthesize knowledge for ERA.
• Despite growing use in environmental modeling, BNs' formal application in ERA is unevaluated.

Purpose of the Study:

• To systematically review the application of Bayesian networks (BNs) in environmental risk assessment (ERA).
• To evaluate the scope, technical aspects, and utilization of BNs in peer-reviewed ERA studies.
• To provide guidelines and development ideas for enhancing BN use in ERA.

Main Methods:

• Systematic mapping protocol to identify relevant peer-reviewed publications.
• Analysis of studies employing BNs within an environmental risk context.
• Evaluation of model scope, technical features, and results interpretation.

Main Results:

• BNs have been increasingly applied in environmental risk assessment, particularly in recent years.
• Identified scope and technical aspects of BN applications in ERA.
• Found opportunities for advancing BN model implementation and participatory modeling in ERA.

Conclusions:

• Bayesian networks are a viable and increasingly adopted tool for environmental risk assessment.
• Further development in model implementation and participatory approaches can enhance BN utility in ERA.
• Guidelines are proposed to support the effective use of BNs in future ERA practices.