You are here:
Characterizing Assets, Threats and Solvability with Bayesian Networks to Support Spatial Prioritization for Environmental Management
Citation:
Carriger, John F. AND Susan H. Yee. Characterizing Assets, Threats and Solvability with Bayesian Networks to Support Spatial Prioritization for Environmental Management. SETAC North America 43rd Annual Meeting, Pittsburgh, PA, November 13 - 17, 2022.
Impact/Purpose:
This poster will be presented at the Society of Environmental Toxicology and Chemistry (SETAC) 2022 North America meeting in Pittsburgh. The poster describes an adaptation of the Assets, Threats and Solvability (ATS) framework with better consideration of uncertainty and the trade-offs among the components of the ATS hierarchy. The approach may be of benefit for screening-level environmental assessments where spatial prioritization is necessary. This type of context is often found in remediation and setting environmental protection priorities for spatial areas or regions.
Description:
A screening-level spatial prioritization framework is proposed using Bayesian networks and Stefan Hajkowicz and colleagues’ Assets, Threats, and Solvability (ATS) framework. The ATS framework is based on multicriteria decision analysis and provides a useful way for organizing information needed in prioritization assessments. In ecological risk assessment, prioritizing spatial regions for protection often includes a consideration of the components of the ATS framework- the assets are often related to valued ecological attributes, the threats are often from the chemical hazards, and solvability is related to the degree of effectiveness for restoring or protecting a location. The calculations necessary to implement ATS are possible with or without a Bayesian network. However, the Bayesian network can also allow incorporation of uncertainties for each of the component attributes. A schematic for using the ATS with Bayesian networks follows the decision analysis framework and starts with a selection of a decision context. Next, the ATS criteria are developed using an objectives hierarchy. An objectives hierarchy is especially useful for defining the criteria for a decision and for ensuring the criteria are internally cohesive and robust for analyzing trade-offs. A Bayesian network model is then developed to calculate the ATS hierarchy and compute prioritization levels as well as levels of the individual components of the ATS. Depending on the context, the ATS Bayesian network can be used for a variety of spatial scales and types from a large region to a grid cell. The ATS framework with Bayesian networks is potentially useful for summarizing complex information needed for protection, conservation, and restoration decision contexts where spatial differentiation and prioritization are key components of the environmental assessment.