Setting Site Specific Critical Loads: An Approach using Endorsement Theory and Dempster–Shafer
Wadsworth, Richard A.; Hall, Jane R.. 2007 Setting Site Specific Critical Loads: An Approach using Endorsement Theory and Dempster–Shafer. Water Air and Soil Pollution: Focus, 7 (1-3). 399-405. 10.1007/s11267-006-9084-8Before downloading, please read NORA policies.
Download (421kB) | Preview
There is an increasing demand from conservation agencies for site-specific critical loads (CL); unfortunately, there is often very little specific information on a site to determine the important parameters needed to calculate the CL or on the spatial location of the “designated feature” in a site. Determining the most appropriate CL therefore involves using expert judegement to make decisions with incomplete and uncertain information. Endorsement Theory (Cohen, 1985) and Dempster–Shafer statistics (Dempster, 1967; Shafer, 1976) are, respectively, a decision-theoretic and a statistical technique for reasoning under those conditions (uncertainty and incompletness). A key reason for applying these techniques is that they make expert opinion explicit and available for scrutiny. Both techniques have been applied to the problem of setting an appropriate site specific CL, using heathland sites as a case study. Inital findings are encouraging; the uncertainty in expert judgement is made explict, the end results are intuitively reasonable and the methodology apparently acceptable to decision makers.
|Item Type:||Publication - Article|
|Digital Object Identifier (DOI):||10.1007/s11267-006-9084-8|
|Programmes:||CEH Programmes pre-2009 publications > Biogeochemistry
CEH Programmes pre-2009 publications > Environmental Informatics
|Additional Information. Not used in RCUK Gateway to Research.:||The original publication is available at www.springerlink.com|
|Additional Keywords:||Endorsement Theory, Dempster–Shafer, uncertainty, site specific critical loads|
|NORA Subject Terms:||Data and Information
|Date made live:||22 Jan 2008 14:52|
Actions (login required)