494 Multiple attribute decision models for BCG calibration

Wednesday, May 20, 2009: 4:30 PM
Vandenberg B
Jeroen Gerritsen , Tetra Tech Inc., Owings Mills, MD
The objective of the Biological Condition Gradient (BCG) model is to bridge quantitative, scientific information to management goals and endpoints.  Biologists are able to describe 5 or 6 distinct levels of the BCG that refer to the maintenance of structural and functional integrity of the original natural community.  The BCG Levels can be directly adopted as management goals, and as water quality criteria.  The central challenge to using the BCG directly in water management is translating the description of BCG levels and aquatic community changes among levels, into quantities that can be readily measured or calculated from monitoring data.  I explore one approach, which consists of formalizing expert knowledge and consensus on the BCG levels into a multiple attribute decision model.  Models developed for several northeastern US states are compared.  In the approach, a panel of biologists develops descriptions of the BCG levels, assigns sites to the levels, and identifies a set of rules that they use for the assessments. Multiple attribute decision models explicitly preserve and follow the rules used by the expert panel.  Consensus assessments built from the professional judgment of many experts result in a high degree of confidence in the assessments.  If formalized properly, any person (with data) can follow the rules to obtain the same level assignments as the group of experts.  The resulting decision criteria are transparent to stakeholders and the informed public.