Decision-making in environmental management is complex due to the multiplicity arid
diversity of management objectives and technological choices. This suggests that modelers
and experts could utilize (I) multiple-criteria decision-making (MCDM) approaches to
assist stakeholder groups in integrating and synthesizing relevant data and information to
address ecological and socio-economic concerns and (2) uncertainty approaches to quantify
the risks related to the impact of decision alternatives. Since decisions made under
uncertainty and MCDM methods have been studied almost independently, most of the
MCDM approaches do not address the uncertainties of real world decision situations.
This dissertation presents the use of a MCDM methodology and its related decision-making
tool, RESTORE. RESTORE is an integrative geographical information system-based
decision-making tool that was developed to help watershed councils prioritize and
evaluate restoration activities at the watershed level. RESTORE's deterministic
performance evaluation module is developed from experts' knowledge and experiences.
However, to filly address the complexity of the various landscape processes and human
subjectivity, RESTORE should involve uncertainties inherent to experts' knowledge. No
single method is able to model all types of uncertainty, therefore the examination of
various uncertainty theories is critical before selecting one best suited to a specific decision
context. This work explores three uncertainty theories: certainty factor model, Dempster-Shafer theory, and fuzzy set theory. To evaluate these methods in a MCDM watershed
restoration context, we (1) identified criteria to assess the suitability of a method for a
specific MCDM context, (2) characterized each theory in terms of the identified criteria
using RESTORE, and (3) applied each theory using RESTORE. Special emphasis was
given to the development of a comprehensive fuzzy MCDM methodology.
Uncertainty-based MCDM approaches provide a valuable tool in analyzing complex
watershed management issues. When used properly, the proposed MCDM methodology
allows decision-makers (DMs) to explore a broader range of drivers and consequences.
The inclusion of uncertainty analysis provides DMs with meaningful information on the
quality of the evidence supporting the impact of a decision alternative, allowing them to
make more informed decisions.