| Watershed management requires holistic, inclusive analyses of decisions and their potential impacts. This can be difficult for many reasons. In most cases, a wide variety of stakeholders with competing interests need to be included in the decision-making process and have their interests fairly represented. Relationships between variables of interest to different stakeholders need to be established and clearly defined. This must often be done in the absence of data of sufficient quantity or quality and in politically heated situations.; For this dissertation, a Bayesian Decision Network (BDN) probabilistic modeling framework was developed that satisfies many of these requirements and complications. This methodology is applied to streamflow management in the Grand Canyon reach of the Colorado River and to phosphorous management in East Canyon Reservoir in Northern Utah. A discussion of the use of BDNs to account for uncertainty in the development of Total Maximum Daily Loads (TMDL) is also presented as an approach for addressing the TMDL margin of safety requirement. Finally, a conceptual BDN-enabled decision support system framework with implementations in both an Internet GIS setting is presented. |
