By John C. Carron, Edith A. Zagona, and Terrance J. Fulp. Published in the Proceedings of the ASCE Watershed Management 2000 Conference, Ft. Collins, CO, 2000.

Abstract: In recent years, management considerations for reservoirs have been expanded to include environmental and recreational objectives. Uncertainty modeling contributes to better reservoir management by quantifying the uncertainties and determining the sources of significant uncertainty in predicting reservoir and river conditions which affect environmental habitats and recreational conditions. RiverWare employs a firstorder, second-moment approach to modeling uncertainty. The method is described in this paper and applied to an example on Lake Mohave on the Lower Colorado River. Comparison of historical forecasts with observed data can provide managers with insight into operational policies which result in greater system uncertainty. Identifying these sources of uncertainty can guide managers in the development of future operational guidelines.