Uncertainty Analysis of Lake Erie Net Basin Supplies as Computed Using the Residual Method

Abstract

<p>The Lake Erie net basin supply (NBS) is defined as the net volume of water entering or exiting the lake from its own drainage basin over a specified period of time. NBS can be computed using either the component or residual method. In this research, an uncertainty analysis was performed on the residual method of computing Lake Erie NBS using both the First-Order Second Moment (FOSM) method and a Monte Carlo simulation approach. Uncertainties in each of the various inputs, including the inflows, outflows, and change in storage, among other sources, were first defined through analysis of data, when available, or with alternative methods when necessary. Estimating the unce11ainty in each of the NBS model inputs was found to be the most difficult and time consuming component of this study, and also the component prone to the most subjectivity. The results obtained using the FOSM and Monte Carlo approaches were found to be nearly identical when applied to the residual method of computing Lake Erie NBS. Comparison of the results of this study to the results from other research showed that the overall uncertainties in NBS are of similar magnitude. However, the uncertainty in the change in storage was found to be greater than estimates given in previous studies, and greater than perhaps generally believed, being of a similar magnitude to the uncertainty in the Lake Erie inflows and outflows, which have normally been cited as the greatest sources of uncertainty in Lake Erie residual NBS.</p>