Uncertainty Analysis of Water Levels of Sistan River and Reliability Analysis of Flood Control System

Uncertainty is always an element in the design process of flood control systems. Hence, efficient flood management needs uncertainty analysis. Computation of water levels at different floods magnitudes is a main task in flood management studies such as flood plain delineation for flood insurance studies, flood hazard mapping, and project planning for flood damage reduction investigations, design of levees, and design of bridges at stream crossings. Two of the sources of uncertainty associated with the computation of water levels, those of hydrologic uncertainty of estimation of design discharge and hydraulic uncertainty of channel capacity, are examined in this paper. The uncertainty of the channel capacity is determined by first-order uncertainty of the Manning's equation using Mean First-Order Second-Moment (MFOSM) method. The uncertainty of flow area, wetted perimeter, and the friction slope, as the parameters of Manning's equation, are expressed in terms of the variance of vertical, lateral, and longitudinal measurements performed with land surveying instruments. The uncertainty of Manning's n is characterized by the equation of standard deviation of n developed by Hydrologic Engineering Center, US Army Corps of Engineers. The uncertainty of design discharge, as hydrologic uncertainty, is investigated considering the parameters uncertainty and different probability distributions.
The uncertainty information is used to derive confidence limits for the water surface elevation of 100 year return period and to analyze the reliability of levees via direct integration method of solving the joint probability of design discharge and channel capacity as the system's loading and resistance respectively. In a case study application to the Sistan River, Sistan and Balouchistan Province, results show that ignoring the hydraulic uncertainties, the water level would be underestimated and the reliability overestimated.