A Monte Carlo Simulation-Based Approach to Evaluate the Effect of Streamflow Measurement Error on Design Flood Estimates

Abstract

Flood frequency estimates are widely used in water resources engineering for the design of drainage structures such as culverts and bridges. These estimates are also essential for effective floodplain management. The discharge values used in flood frequency analysis are frequently obtained using stage-discharge rating curves. Consequently, errors inherent in stage measurements and stage-discharge rating curves adversely affect both river discharge and flood frequency estimates. These errors are especially significant for high-magnitude river discharge estimates produced by catastrophic flood events that are accompanied by unsteady flow conditions, heavy sediment loads, and debris. This study uses a Monte-Carlosimulation-based approach to evaluate the impact of these errors on design flood estimates. A log-Pearson type 3 distribution is used to obtain the necessary flood frequency estimates. The results are analyzed for several return periods that are commonly used in water resources engineering. The study’s findings show that for flood frequency analysis results to be useful for floodplain management and engineering design analysis, it is important that they explicitly account for discharge measurement errors.