Fitting the Curve
Finding the best fit for flood frequency analyses
No matter how much data we collect, there is bound to be inherent uncertainty in our rainfall and runoff estimates. Flood frequency analyses rely on curve fitting approaches that can vary substantially with the selected distribution. Why would you choose one or the other?
Join us for an overview of the available options and learn how sensitive the results are to the selected parameters. We’ll hear from Haden Smith, the developer of BestFit, a public domain software that has recently been validated against HEC-SSP, FLIKE, and other FFA tools. BestFit performs distribution fitting and Bayesian estimation from a choice of thirteen probability distributions.
We’ll also cover an overview of regional regression techniques, how to identify outliers, and how to interpret confidence limits. Our panelists will also demonstrate the effect of incorporating new extreme event observations on predicted trends along with the role of climate change in Bayesian inference.
US Army Corps of Engineers, Risk Management Center