Extreme Precipitation Analysis for Engineering Applications: Analyzing, Moving, Scaling, and Stochastically Generating Storms

An evaluation of hydrologic risks to dams and infrastructure requires examination of extreme precipitation events and determination of watershed precipitation frequency estimates for application in hydrologic modeling efforts. These models require detailed spatial and temporal information on precipitation and temperature of past storm events to characterize the dynamics of hydrologic and meteorological processes (i.e., snowmelt, soil moisture, runoff). Furthermore, quantifying and understanding the synoptic and mesoscale characteristics of extreme precipitation events is critical in identifying the key drivers of significant floods. In recent years, advances in the academic, private, and public sectors have provided the opportunity to investigate new data sets and new standard-of-practice methods for application in extreme precipitation studies.

Among the most significant meteorological advancements are the analysis, transposition, scaling, and stochastic generation of precipitation events to support watershed precipitation frequency analyses. An overview of the meteorological datasets and techniques used to determine extremely rare precipitation frequency estimates will answer the question: “How are annual exceedance probabilities (AEPs) of precipitation out to 10^-7 computed?” The key is assembling a large enough sample of real and stochastically-generated plausible storm precipitation events to confidently extrapolate out to these AEPs. Using examples from watersheds in the U.S. and Canada, this presentation will feature the assembly of storm data required to derive much-needed, rare AEPs for high-consequence infrastructure.