The prediction of extreme runoff events has significant risk and financial implications when dealing with hydraulic infrastructure. This is especially true in highly urbanized areas such as Southern California. Two methods for determining extreme runoff exist: extrapolation of existing runoff data using extreme event probability distributions, or hydrologic modeling using design rainfall events and watershed characteristics to generate an estimate of the extreme runoff event.

This research investigates both methods to evaluate usefulness and limitations in providing guidance for risk and financial analysis. Design of levees and flood protection channels often focuses on providing protection from events with a 50- to 100-year recurrence interval. In many areas of the country, stream gage record sets do not contain records of this length. The effects of record length, probability distribution selection, and the method of parameter estimation are evaluated to determine the impacts on prediction of the extreme runoff event used for levee and channel design.

Design storms are often used with hydrologic models to predict runoff for events larger than those measured through systematic stream gaging. The Probable Maximum Precipitation (PMP) - Probable Maximum Flood (PMF) methodology is widely used. This research evaluates the use of this standard on design for major hydraulic structures such as dam spillways within Southern California. Two standard PMP methodologies are evaluated based on rain gage frequency analysis within Los Angeles County. The effects of soils, watershed characteristics, and wild fire on extreme runoff events are also evaluated using Monte Carlo Simulation of 27 watersheds within the County.

The Monte Carlo Simulations evaluates two design storms, two soil loss methodologies, and the effects of fire within a watershed. The structure and development of the model will be discussed, as well as the results for the different cases in determining extreme runoff events. Conclusions will be drawn regarding prediction of extreme runoff events in Southern California.