Application of the Monte Carlo method to estimate warning times to critical levels in a high consequence dam

The Monte Carlo method of design flood estimation is widely accepted as being a rigorous approach to modelling flood frequency of dams, as it can include representative distributions of storm, catchment, and storage characteristics. These include temporal patterns, antecedent catchment conditions, and starting lake storage level. The Monte Carlo method also allows for analysis of other model variables, such as hydrograph timing. Flood frequency of the outflow peak discharges is not the only variable of interest when modelling design floods dams and reservoirs. Also of importance is the time of the storage to rise from a series of warning levels in order to assist with planning for downstream flowing. A catchment and storage model for the dam has been developed, and adapted to Monte Carlo analysis. The suitability of this approach to determining warning times has been assessed. Tens of thousands of design event simulations of the model were executed, with a number of variables recorded for each simulation including peak inflow, peak outflow, peak storage level, and time to reach a given target level from a range of start trigger levels. An attempt was made to fit a statistical distribution to scenario rise times, with mixed results. The scenario rise times were used as qualitative information to guide further investigations, and could also be used for sensitivity analysis to determine parameters critical to short warning times.