Uncertainty in the summary relationships is quantified and incorporated into the economic and engineering performance analyses of alternatives. The process applies Monte Carlo simulation, a numerical-analysis procedure that computes the expected value of damage while explicitly accounting for the uncertainty in the basic parameters used to determine flood inundation damage. HEC has developed the FDA computer program to assist in analyzing flood damage reduction plans using these procedures.
Using this approach, each input parameter can be expressed as an empirical or analytical distribution and input parameter values for each Monte Carlo simulation are selected using randomly drawn numbers. For example, as part of the consequences compute, a structure value is drawn from a structure value distribution. During the risk compute, a stage-discharge function is sampled, which means that a stage is drawn from a stage distribution for each flow in the relationship. This process takes place for all input variables that are defined with uncertainty. HEC-FDA relies on a random number generator and the empirical and analytical distributions in the Statistics Engine to power the Monte Carlo. Each object that is sampled is provided its own random number generator seeded separately so that the series of random numbers provided for sampling is independent of the addition or removal of other objects for sampling and so that each object is sampled independently of the other objects.
Computing Expected Annual Damage with Uncertainty is an application of the Monte Carlo simulation.