Estimation of the flood flow in the absence of at-site recorded data generally requires either a regional method (Project 5) or a rainfall based approach. Rainfall based approaches are based on using a catchment modelling system which may be conceptually complex or simple to predict the response of a catchment to a storm burst (or event) or to a sequence of storm events.

At the time of the publication of the current edition of Australian Rainfall and Runoff, the recommended approach implicitly assumed that the frequency of the rainfall was translated into the desired frequency of flood flow through use of median values of other input to the catchment modelling system. This implicit assumption has been tested in the period since publication of the current edition of Australian Rainfall and Runoff and a number of alternative approaches have been proposed as a means of circumventing the need for this implicit assumption. These methods include `Monte Carlo¿ approaches and `continuous simulation¿ approaches. The reliability of these approaches and their associated uncertainty in prediction need to be defined for different types of problems so that suitable guidance can be presented in the revision to Australian Rainfall and Runoff.