Rivers

Flood forecasting models for rivers can range from simple empirical relationships to complex integrated catchment models. Forecasts may be based simply on river level or flow observations at locations upstream of the site of interest, or use observations and possibly forecasts of rainfall to gain additional lead time. This chapter begins with a discussion of the main factors which can influence the design of a river flood forecasting model, including the forecasting requirement and the availability of real time data, and then describes two main categories of model; rainfall runoff models, and river channel (flow routing) models. Examples are provided for a range of process-based and conceptual modelling techniques, and for data-based approaches such as transfer function and artificial neural network models.