Historical rainfall for urban storm drainage design

Abstract

Traditional design method for urban drainage systems is based on design storms and its major drawback is that frequencies of peak flows in the system are considered equal to frequencies of design storms. An alternative is to use historical storms with rainfall-runoff models to produce a series of possible flows in the system and their frequencies. The latter approach involves more computations and can be laborious for larger catchments. This paper considers ways to reduce the set of historical storms to be involved in design procedure and yet to lead to realistic flow frequencies. Frequencies obtained by rainfall-runoff simulation at an experimental catchment are compared with frequencies of observed peak flows in the system.