Design flood synthesis by excess rain routing

Trials of several routing procedures for hydrograph syn­ thesis revealed that routing of the excess rain hyetograph through a single linear reservoir-type storage could yield a plausible direct runoff hydrograph provided the shape and drainage pattern of the catchment was not abnormal. The method requires evaluation of only one parameter, the routing constant K, to define the hydrograph for an excess rain of given depth and duration. Kvalues were determined for 92 South African catchments by seeking best-fits be­ tween routed and observedunitgraphs. To aid estimation of Kfor ungauged catchments the country was subdivided into veld-type zones for each of which a relationship between K and area of catchment is provided. A worked example de­ monstrates the method. 1 of depth-area-duration-frequency analyses for a large number of major storms. Secondly, with the aid of a plot based on a large number of observed storm losse$!, the synthetic design storm is converted to a hyetograph of excess rain and, thirdly, the design hyetograph is transformed to a design flood hydrograph by unit­ graph technique with the aid of regionalized dimensionless unitgraphsl, derived from analyses of observed hydrographs in 92 South African catchments. The procedures are summarized and illustrated by examples in a design flood manual 2 issued by the Hydrological Research Unit for the guidance of the practis­ ing engineer. This paper offers a quick method of performing the third step, transformation of excess rain to design flood hydrograph, by a sifT)ple routing procedure. During a recent study it was found that the flood hydrograph associated with given uniformly distr(buted excess rain could seemingly be approximated with fair' accuracy by routing the . excess rain in finite increments through a single linear reservoir. As there are in South Africa few adequately instrumented cat:-.;,­ ments in which both direct runoff and the duration and distribu­ tion of the causative excess' rain have been measured there is as yet no direct means of testing the findin8s. Resort was therefore had to the work of Pullen 3 who used the S-hydrograph technique to establish the durations of excess rain associated with several major floods in 92 South African catch­ ments and to derive synthetic dimensionless unitgraphs. These could be dimensionalized from catchment characteristics and regionalized lag factors. This paper summarizes the results of a study· in which the Pullen unitgraphs were used to determine the routing parameter K and to some extent to test the supposition that the direct run­ off flood hydrograph could be synthesized with reasonable accur­ acy by routing the excess rain through a single linear storage.