Flood Peak Estimation Using Rainfall Run off Models

Abstract

The design of hydraulic structures such as road culverts, road bridges and dam spillways requires the determination of the design food peak. Two approaches are available in the determination of the design flood peak and these are: flood frequency analysis and rainfall runoff models. Flood frequency analysis requires a selection of an appropriate probability distribution function and the availability of a long record of instantaneous flows and therefore, is not appropriate for the estimation of a design flood peak for catchments with no flow measurements. Rainfall runoff models on the other hand do not require the availability of stream flow data for the estimation of the design flood magnitude. They only require rainfall data and the physical characteristics of the catchment as input variables. Five rainfall runoff models (rational formula, synthetic unit hydrograph, TRRL East African Model, HEC-1 Model and MULTSED Model) have been used for the estimation of the 50 year design flood peak for the Sinza river in Tanzania. Simulated design flood peak values (50-yr flood) range from 37.6 to 44.34 m/s and have a standard deviation of 2.4m/s and therefore, are regarded as satisfactory.

UNISWA Research Journal of Agriculture, Science and Technology Vol 3 (1) 1999: pp 13-20