Alternative well calibrated rainfall-runoff model: Genetic programming scheme

Abstract

Genetic Programming (GP) has been explored as a flow forecasting tool. A catchment in Singapore with a drainage area of about 6 km2 is used for this case study. GP was trained to simulate runoff from a conceptual rainfall-runoff model, Storm Water Management Model (SWMM), which was first calibrated using Shuffled Complex Evolution (SCE) algorithm. Four storms of different intensities and durations are used for training and verification of the GP models. The results show that the runoff prediction accuracy of genetic programming based tool, measured in terms of root mean square error and correlation coefficient, is reasonably high. Thus, GP coupled with a robust optimization scheme such as SCE is a viable complementary tool to traditional conceptual rainfall-runoff models.