Urban hydrological modelling in a developing country with limited data

Abstract

Accurate hydrological modelling can be challenging in heterogeneous urban catchments, especially in regions with limited data. SWMM configurations were set up for two urban catchments in South Africa to test the performance of this model for urban areas in developing countries. Despite the relatively small catchment sizes, inconsistencies are evident between observed rainfall and measured runoff for most events. Model scenarios were created with various sets of parameter values, first from literature and then from values physically measured in the catchments. The observed and simulated flow comparisons were improved by incorporating model parameter values derived for this study, although the results were inconsistent. As the region experiences storms with considerable spatial distribution, the flow gauging stations sometimes record runoff which does not correspond to an observed rainfall event with corresponding intensity, and vice versa. Recommendations are made for further improvement of the modelled results using radar data for significant rainfall events and incorporating storage areas throughout the catchments to simulate unintended attenuation.