Analysis of the flood patterns using hydrologic and hydraulic models in river Kafu catchment :

Abstract

Floods are the most frequent type of natural disaster occurring all over the world. This research project shows that River Kafu floods almost every year with the most severe event in October 2021. Those floods caused death, destruction of property and displacement of over 10,000 people. Flood forecasting, Flood hazard mapping and flood risk zoning are quite effective non-structural procedures in managing floods that lower the risks floods may cause. The study was conducted to build a hydrological and hydraulic model in combination with Geographical Information System (GIS) for flood inundation mapping in Kafu Catchment basin. HEC-HMS and HEC-RAS models were employed to achieve the overall specified objective. Time series analysis of hydrological data has been done using Log Pearson type III frequency analysis method to obtain the rainfall intensities for different return periods such as 50yr, 100yr, 200yr and 500yr, Results of the rainfall frequency analysis which were the point depths of 52.78mm, 58.97mm, 65.09mm and 73.30mm for the respective return periods were used as an input for the hypothetical storm method in the hydrological model to generate the corresponding catchment discharges which was then used in the hydraulic model to develop the flood inundation maps, Water surface elevations and flow velocities in HEC-RAS. A sensitivity analysis of the hydrological model parameters used in this study showed that the model out puts are more sensitive to curve number compared to other parameters (percentage of imperviousness and lag time). The HEC HMS model was calibrated and validated using both manually and automatically by the Nash- Sutcliffe Efficiency (NSE) obtained was satisfactory. From the results of the hydraulic model, it was observed that, the flood plain inundated areas increase with the magnitude of flow within the modelled network indicating a high flood risk level for activities and settlements adjacent to the river banks. The simulated peak discharges of 46m3/s, 113.3m3/s, 202.6m3/s, and 345.7m3/s for the respective 50-, 100-, 200-, and 500-year return periods produced maximum channel flood depths of 2.497m, 2.887m, 3.429m, and 3.698m respectively as observed from the gauging station at Namyekudo, Masindi. Flood hazard maps were generated for the return periods in the Ras Mapper environment.