Calibration and verification of a hydrological SWMM model for the ungauged Kinyerezi River catchment in Dar es Salaam, Tanzania

Floods are one of the most devastating natural disasters in urban areas, causing infrastructure damage and loss of lives. Flash flood prediction in ungauged catchments remains a challenge due to the unavailability of historically recorded river flow data. Long-term and extreme discharge records of rivers are essential for planning flood mitigation and the safety of infrastructures. However, the lack of recorded river flow data in Dar es Salaam city has hindered the application of rainfall-runoff models. This study aims to calibrate and verify the EPA SWMM 5.2 model for runoff prediction of ungauged Kinyerezi River using the Msimbazi River flow data. The sub-catchment geospatial characteristics were delineated using the Digital Elevation Model of the study area in ArcGIS 10.4. The input parameters of the model were sub-catchment characteristics, rainfall data, and infiltration data (in form of Curve Numbers). The historical recorded water levels of the Msimbazi River were converted to flows using Manning's equation. The model performance was evaluated using the coefficient of determination (R²), Nash–Sutcliffe efficiency (NSE), and index of agreement (d). The results showed that during calibration, R², NSE, and d were 0.664, 0.63, and 0.877, respectively, while during verification, were 0.665, 0.648, and 0.897, respectively. The study concluded that the model was effective in predicting runoff discharges of the ungauged Kinyerezi as well as Msimbazi Rivers. The model will be helpful in assessing the adequacy of drainage systems, generating extreme discharge records, and evaluating the performance of Low Impact Development (LID) practices in mitigating runoff floods.