Abstract
Floods are a widespread natural hazard globally and significantly negatively impact the public and the environment. Deforestation, uncontrolled urbanization, and global warming are the leading causes of floods triggered by heavy rain that develops over a short period. In reducing flood destruction, it is crucial to recognize the reasons for significant floods, their intensity, and their frequency. Estimating flood intensity is a critical element for preparation, designing, and overseeing water resources projects, and assessing the extent of likely future flooding is an essential task in hydrology. Most water projects and flood management responsibilities necessitate accurate estimates of flood designs. For this logic, design flood assessment remains an area of great interest in flood hydrology and is a subject of research worldwide. Flood frequency analysis (FFA) provides a practical method for determining a robust probability distribution consistent with streamflow data at a location of interest.
This study aims to analyze the occurrence and magnitude of floods in the Mngeni River Basin at 11 different gauging stations, using annual peak or maximum flow data, which can also be an essential input for flood management in the region. Analyzed the annual highest discharge of different river stations within the Mngeni River Basin and compared the statistical methods (Log-normal, Normal, Gumbel Max, and Log Pearson type III distributions) against the recorded hydrological data. Determine the recurrence intervals of significant floods within the Basin, find the best statistical distribution method for the Mngeni basin, and evaluate whether flood values have changed for the same return periods.
Peak flow in the Mngeni River Basin is estimated to be 1300 cubic meters per second. Based on the flood frequency analysis, all return periods indicated increasing flows with higher likelihoods of exceedance. The best-fit distribution was determined using three goodness-of-fit tests—Kolmogorov–Smirnov, Anderson–Darling, and chi-squared—using the EasyFit program. In the Mngeni River basin, LP3 is the optimal preferred Probability Distribution (PD), with the Log-Normal distribution coming in second. But under some situations, LP3 isn't the Mngeni River Basin's best-fit PD. To design new engineering structures in the Mngeni River Basin, peak flow values of Log-Pearson type III (LP3) would be helpful. Designing systems in or adjacent to the river that may be flooded and protecting against the largest expected event is critical.