Flood Frequency Analysis of Annual Maximum Stream Flows for Luvuvhu River Catchment, Limpopo Province, South Africa

Abstract

Luvuvhu River Catchment (LRC) is one of the regions in South Africa where floods have caused enormous damage and impacted negatively on fauna and flora. Over the years, the catchment has experienced floods resulting from heavy rainfall associated with the ITCZ. Annual maximum flow data from 8 stations with 50 years hydrological data were used to analyze flood frequencies in the catchment. To derive the probability of occurrence of flood events, the frequency distributions which could best describe the past characteristics and magnitudes of such floods were tested. This involved the determination of the best flood frequency models, which could be fitted to the available historical recorded data. The distribution models used included the Generalized Extreme Value, Gumbel or Extreme Value type 1, Log-Normal and the Log Pearson type III distributions. The extreme value analysis showed that the Gumbel and Log Pearson type III distributions provided the best fit.