Major Flood Occurrence Modelling for Nata, Botswana

Abstract

For an accurate assessment of flood risk, taking into account the relationship between heavy rainfall and storm surge can be crucial. There are several statistical techniques for modeling such severe dependence, but it is unknown how well they function in terms of calculating the probability that infrequent river floods will exceed a given threshold. The Gumbel mixed model, a bivariate extreme value distribution model with Gumbel marginals, is suggested in this research in order to explore the joint probability distribution of connected flood peaks and volumes as well as the joint probability distribution of correlated flood volumes and durations. The joint distributions, conditional probability functions, and appropriate return periods are obtained from the marginal distributions of these random variables. This study examines how the Gumbel distribution equation model and rainfall data set can be utilized to analyze flood frequency and flood extreme ratio of any given spatial domain in order to underline the significance of employing the model in the geo-analysis of diverse environmental phenomena. For the building of water projects in ungauged locations without records of rainfall intensity or climate conditions, it is essential to produce suitable rainfall Intensity-Duration-Frequency (IDF) curves. Hydrological engineering planning, design, and management problems frequently call for a thorough understanding of flood event characteristics, such as flood peak, volume, and duration. Flood frequency analysis often focuses on flood peak values and so provides a limited assessment of flood events. This study looked at a 31 years span of rainfall data and found out that according to the gumbel distribution annual precipitation of 63.28mm is most likely to occurevery year, and annual precipitation of 300.59mm has a 100% chance of occurring every 32 years. It can be deduced from the graphs that there is 3% of  301 mm rainfall occurrence. The lower percentage exceedance probability of 97% shows that there are higher chances of having 63mm. This method is appropriate for estimating discharge while designing flood control structures.