Watershed Regionalization For Regional Flood Frequency Analysis In The Rift Valley Lake Basin Ethiopia

Africa Center Of Excellence For Water Management Project Topics

Get the Complete Project Materials Now! »

The use of regional information to predict magnitude of flow both at site and ungauged catchments are useful for planning and management of water resources. The main objective of this study was to regionalize watersheds in the Rift Valley Lake Basin (RVLB) and flood frequency analysis for the delineated homogeneous regions. In regionalization of the watersheds; Physiographic, drainage, meteorological, soil cover, land-use pattern characteristics and geographical location attributes have been used. Cluster analysis was done by Hierarchical clustering to obtain number of clusters, and final clustering by K-mean method. Accordingly four regions have been identified and checked using homogeneity tests. Using goodness of fit tests (Chi-square test, Kolmogorov–Smirnov, and Anderson–Darling), the best fit distribution models have been selected. Generalized extreme value distribution is the best fit for region I, Log-normal (2P) is selected for region II, Wakeby distribution is found to be the best for region III, and Generalized pareto is chosen for region IV. For the selected distributions efficient parameter estimation technique was selected by performing standard error analysis. Thus, method of moment (MOM) is the one with the lowest error so, selected for region I, and maximum likelihood (ML) method is found the most efficient method for the regions II to IV. For each region unique regional frequency curve is developed with standardized annual maximum flow series (AM), which is a crucial to estimate flood quantile in ungauged areas of the basin. Regional regression model was developed for all region except for region I which consists only one gauged catchment based on their R2 values. Accordingly 0.82, 0.83, and 0.79 of R2 values respectively for all the three regions. For checking performance of the model, validation of regional model was carried out by computing the relative errors, over five gauged watersheds that is representative for each region considering as pseudo ungauged. The relative errors between observed and estimated mean annual maximum flows resulted all regional model performs good having maximum of 10.6% of relative error. So, for any current and future water resources developments in the area, the developed regional model can be applied.

Get Full Work

Report copyright infringement or plagiarism

Be the First to Share On Social



1GB data
1GB data

RELATED TOPICS

1GB data
1GB data
Watershed Regionalization For Regional Flood Frequency Analysis In The Rift Valley Lake Basin Ethiopia

98