Flood estimation at Hathnikund Barrage, River Yamuna, India using the Peak-Over-Threshold method

Abstract

The paper presents a methodology for the estimation of flood magnitude using the peaks-over-threshold (POT) data. A semi-automatic approach for the determination of POT values has been presented. Two frequency analysis methods, namely Gumbel’s method and Log-Pearson type III distribution, have been used. Two major issues in implementing a POT approach are: (i) determining an appropriate threshold and (ii) ensuring the independence of events. The present paper addresses these issues simultaneously. Using the daily discharge data, various return-level flood magnitudes have been estimated. A comparison of flood magnitudes obtained using the annual maxima (AM) approach and POT approach clearly indicates that the flood magnitudes in general are higher when Log-Pearson method is used as compared to when Gumbel’s method is used. The result of the research conducted clearly indicates that the flood magnitudes estimated on the basis of POT sample are more reliable, particularly when the data length is short.