FLOOD FREQUENCY ANALYSIS IN BAITARANI RIVER BASIN, ODISHA

Abstract

Floods are the most common natural hazards, and they occur when a large amount of water overflows and submerges usually dry land. In coastal areas, floods are often induced by heavy rainfall, rapid snowmelt, or a storm surge from a tsunami. Every year many lives and property is lost in various parts of India e.g. Bihar, coastal regions etc and across the world.The various measures are used to reduce the effects of floods and losses due to it. Floods affected over 2 billion people worldwide between 1998 and 2017. Floods are most harmful to people who live in floodplains or in buildings that aren't flood-resistant, or without having flood warning systems.Flood frequency analysis is used to forecast flow values corresponding to particular return periods or probabilities along a river.Statistical information i.e. mean values, standard deviations, skewness, and recurrence intervals are determined using observed annual peak flow discharge data.These statistical data are then used to create frequency distributions, which are graphs and tables that show the probability of different discharges as a function of recurrence interval or exceedenceprobability.Flood frequency analysis is also useful in flood insurance and flood zoning activities. Flood frequency analysis (FFA) of the Baitarani River is very essential for assessing flood hazards, mitigating them, and preparing and managing the river effectively for all people. Since the river flows through Odisha and Jharkhand State, it is one of the most frequently flooded rivers in India. The aim of this analysis is to estimate the predicted flood in the Anandpur &Champua watersheds of the Baitarani River Basin, using three probability distributions: the Gumbel distribution, the Log Pearson III distribution, and the Weibull distribution. Annual maximum flood time series from 1981 to 2017 (37 years) and 1991 to 2017 (28 years) were obtained from the Central Water Commission at the Anandpur and Champua gauging stations. Three Goodness of fit methods for determining the best model namely the Kolmogorov Smirnov, Anderson Darling, and Chi Squared test are used. The study determines the predicted flood for return periods of 2, 10, 25, 50, 100, and 1000 years. The analysis demonstrates that the Log Pearson-III model is more readily recognized than other models. To illustrate the seasonal discharge pattern, the Mann-Kendall statistical test was used, which reveals a decreasing trend in monsoon discharge but an increasing trend in summer and winter discharge.