Annual maximum rainfall prediction using frequency analysis for roorkee, uttarakhand, india

Abstract

Rain is a meagre and important hydrological variable in dryland areas. Due to population growth, economic development and urbanization, the demand for water in these areas is increasing day by day and with all available resources, water management becomes more and more crucial. Therefore, rain probability analysis is needed to address a number of water management issues and to prevent crop damage due to deficits or excess rainfall. Scientific rainfall forecasting and crop planning can prove to be an important tool in the hands of farmers for better economic returns. An attempt was made to analyze the frequency of different rainfall data. Frequency analysis helps in determining the expected rainfall. An important step in analyzing the frequency of rainfall is selecting a suitable distribution to represent the depth of the rainfall to investigate the precipitation. The distributions namely, Gumbel, Log Normal, Log Pearson Type III and Van Te Chow were used for the prediction of the maximum annual rainfall. Annual rainfall data for 20 years (1991-2010) were collected from the National Institute of Hydrology (NIH), Roorkee, India. This research attempts to fit several theoretical probability distributions to the maximum annual rainfall for daily, two consecutive days, three consecutive days; and to select the best probability distribution for the prediction of maximum annual rainfall for daily, two consecutive days, three consecutive days.For determination of goodness of fit chi-square, percentage absolute deviation, integral square errorwere carried out by comparing the expected values with the observed values. The results found showed that the Gumbel distribution emerged to be the best fit for the prediction of annual maximum rainfall values of Roorkee for one, two and three consecutive days. The second best fit was Ven Te Chow distribution for one, two and three consecutive days.