A thesis submitted in partial fulfillment of the requirement for the degree of Master of Engineering, school of Engineering and Technology

Thesis (M.Eng.) - Asian Institute of Technology, 1982

The present thesis work deals with the probability distribution function of flood volume. The definition of flood volume and the procedure used to obtain it have been presented. Six probability distribution functions which are commonly used in t he frequency analysis of flood peaks namely: exponential, gamma, lognormal , Gumbel , Pearson Type III and Log Pearson Type III distributions are compared for their goodness of fit to the observed frequency of flood volume . The maximum likelihood method is used for estimating the distribution parameters from the observed flood volume, except for the Pearson Type III and Log Pearson Type III distributions whose parameters are estimated by the moment method. Goodness of fit is evaluated by using chi- square and Kolmogorov-Smirnov statistics. The result drawn from the 7 selected streamflow gaging stations in Thailand, and for various considered threshold discharges is that t he two parameter gamma distribution seems to be the most suitable for describing flood volume in Thailand. Finally, the application of the concept of the distribution of flood volume to Sirikit reservoir operation is also presented.