Frequency analysis of rainfall characteristics in R software platform

Having the values of rainfall characteristics (intensity, depth, duration) in different return periods is highly important in many of the projects related to water resource management, drought studies, flood management, and design of water structures. These parameters are also the main inputs in hydrological models. In this study, different available software for frequency analysis and the properties of several related packages presented in the R programing platform were reviewed and the best distribution functions fitted on each rainfall characteristic was determined using the L-moment method through lmomco package. Further, quantile values of each variable were determined for different return periods based on the selected distributions. For this purpose, 40-years rainfall data recorded in Minoodasht station in Chehelchay watershed located in Golestan Province was used. The results showed the most appropriate distribution fitted on intensity and depth of the rainfall to be the Wakeby distribution and the Weibull distribution for duration. Furthermore, the quantile values were computed for each variable in return periods of 2 to 1000 years. The results indicated the application of L-moment through lmomco package in the R software platform as a very useful tool for achieving the best fitted distributions on the sample data, showing high efficiency in determination of the governing distribution on each variable and its quantile value for a given return period.