Development of synthetic unit hydrograph using probability models

Use of probability distribution functions (pdfs) in deriving synthetic unit hydrographs (SUHs) has received much attention because of their similarity with unit hydrograph properties. In this study, the potential of four pdfs, i.e., two-parameter Gamma, three-parameter Beta, two-parameter Weibull and one-parameter Chi-square distribution to derive SUH are compared with traditional synthetic methods, i.e., SCS method and Snyder method using field and simulated data in Lighvan catchment, Iran. The distribution parameters are computed using numerical methods by programming in Mathematica. Based on visual comparison as well as statistical measures, such as square root of error (STDER), root mean square error (RMSE), coefficient of determination (R2), and weighted error, it was found that the Gamma and Weibull distributions yield SUHs precisely using the field and simulated data, respectively. The Gamma, Beta and Weibull distributions were more flexible in preserving SUHs shape than the Chi-square. All of the pdfs outperformed the traditional methods.