The parameter estimation of statistical distributions is important for regional frequency analysis (RFA). The accuracy of different parts of RFA such as estimating the regional quantiles of the selected statistical distribution, determining the heterogeneity measure, and choosing the best distribution based on the Monte Carlo simulation, may be influenced by using the different values of regional parameters. To fulfill this aim, in the present study, a new model is developed for regional drought frequency analysis. This model utilizes the L-moments approach and the adjusted charged system search as an advanced meta-heuristic algorithm, in which some modifications on the equations of the algorithm are performed to improve its standard variant. The verification of the regional parameters estimated by the new methodology yields accurate results compared to other models. Furthermore, this study illustrates the usefulness of the robust discordancy measure against the classic one. For this purpose, different values of the subset factors (α) are utilized in the robust discordancy measure, and finally, the best value of subset factor is found equal to 0.8, which can accurately recognize discordant sites within the region.

You do not currently have access to this content.