Simulation of Flood Characteristics of Ajichay River Using a Multivariate Structure

In this study, the first flood time series of vanyar station on Ajichay rivr was separated from base flow using local minimum method, 34 flood events were distinguished and flood event characteristics were extracted and then dependence structure between the characteristics of the flood event (peak flow, peak time, the total volume and base time of flood) diagnosed using the D-vine structure and then mentioned characteristics were simulated using the most accurate D-vine structure. To get the best D-Vine structure first structures were constructed through permuting the features of flood and then different Elliptical and Archimedean copula families were fitted on each pair-copula and the most accurate copulas were selected according to Akaike and Bayesian information criteria for each of pair-copula. In this way, the best combination of copulas was obtained for each of D-vine structures and the best structure was selected to simulate feature of flood and ultimately to evaluate the effectiveness of selected structure in the simulating of flood characteristics, the main statistics of simulated flood features were calculated and results showed that the mean and standard deviation of all simulated properties of flood were maintained well. However, the minimum, maximum and skewness of simulated features of flood in some features had been maintained well and also were preserved relatively well in some other features. Finally, the results showed that for obtaining the best D-vine structure should be performed all permuting of variables and also mentioned structure is capable in accurate simulation of flood characteristics even in short time series.