Long-Term Stochastic Modeling of Monthly Streamflow in River Nile

Synthetic time series created from historical streamflow data are thought of as substitute events with a similar likelihood recurrence to the real event. This technique has potential greatly reduce uncertainty surrounding measured streamflow. The goal this study is create synthetic model using combination Markov chain and Fourier transform techniques based on long-term for Nile River. First, chain’s auto-regression applied, in which data’s trend seasonality discovered eliminated before applying Pearson III distribution function. function substituted by discrete (DFT) second approach. applicability two simulate between 1900 1999 evaluated. ability generated maintain four most important statistical properties samples monthly flows, i.e., mean, standard deviation, autocorrelation lag coefficient, cumulative distribution, was used assess quality series. results reveal that techniques, small differences accuracy, reflect variation well terms three mentioned parameters. According coefficient determination (R2) normalized root mean square error (NRMSE) statistics, approach somewhat superior simulating predicted discharge.