River flow forecasting constitutes one of the most important applications in hydrology.
Several methods have been developed for this purpose and one of the most famous techniques
is the Auto regressive moving average (ARMA) model. In the research reported here, the
goal was to minimize the error for a specific season of the year as well as for the complete series. Goal programming (GP) was used to estimate the ARMA model parameters. Shaloo Bridge
station on the Karun River with 68 years of observed stream flow data was selected to evaluate
the performance of the proposed method. The results when compared with the usual method of
maximum likelihood estimation were favorable with respect to the new proposed algorithm.