Abstract
Stochastic Modeling for The Daily Extreme Flows of Kelkit Stream
The aim of this study is to determine whether the daily extreme flows for Kelkit Stream could be forecast as a stochastic model. For this aim, the autoregressive models (the first and second order Markov models) and Arima(l.O.l) model (mixed autoregressive-moving average model) were used. The flows forecasted by using the models mentioned were compared to the observed flows. The results showed that the flow predictions based on the first order Markov model are fitted to the data better than the other models.
Keywords
Extreme flows autocorrelation coefficient Markov model Arima model