Stochastic models for the maximum and minimum daily flows of Tigris and Khabur rivers

Abstract EN

In this paper, time series for the available recorded data of Tigris and Khabur Rivers at Tusan and Zakho stations were analyzed, stochastic models were investigated for both stations using several time intervals (maximum and minimum daily flow in a month ), normality was tested using Kolmogrov-Smirnov approach, series transformation to normal distribution was carried out using box-cox method , split-sample method and kendall's correlation tests were used to check and remove the jump and trend components from the series of the two rivers, and the periodicity component also was detected.

stochastic models, SARIMA (Seasonal autoregressive integrated moving average) and matals (multisite), were used for forecasting and generating the time series for the above two mentioned rivers.

autocorrelation function (ACF) and partial Autocorrelation Function (PACF) were used to identify the suitable SARIMA model.

the akaike information Criteria (AIC) formula was used to find the suitable model for the series with the time intervals as described previously.

SARIMA models of order SARIMA(1,0,0)(0,1,1)12 and SARIMA(1,0,0)(0,1,1)12 were found to be suitable for maximum and minimum daily flow in a month time series for Zakho station respectively, SARIMA(1,1,1)(0,1,1)12 and SARIMA(1,0,0)(0,1,1)12 were found to be suitable for maximum and minimum daily flow in a month time series for Tusan station respectively .Matalas model of first order AR(1) was applied for the time series of both stations.