The effect of dam reservoirs on the statistical properties of inlows

The change in the climate caused by dam arrangements and natural transformations will evidently result in a change in the flow characteristics. A time series could be developed employing a cross-sectional measurement of the current flow of a river, this being a random variable. With the help of studies involving time series modeling techniques, lineer stochastic models could be developed with data large enough for the modeling before the construction of a dam reservoir. The modeling is applicable only if the current values after the construction of the reservoir are measured detailed enough. This study involves the change of the models made before and after the construction of dam. Climatic changes in the area and the effect of the human activities such as dam construction on the change of regional hydrological regime are analyzed in this study along with the modeling.The present study employs the average monthly current flow between 1964 and 2005 recorded in the Sinanhoca Station which is situated on Manavgat River in the Middle Mediterranean basin. The data after 1983 represent data recorded after the construction of Oymapınar Dam. Having formed modeling before and after the construction of the dam respectively, the study first revealed that the average annualFlowrate was in line with normal distribution; and then it was tested via Probability Plot Correlation Coefficient Test (PPCC). AR(1), MA(1), AR(2), ARMA(1,1) models were employed for both situations for annual values. However, Akaike Information Criterion (AIC) was used in order to identify the most suitable model that best represents the process, and the AR(1) model was chosen for both situations. Consequently, after conducting the Portmanteau test, in which the independency of the independent random component was identified, the study revealed that the AR(1) model passed the test for both situations.