The monthly time series of the Total Suspended Solids (TSS) concentrations in Euphrates River at Nasria was analyzed as a time series. The data used for the analysis was the monthly series during (1977-2000).
The series was tested for nonhomogenity and found to be nonhomogeneous. A significant positive jump was observed after 1988. This nonhomogenity was removed using a method suggested by Yevichevich (7). The homogeneous series was then normalized using Box and Cox (2) transformation. The periodic component of the series was fitted using harmonic analyses, and removed from the series to obtain the dependent stochastic component. This component was then modeled using first order autoregressive model (Markovian chain). The above analysis was conducted using the data for the period (1977-1997), the remaining 3-years (1998-2000) of the observed data was left for the verification of the model. The observed model was used to generate future series. Those series were compared with the observed series using t-test. The comparison indicates the capability of the model to produce acceptable future data.