Portal de Periódicos da CAPES

Analyzing And Forecasting The Ten-daysNatural Flow Of The Blue Nile Using ThePeriodogram Technique

1970; Volume: 12; Linguagem: Inglês

10.2495/cmem950481

1746-4064

Abdelazim M. Negm, Abdulaziz S. Al‐Turbak, A.A. Alhamid, Ashraf Elsayed,

Stock Market Forecasting Methods

The periodogram or the spectrum analysis is a tool to investigate both the pattern and the hidden periodicities of the time series data. In this study, the continuous spectrum is applied to the series of the Blue Nile ten-days natural flow during the flood season at three main stations (Khartoum, Sennar, and Roseires). Also, the analysis includes the ten-days annual flow series. It was found that the periodicities of the Blue Nile natural flow are of complex nature. The pattern of the times series shows two main components viz. (periodicity and stochasticity). The tendency components appeared only in the month of July. The contribution of the stochasticity component to the total explained variance is dominant. Based on the obtained periodicities, attempts are made to make a long-term forecast for ten-days natural flow of the Blue Nile, but it was found that the obtained forecast has some gaps at some periods. It is concluded that a better forecast may be obtained if each component is forecasted separately by the proper forecasting technique and then superimposed to obtain more accurate and logical forecast. Accordingly, an algorithm for forecasting the ten-days natural flow of the Blue Nile is presented.