A stochastic Model for Predicting Irrigation Water Requirements (IWR)

Abstract

The main objective of this paper is to develop a stochastic time series model with trend, periodic and irregular components using a ten years IWR decade data for three different types of cotton crops cultivated in Gezira Scheme, SUDAN. The model was applied to cotton Brackat and then used to Shmbat & Akala cotton. In the analysis of IWR time series the correlogram technique was used to detect the periodicity which then smoothed by Fourier series method. The series is then tested for stationary and the dependent part of irregular component is found to be well expressed by the first order autoregressive model for all the crops. The developed model superimposes a periodic-deterministic process and an irregular component