FORECASTING THE CROP YIELD OF A COCONUT ESTATE
Abstract
Seasonal Autoregressive Integrated Moving Average (ARIMA) process of (0,1,2) x (0,1,1) x 6 that best fits a set of crop‑wise coconut yield data, in Bandirippuwa, Lunuwila is identified without using variance stabilization transformation. In this process the present value of the series may be described as a linear function of the past observation of the series and past disturbances. The physical factors such as rainfall, temperature, day length etc. are not required for this method, however the past crop figures in the estate is needed. While such model is useful for short term forecasting, it also gives the upper and lower limits of the forecasts at a given probability. These intervals would provide the quantified information on the degree of duration of the forecasts.
References
2. Abeywardena, V. (1971) "Yield variation in coconut" C.C.Q. 22 97 103.
3. Chatfield, C. (1979) "Inverse autocorrelations" J.R. Statist. Soc. A. 142, 363 372.
4. Godfrey, L. G. (1979) "Texting the adequacy of a time series model" Biometrika 66 67 72.
5. Hamilton, D.C. and Watts, D.G. (1978) "Interpreting partial autocorrelation function of seasonal time series models". Biometrika 65, 135 140.
6. Hull, C.H. and Nil, N.H. (1981) "SPSS update 7 9" Mcgrow Hill Book Company.
7. Makindatics, S. and Wheelwright, S.C. (1978) "Forecasting Methods and application" John Willey and on, New York.
8. Medhi, J. (1982) "Stochaxtic Process" John Willey and Sons, Inc., New York.
9. Newbold, P. (1981) "Some recent developments in Time Series Analysis" International Statistical Review 49, 53 6.
10. Smith, G.W. (1966) "The relation between rainfall, soilwater and yield of copra on a coconut estate in Trinidad" J. Appl. Ecol 3(l) 117 125.
