Using Box-Jenkins (ARIMA) Models to Forecast Syrian Olive Oil Production and Estimate the Losses Resulting from the Climate Changes and Syrian Crisis

Wael Habib (1) Fayez Al-Mikdad(1) and Mohammad Ghoush(2)    

(1). Scientific Agricultural Research Center in Latakia, (GCSAR). Damascus, Syria.

(2). Agro-Economy, General Commission of Scientific Agricultural Researches, Syria.

(3). Agro-Economy, Tishreen University, Syria.

(*Corresponding author: Dr. Wael Habib. E-Mail:wael.ha76@gmail.com).

Received: 06/08/2020                               Accepted: 01/09/2020

Abstract

The objective of the research is to analyze the time series of olive oil production in Syria (1961-2018) and defining its components and characteristics in order to determine the appropriate predictive mode. In addition to estimate, the predictive value of the supposed production during the period of the Syrian crisis (2011-2018) thus estimating losses as a result of this crisis. The results showed that the time series of olive oil in Syria is not stationary. It characterized by three trends of development, the first is a general trend that tends to rise, and the second is a cyclical trend, ie, biennial bearing (BB). The third is a random trend resulting from abnormal climatic changes or security disturbances. The timing of biennial bearing Phenomena id difficult to predict. It was found that, the BB was responsible for 21.7% of the increase in production in the years of positive /bearing production (the increase above the level of the general production of the series). In contrast, this factor is responsible for 18.9% of the decline in production in the years of negative/nonbearing production. The best model for predicting olive oil production in Syria was ARIMA (3.1.1), but the parameter of the lagYt-2 was insignificant. The self-regression parameters (AR) declare that the behavior of this time series is often determined by its values in the first and third recent years. While the moving average parameters (MA) indicates that, the behavior of the time series (Yt)is often determined in terms of the current random noise and the previous random noise. According to estimates of this model, the total losses of olive oil production in Syria as a result of the Syrian crisis amounted to (277.3) thousands tons. These losses are mostly due to the security unrest and the abnormal climate changes that accompanied the crisis years as well.

Keyword: Olive Production in Syria, Box-Jenkins Models, Time Series, Biennial/Alternative   Bearing, Cyclical Trends.

Full paper in Arabic: PDF