Prediction of Monthly Fluoride Content in Tigris River using SARIMA Model in R Software

  • Awatif Soaded Alsaqqar, Ass. Prof. Dr. College of Engineering-University of Baghdad
  • Basim Hussein Khudair, Ass. Prof. Dr. College of Engineering-University of Baghdad
  • Rasha Attwan Faraj, Ass. Lect. College of Engineering-University of Baghdad
Keywords: water quality management, time series analysis and prediction, SARIMA model, R software.

Abstract

The need to create the optimal water quality management process has motivated researchers to pursue prediction modeling development. One of the widely important forecasting models is the sessional autoregressive integrated moving average (SARIMA) model. In the present study, a SARIMA model was developed in R software to fit a time series data of monthly fluoride content collected from six stations on Tigris River for the period from 2004 to 2014. The adequate SARIMA model that has the least Akaike's information criterion (AIC) and mean squared error (MSE) was found to be SARIMA (2, 0, 0) (0,1,1). The model parameters were identified and diagnosed to derive the forecasting equations at each selected location. The correlation coefficient between the actual and predicted values for fluoride concentration at the six locations, Al-Karakh, East Tigris, Al-Wathbah, AL-Karamah, Al-Rashid and Al-Wahda WTP intakes, was 0.93, 0.82, 0.86, 0.90, 0.83 and 0.89, respectively. Model verification results indicated that the model forecasting outputs rationally estimated the actual monthly fluoride content in the selected locations.

 

Downloads

Download data is not yet available.
Published
2016-08-01
How to Cite
Alsaqqar, A., Khudair, B. and Faraj, R. (2016) “Prediction of Monthly Fluoride Content in Tigris River using SARIMA Model in R Software”, Journal of Engineering, 22(8), pp. 75-85. Available at: http://joe.uobaghdad.edu.iq/index.php/main/article/view/156 (Accessed: 18September2020).

Most read articles by the same author(s)