Short Term Prediction of PM10 Concentrations Using Seasonal Time Series Analysis

• Main Authors: Hamid, HA (Author), Ramli, NA (Author), Ul-Saufie, AZ (Author), Yahaya, AS (Author), Yasin, MN (Author)
• Format: Article
• Language: English
• Published: 2016
• Subjects: AIR-POLLUTION
• Online Access: View Fulltext in Publisher

 Air pollution modelling is one of an important tool that usually used to make short term and long term prediction. Since air pollution gives a big impact especially to human health, prediction of air pollutants concentration is needed to help the local authorities to give an early warning to people who are in risk of acute and chronic health effects from air pollution. Finding the best time series model would allow prediction to be made accurately. This research was carried out to find the best time series model to predict the PM10 concentrations in Nilai, Negeri Sembilan, Malaysia. By considering two seasons which is wet season (north east monsoon) and dry season (south west monsoon), seasonal autoregressive integrated moving average model were used to find the most suitable model to predict the PM10 concentrations in Nilai, Negeri Sembilan by using three error measures. Based on AIC statistics, results show that ARIMA (1, 1, 1) x (1, 0, 0)(12) is the most suitable model to predict PM10 concentrations in Nilai, Negeri Sembilan.