Time Series Analysis for Prediction of PM2.5 Using Seasonal Autoregressive Integrated Moving Average (SARIMA) Model on Taiwan Air Quality Monitoring Network Data

Abstract

Nowadays, in Taiwan, due to the increasing number of vehicles, industrialization of large energy consumption, uncontrolled constructions and urbanization, air pollution is becoming a major problem. Hence, it is necessary to control air pollution by applying air quality monitoring actions. The particulate matter (PM2.5) of the air pollution in TAQMN data is the main pollutant accountable for at least two-thirds of the severely polluted days in the major cities of Taiwan. In this work, machine learning (ML) techniques are widely used in developing models that can be used to control the air pollution. Seasonal Autoregressive Integrated Moving Average (SARIMA) model is used to predict the air pollution concentration, where the dataset chronologically from 2012 to 2016 are used to train the proposed method and testing data set from 2016 to 2017. The result of the SARIMA model shows high precision in forecasting the future values of particulate matter (P2.5) level with minimum error.