A 24-hour forecasting of PM₁₀ concentration in urban area

Air pollution is one of the critical issues faced globally due to industrial emissions, traffic emissions, and open burning. Therefore, air quality monitoring is very important in ensuring the level of air pollution would not harm people. PM₁₀ is more dominant in air pollution compared to other pollutants (NO₂, SO₂, CO). Besides that, PM₁₀ has been a challenge to Malaysia’s air quality management. Therefore, the purpose of this study was to develop PM₁₀ concentration forecasting model by using multiple linear regression (MLR) for 24-Hour forecasting. The hourly-data from the 2002-2016 in Klang, Malaysia was used in this study. The model used gaseous (NO₂, SO₂, CO), PM₁₀ and meteorological parameters (temperature, relative humidity and wind speed) as predictors. The performance indicators such as Prediction Accuracy (PA), Coefficient of Determination (R²), Index of Agreement (IA), Normalized Absolute Error (NAE), and Root Mean Square Error (RMSE) had been used to measure the accuracy of the models. Performance indicators indicates that the model was adequate for the next 24-hour prediction (PA=0.780, R²=0.609, IA=0.865, NAE=0.209, and RMSE=21.629). The findings indicated that multiple linear regression method can be used to forecast PM₁₀ concentration for the next 24-hour. The prediction of PM₁₀ concentration will support the relevant local authorities to give an early warning to people who are at risk of acute and chronic health effects resulted from air pollution.