Summary: | Ambient PM10 (i.e particulate diameter less than 10 um in size) pollution has negative impacts on human health and it is influenced by meteorological conditions. Although the correlation between meteorological parameters and PM10 concentrations is significant in most cases, the linear relationship between them implies that the fraction of the variance, R2 rarely exceeds 25%. However, considering the previous day's concentration of pollutants to the multi-linear regression enhances the model performance and increases the value of R2. Alternatively, artificial neural networks (ANN) are used to capture the complex relationships among many factors considered which present a better prediction. Thus, this study presents the results of predicting ambient PM10 concentration and the influence of meteorological parameters based on the data sampled from 2008-2010 in an industrial area of Pasir Gudang, Johor
|