Augmentation of PM measurements based on machine learning model and environmental factors.

PM, particulate matter with an aerodynamic diameter smaller than 1.0 µm, can adversely affect human health. However, fewer stations are capable of measuring PM concentrations than PM and PM concentrations in real time (i.e., only 9 locations for PM vs. 623 locations for PM or PM) in South Korea, making it impossible to conduct a nationwide health risk analysis of PM. Thus, this study aimed to develop a PM prediction model using a random forest algorithm based on PM data from the nine measurement stations and various environmental input factors. Cross validation, in which the model was trained in eight stations and tested in the remaining station, achieved an average R of 0.913. The high R value achieved under mutually exclusive training and test locations in the cross validation can be ascribed to the fact that all the locations had similar relationships between PM and the input factors, which were captured by our model. Moreover, results of feature importance analysis showed that PM and PM concentrations were the two most important input features in predicting PM concentration. Finally, the model was used to estimate the PM concentrations in 623 locations, where input factors such as PM and PM can be obtained. Based on the augmented profile, we identified Seoul and Ansan to be PM concentration hotspots. These regions are large cities or the center of anthropogenic and industrial activities. The proposed model and the augmented PM profiles can be used for large epidemiological studies to understand the health impacts of PM.