A Study of Particulate Matter Area Division Using PM10 Data Clustering: Focusing on the Case of Korean Particulate Matter Observatory

Seonghee Min and Yoosoo Oh

Recently, In Korea, interest in air pollution has increased. It provokes an issue about the risks of asthma and lung diseases. Especially, the particulate matter is a substance that causes respiratory diseases and cancer. The WHO reported the severity of it. Currently, particulate matter observatories of Korea are 353, those intensively located in Seoul and Gyeonggi-do. Local cities have difficulties in estimating the exact PM10 value. In this paper, we classify the particulate matter area to predict a distribution of the particulate matter by utilizing the PM10 data clustering. We divide the area of Korea territory through labeling data and center points of particulate matter observatories. We perform the labeling of particulate matter observatories through the clustering. Also, we divide the area of Korea territory by using the Voronoi algorithm based on the center point of the cluster. For the particulate matter area classification, we consider the density distribution of the PM10 values compared to the conventional administrative district. Besides, the proposed method can help to grasp the particulate matter distribution in shadow areas that have no observatory.

Volume 11 | 12-Special Issue

Pages: 959-965

DOI: 10.5373/JARDCS/V11SP12/20193300