Analytical Study On The Trends Of The High Pm2.5 Concentrations And Interregional Association

The standards for PM2.5 forecast grade were tightened after March 27, 2018 (bad grade: daily average more than [51 mu g/m(3)] -> [31 mu g/m(3)]). In this study, The characteristics of high concentration according to the tightened standards were analyzed from two perspectives. The daily mean data for each forecast areas over 3 years (December 2015 to November 2018) was used for analysis. First, The association between the forecast areas of the same time zone was analyzed using the conditional probability. On the whole season, Gangwon-east and Jeju had the lowest association with other regions. On the contrary, Seoul metropolitan area and Chungbuk were highly interconnectivity in spring, and Gangwon-west, Chungbuk and Gyeonggi northern were the most important areas that affected other regions in winter. Second, The association between the forecast areas considering time continuity using high concentration prediction models for each areas that was developed by logistic regression analysis. The high temporal association was found in most of the area except Gyeonggi southern, Daejeon and Gyeongbuk that means if high concentration occur in A area in today, high concentration will occur in A area in tomorrow. Based on the odds ratio of the prediction model, Gyeonggi northern was selected as the most influential area impact on the probability of occurring high concentration of the other areas, and Busan, Daegu and Gyeongnam were selected as the least influential region.