A Development of Particulate Matter Forecasting System with Inverse Modeling using Source Contribution

PM (Particulate Matter) forecasting system using inverse model based on the surface air quality measurements in East Asia was developed and its performance was evaluated. The PM forecasting using inverse model showed a posteriori PM10 emissions derived from inverse model decreased in Asian dust regions (R01, R02, R03, R04) and Southwestern regions (R08, R09), but increased in urban and industrial regions (R05, R06, R07) in East Asia. In the Korean Peninsula, a posteriori PM10 emissions increased in most of regions except for Seoul and Incheon.The predicted PM10 without inverse modeling (CASE02) is underestimated compared to the observations, and the forecasted PM10 with the inverse model (CASE05) showed an good agreement with the measurements. The performance of PM forecasting model with inverse model also displayed that the forecasting index (Accuracy (A), Probability Of Detection (POD), False Alarm Rate (FAR)) was improved in most regions in Korea, compared to basic model without inverse model during 2017 and 2018. Therefore, the developed air quality forecasting model with inverse model was improved the forecasting performance of PM10 and can be proposed as a representative PM10 forecast model in South Korea.