Modified Inverse Distance Weighting Interpolation for Particulate Matter Estimation and Mapping

Abstract

Various studies are currently underway on PM (Particulate Matter) monitoring in view of the importance of air quality in public health management. Spatial interpolation has been used to estimate PM concentrations due to that it can overcome the shortcomings of station-based PM monitoring and provide spatially continuous information. However, PM is affected by a combination of several factors, and interpolation that only considers the spatial relationship between monitoring stations is limited in ensuring accuracy. Additionally, relatively accurate results may be obtained in the case of interpolation by using external drifts, but the methods have a disadvantage in that they require additional data and preprocessing. This study proposes a modified IDW (Inverse Distance Weighting) that allows more accurate estimations of PM based on the sole use of measurements. The proposed method improves the accuracy of the PM estimation based on weight correction according to the importance of each known point. Use of the proposed method on PM10 and PM2.5 in the Seoul-Gyeonggi region in South Korea led to an improved accuracy compared with IDW, kriging, and linear triangular interpolation. In particular, the proposed method showed relatively high accuracy compared to conventional methods in the case of a relatively large PM estimation error.