UNDERSTANDING MISSING DATA IN REAL-TIME POLLUTION MONITORING SYSTEM IN CHINA

Abstract

Using both remote sensing data on air pollution and publicly reported hourly PM2.5 data from ground-level monitoring stations, this paper examines whether the quality of the publicly reported PM2.5 is affected by selective reporting whereby high-level hourly pollution readings are dropped in the reported data. Our analysis shows that the contemporaneous level of air pollution measured by the Aerosol Optical Depth (AOD) has a negative relationship with the frequency of data missing. This relation- ship is weaker in dirty cities measured by the average AOD during the sample period and is reversed in very dirty cities.