Using social media for air pollution detection-the case of Eastern China Smog

Air pollution has become an urgent issue that affecting public health and people’s daily life in China. Social media as potential air quality sensors to surveil air pollution is emphasized recently. In this research, we picked up a case-2013 Eastern China smog and focused on two of the most popular Chinese microblog platforms Sina Weibo and Tencent Weibo. The purpose of this study is to determine whether social media can be capable to be used as ‘sensors’ to monitor air pollution in China and to provide an innovative model for air pollution detection through social media. Based on that, we propose our research question, how a salient change of air quality expressed on social media discussions to reflect the extent of air pollution. Hence, our research (1) determine the correlation between the volume of air quality-related messages and observed Air quality index (AQI) with the help of time series analysis model; (2) investigate further the impact of a salient change of air quality on the relationship between the people’s subjective perceptions regarding to air pollution released on the Weibo and the extent of air pollution through a co-word network analysis model. Our study illustrates that the discussions on social media about air quality reflect the level of air pollution when the air quality changes saliently.