A24A-07:
Evaluating the Impacts of Transboundary Air pollution from China on Air Quality in the U.S. Using a Regression Framework

Tuesday, 16 December 2014: 5:30 PM
Nicole S Ngo, University of Oregon, Eugene, OR, United States, Xiaojia Bao, Xiamen University, Xiamen, China and Nan Zhong, Columbia University of New York, Palisades, NY, United States
Abstract:
China is the largest emitter of anthropogenic air pollution in the world and previous work has shown the environmental impacts of the long-range transport (LRT) of air pollution from China to the U.S. via chemical transport models, in situ observations, isentropic back trajectories, and to a lesser extent statistical models. However, these studies generally focus on a narrow time period due to data constraints. In this study, we build upon the literature using econometric techniques to isolate the impacts on U.S. air quality from the LRT of air pollution from China. We use a unique daily data set of China's air pollution index (API) and PM10 concentrations at the city level and merge these information with daily monitor data in California (CA) between 2000 and 2013. We first employ a distributed lag model to examine daily patterns, and then exploit a "natural experiment." In the latter methodology, since air pollution is rarely randomly assigned, we examine the impacts of specific events that affect air quality in China, but are plausibly uncorrelated to factors affecting air pollution in CA. For example, Chinese New Year (CNY) is a major week-long holiday and we show pollution levels in China decrease during this time period, likely from reductions in industrial production. CNY varies each calendar year since it is based off the lunar new year, so the timing of this pollution reduction could be considered "as good as random" or exogenous to factors affecting air quality in CA. Using a regression framework including weather, seasonal and geographic controls, we can potentially isolate the impact of the LRT of air pollution to CA. First, results from the distributed lag model suggest that in the Spring, when LRT peaks, a 1 μg/m3 increase in daily PM10 from China between 10 and 14 days ago is associated with an increase in today's PM2.5 in CA of 0.022 μg/m3 (mean daily PM2.5 in CA is 12 μg/m3). Second, we find that if CNY occurred 5 to 9 days ago, today's PM2.5 in CA decreases by 3 μg/m3. We also conduct other tests and sensitivity checks, like observing impacts from individual cities in China or other events, and using daily leads as a falsification test. Our results have important policy implications regarding the consequences of foreign pollution sources and suggest a causal relationship between pollution from China and air quality in CA.