The Interactive Effects Between Meteorological Factors and Air Pollutants on Circulatory and Respiratory Mortality in Beijing

Meimei Xu,Yunzheng Mo,Xiaochuan Pan
DOI: https://doi.org/10.1289/isee.2013.p-3-03-39
2013-01-01
Abstract:Background The interactive effects between meteorological factors and air pollutants were usually examined using generalized additive model (GAM) based on hierarchy in previous studies, but this may break the continuous. Aims We explored and compared the interactive effects between meteorological factors and air pollutants from continuous perspective. Methods Air pollution data (PM10, NO2 and SO2), meteorological data [temperature(T), relative humidity(RH), atmosphere pressure(AP) and wind speed(WS)] and mortality data were collected from 1 January 2006 to 31 December 2009 in Beijing. GAM was fitted with bivariate thin-plate smoothing spline to examine the interactive effect between meteorological factors and air pollutants on circulatory and respiratory mortality. The best model was chosen based on the least deviance criterion. The interactive effects were roughly compared by the absolute value of the estimates and visualized using two-dimensional contours. Results The daily average mortality of circulatory and respiratory disease in Beijing was 82 and 18 respectively. The relatively largest interactive effects for circulatory mortality were NO2-AP, followed by PM10-AP and T-AP. For respiratory mortality, the effect of T-NO2 was largest, followed by NO2–AP and T-AP. Conclusion GAM can be used to explore interactive effect between meteorological factors and air pollutants from continuous perspective. There was difference in the relatively largest interactive effect item for circulatory and respiratory mortality.
What problem does this paper attempt to address?