Non-Parameter Statistical Analysis of Impacts of Meteorological Conditions on PM Concentration in Beijing

Based on monitoring data of particulate matter(PM) and meteorological data from September 2005 to September 2006 in Beijing,the influence of meteorological conditions on concentration levels of PM with different size ranges was investigated for four seasons using non-parameter statistical analysis such as Spearman correlation matrix.The results showed that PM concentrations were significantly correlated with one or more meteorological parameters in different seasons. PM_(2.5),PM_(2.5-10) and PM_(10) had strongly negative correlation with surface atmospheric pressure in spring.PM concentrations were influenced greatly by rainfall in summer.PM_(2.5) and PM_(10) had significantly negative correlation with solar radiation in autumn and winter.The concentrations of PM with three size ranges in winter were significantly and negatively correlated with wind speed,but positively correlated with ambient temperature as well as relative humidity.The concentrations of both fine particles(PM_(2.5)) and coarse particles(PM_(2.5-10)) were influenced by meteorological conditions,but to different extents.The concentrations of coarse particles in dusty spring and humid summer were more significantly correlated with wind speed than those of fine particles,so the PM_(2.5)/PM_(10) increased with stronger wind.In autumn,the concentrations of fine particles were more significantly correlated with solar radiation than those of coarse particles,so the PM_(2.5)/PM_(10) reduced when solar radiation lasted longer.The concentrations of coarse particles in winter were more significantly correlated with humidity than fine particles.With increasing humidity,the PM_(2.5)/PM_(10) got lower.