Temporal and Spatial Variation of PM10 Concentration and Its Influence Factor Analysis
可吸入颗粒物(PM10)浓度时空变异性及影响因素分析

The temporal changing trend, periodicity and spatial variation of PM10 concentration in Xiamen City measured in four different monitoring sites have been analyzed by using time series and simple statistics approaches including correlation analysis, principle component analysis and cross-correlation analysis. The results show the PM10 concentration had no significant increasing and decreasing trend from 2001 to 2002, but had significant periodicity (the most significant periods are 14 and 7 days for different monitoring sites). The spatial distributions of PM10 concentrations, however, are different with monitoring sites. The second PCA can describe the distribution difference exactly. The influences of meteorological factors, including air temperature, rainfall intensity, wind direction and rainfall days, on PM10 concentration have also been analyzed. Especially, the cross correlation analysis (CCA) was applied to estimate the influence of monthly rainfall days (MRD) on PM10 monthly average concentration (MAC). Meteorological factors can change the variation of PM10 concentration directly. The influence of MRD on MAC lags for about 2 or 3 months. Moreover, land cover types are also important to the spatial distribution of PM10 concentration. The local regions where the vegetation cover ratio (VCR) extracted from relatively remote sense data is high have relative low PM10 annual average concentration.