| The Chemical Mass Balance (CMB) receptor model was applied to the volatile organic compound (VOC) data acquired during the Coastal Oxidant Assessment for Southeast Texas (COAST) project in order to examine temporal and spatial patterns within Houston area. According to CMB results, motor vehicle exhaust is the largest contributor to nonmethane hydrocarbons (NMHC) at the Galleria and Aldine sites, while industrial sources are the largest contributors to NMHC at the Clinton, Baytown, Seabrook, and Port Arthur sites. Vehicle exhaust contributions are greatest during morning commute periods, and are significantly lower during weekend mornings. Biogenic emissions contributions are relatively low and follow expected diurnal variations. Average source contributions aloft were approximately one fifth of those found at the corresponding surface sites with industrial sources and motor vehicle exhaust accounting for the greatest contributions. Mobile source contributions from CMB calculations agree well with inventory estimates at the Galleria site, but are consistently higher than inventory estimates at the Clinton site. The sum of liquid gasoline, gasoline vapor, industrial and CNG contributions from CMB agrees reasonably well with the corresponding sum of point and area sources from inventory estimates, and contributions of biogenic emissions are consistently higher in the emission inventory compared to CMB estimates. Large temporal and spatial variations in source contributions suggest that the actual emission rates, particularly for point sources, are more variable in time than is assumed in the emissions inventory. These variations may have significant implications for air quality model performance and evaluation.; The periodic patterns in the source contributions were examined using a time series analysis (Fourier analysis) approach developed in this study. The Fourier analysis results agree well with the periodic patterns existed in the original data series. The procedure can be applied to other areas in which the finding of periodicities among data series is desirable. |
