Source apportionment of sub-micron prague aerosol from combined particle nUmber size distribution and gaseous composition data by bilinear positive matrix factorization

Abstract

This study has identified and apportioned the sources of ambient aerosols in the urban atmosphere of Prague, the capital of Czech Republic, using Positive Matrix Factorization (PMF). Prior to this work, limited use of PMF technique has been applied to Prague's aerosol while elsewhere around the world, it has been actively used by aerosol scientists to reap the benefit since its first introduction in the year 1993. In the current study, PMF has been applied to the ambient particle number concentrations obtained using a scanning mobility particle sizer (SMPS) in the size range between 14.6 nm and 736.5 nm midpoint diameter with 10 min time resolution along with the ambient gaseous concentrations of CO, SO2, NOx (NO + NO2) and O3 obtained at the receptor site with 15 min time resolution. The concentrations of CH4, NMHC and THC were also measured but were not included in the analyzed data. The 15 min time interval meteorological data of wind direction (grad), wind speed (m/s), temperature (°C), relative humidity (%), and solar intensities in the UV-A and UV-B (W/m2) regions were also recorded to assist in the interpretation and to support the use of Conditional Probability Function (CPF) plots to estimate the directionality of the sources. For the convenience of data handling and matrix preparation, all of the data obtained were reduced to hourly concentrations. Although there was 5 min time shift between the gaseous and meteorological measurement intervals as compared to the SMPS data, it was assumed that the atmospheric conditions did not change appreciably and the data were uniformly normalized to one hour time intervals. The precipitation during the sampling period was also recorded using LNM Disdrometer. The use of PMF as implemented in the program PMF2 identified 5 factors: motor vehicles, heating, ozone influenced secondary particles including two unidentified mixed sources I and II Further research is in progress to identify these two sources that may represent long range transport or background sources.