Application of Positive Matrix Factorization in Source Apportionment of Fine and Coarse Particulate Matter in Bandung City

Abstract:

A Receptor Model method of Positive Matrix Factorization (PMF) was used to identify the emission sources of fine and coarse particulates in Bandung, a city located about 150 km south – east of Jakarta. Samples of fine and coarse particulate matter (PM) were collected in a rural area of Dago Pakar in Bandung city during wet and dry seasons from 2001 to 2007. The samples were collected using Dichotomous Samplers to collect samples which were analyzed for black carbon and element concentrations and using Mini-Vol Samplers to collect samples which were analyzed for ions. The species analyzed in this study are Na, Mg, Al, Si, K, Ca, Ti, Cr, Mn, Fe, Ni, Cu, Zn, Pb, , Cl-, NO3-, SO4=, NH4+. The data set were then analyzed using positive matrix factorization model to determine the source of fine and coarse particulate in this area. The results produced seven source factors for fine particulate and 6 factors for coarse particulate during both seasons. The sources of fine particulate are sea salt, vehicles icluding motor cycle, biomass burning, volkanic dust, secondary aerosol (NH4)2SO4/(NH4)NO3 and soil Dust, while source factors for coarse particles are sea salt, industry, construction dust, volcanic dust, soil dust, and lime dust. The PMF results show that secondary aerosol, biomass burning and vehcles contribute more than 50% to the total fine particles in both seasons. Large fraction of about more than 50% of coarse particles come from soil dust and lime dust.

Key Words: Receptor Model, PMF, Emission Source, Fine particle, Coarse particle